Writing is fundamentally the transfer of information from your brain to my brain. If you have 1000 bits of semantic information you want to transfer, you can't give 300 bits of semantic information to an LLM and have it fill in the remaining 700, because it doesn't know what those 700 bits are. If it's able to guess those 700 bits correctly, then they aren't true semantic information, and you really only have 300 bits you want to transfer. You might as well transfer those bits to me directly, rather than having the LLM add on an extra superfluous 700 bits that I then have to filter out.
For a long time after the internet arrived on the scene, a lot of online news stories would reference websites, papers, polls, etc. without linking to them. There are still news sources doing this today. Sometimes such articles interpret or place context around their hidden references, but a lot of the time they just summarize.
Giving someone the text output of a LLM is very similar to publishing a summary without links to the referenced material. When you were querying your LLM, you could have asked specific questions or asked for a custom focus or point of view. Your intended audience might have questions or different concerns, but they're unable to interact with your LLM. What you have delivered is static and unresponsive. It has all the disadvantages of being machine output without the advantage of being interactive, the way your LLM was for you.
It may have to wait until compute is cheap enough that tokens are essentially free, but we need a system to pass "hyperlinks" to LLM's primed with context, ready to be interactively queried on a chosen context. It's being overly generous to assume that people are putting even 300 bits into a LLM for every 1000 bits of regurgitated writing they try to pass off as their own. When people post LLM output as if it were their own, I have no choice but to assume they had zero knowledge of the subject, but this query taught them what they wanted to learn, and now they're sharing that. That's fine, but please pass an interactive LLM link rather than static text.
Once we have "hyperlinks" for LLM sessions, perhaps we can share LLM output a little more usefully and honestly.
I've seen professional journal pieces refer to science journal articles only to go and read the original article and find that it draws a different conclusion than what is implied by the journalist.
My father used to complain about that 40 years ago, though in his case he was reading newspapers rather than professional journals. But he'd point it out to me often enough that I started to see the pattern. Scientist publishes paper saying "We may have found evidence of X, which suggests the possibility that Y may also be occurring". Journalist: "Scientists find X which proves Y".
This has been happening for decades; I still see it happening today*. My cynical suspicion is that words like "maybe" and "suggests the possibility" don't sell enough papers.
* Worst offender I can remember was actually from the summary of a paper published on the research institution's own website, so I couldn't blame it on "Oh, the journalist misunderstood what the scientist wrote". Summary said "Exposure to X can, on average, cause a 40% higher chance of Y" (where Y was a negative health outcome). I clicked through to the study and read it. Turned out the confidence interval on that chance of Y was so wide, all you could say with 95% confidence was that exposure to X could do anything from reduce your chance of Y by 5 percent, or increase it by 85 percent, or somewhere in between. They had averaged -5 and +85 to get the scarier-sounding 40% number that they published in the summary, but the truth would have been far closer to "this confidence interval is so wide that we really can't conclude anything from this data". But that wouldn't be nearly as likely to get them grants, so they tortured the data in their summary so that it would look better.
In most cases, they are only useful to other scientists in their field, which does mean that they are useless to the general person until they show up in the form of a new commodity.
There is also incentives to adapt the message to the outlet. If you send that data to peer review and say we found 40% increase, the reviewers would reject it, so they have to moderate themselves. But if you send a summary to the university’s outreach outlet saying that we found something or nothing we don’t know, then they would also reject putting it out. So even for the authors, the incentive is to send a careful conclusion to peer review and an overblown one to popsci.
Btw, I also think a 95% confidence interval is just the wrong statistic to look at given that data, and that they could probably have analyzed it better.
> Worst offender I can remember was actually from the summary of a paper published on the research institution's own website, so I couldn't blame it on "Oh, the journalist misunderstood what the scientist wrote".
It's a similar thing. We live in the "attention economy", and research institutions - particularly after the US President openly went and had his minions cut funding to research purely on ideological reasons, but it's been a problem for decades - are just as susceptible to blow stuff out of proportion to make headlines and thus increase the chance someone might throw some money over the fence.
And media does the same, just to manufacture artificial debate. And so do politicians.
And frankly, I'm fed up with that, we will drive ourselves into a wall.
> so I couldn't blame it on "Oh, the journalist misunderstood what the scientist wrote"
It's never the case that someone misunderstood what scientist wrote. Much like the scientific papers, news articles, including those reporting specifically on the discovery, have their own goals, and the paper being cited is used as evidence or argument for article's own "study". Except for press, the standard is rhetorical, not scientific, it's the conclusions and not the methods that are "pre-registered" at the start, and claims are defended by "hey it's just a point of view", not by statistical significance.
In your own example of worst offender: the scientific study was trying to establish and quantify the connection between X and Y. The summary article was trying to push the angle that "this institution is doing important work". It started with that conclusion, and the paper cited was just the first thing the author found that could be easily massaged into supporting that conclusions by rhetorical standards.
Same paper might get cited by journalist trying to push for "X is bad for you", and they'll do roughly the same as the summary article. And, same paper may be cited by someone claiming they have a miracle cure for Y, and they'll make a honest observation that "absence of X reducing Y is a common bullshit claim based on misunderstanding the paper [citation], that actually shows there's no correlation there, I mean look at the confidence intervals, even the author says that in text nobody bothers to read"... - citation may be honest, but the article itself is still using it to prop up a different flavor of bullshit.
TL;DR: don't believe news. It's bad for your mental and physical health (p<00.05).
> It's never the case that someone misunderstood what scientist wrote.
This is wrong. Reporters frequently don’t understand the science or the nuance in the science.
Reporting and science are two very different disciplines. Reporters rarely have a deep background in science and almost never have a background in the specific area that they’re reporting on.
Hell, even scientists have trouble accurately describing the work of a different scientific discipline.
Don’t invent bad faith motivations; they exist but most of the time it’s just two people slightly talking past each other.
> Don’t invent bad faith motivations; they exist but most of the time it’s just two people slightly talking past each other.
I'm not inventing them, but maybe conflating two sources:
1. Malice directly intending to hurt or defraud people. Probably not as common as how I make it seem.
2. Not caring. Well, I subscribe to the view that not expending effort to be accurate when talking to other person is as bad as slashing their tires (paraphrasing an old quip), so I very much consider bullshitting and picking a conclusion and then massaging facts to fit it, to be acting in bad faith too.
Nah, even in middle and high school it was plain that people sometimes misunderstood what the teachers meant. It's not like that goes away in adulthood; even when people "care", they still often misunderstand nuance or details, and sometimes even the bigger picture.
Honestly, it's plain weird to say that people never just make mistakes.
PS - worth adding that "I misunderstood" and "I didn't care enough" are not mutually exclusive. You can do both, so saying "they didn't misunderstand, they just didn't care" isn't a reasonable rebuttal. But even setting that aside, there'll be plenty of folks who care but still don't understand.
I'm gonna invoke Hanlon's handgun here: not attributing stupidity to what's adequately explained by systemic incentives promoting malice.
I normally assume people make mistakes. I don't believe this is a good explanation for news publications, university press releases, politicians, etc. because those are organizations with agenda, and commit "misunderstanding" of this type pretty much in every thing they publish. The pattern here is pretty conclusive, IMO.
> I normally assume people make mistakes. I don't believe this is a good explanation for news publications, university press releases, politicians, etc. because those are organizations with agenda, and commit "misunderstanding" of this type pretty much in every thing they publish. The pattern here is pretty conclusive, IMO.
The pattern of personal motives fits misunderstanding. You need to show that there is an organizational pattern of "malice" (your word, not mine), rather than an organizational pattern of "we are trying to publish quickly, and quality accidentally falls to the wayside". I.e., negligence, not malice.
You haven't provided even a shred of evidence suggesting there's malice at the journalist level. Every science journalist I have met genuinely cared about the science (which is why they were writing on it), but they didn't have time to learn enough about the subjects to understand they were oversimplifying things.
I have recently reviewed a paper that referenced my own article… but the conclusion was so off, I actually went back and re-read that entire article just to be sure there was no hint to the conclusion that the author derived. There was none. Uncanny experience.
Not just news stories. This is a huge problem with social media and forums in general too. Lots of people making bold claims about random things, lots of stories they say are based on a third party source, but very few actual links to said sources in question.
I still remember a recent example where one of those trivia accounts on Twitter posted an interesting story about some guy whose life completely changed after an accident, but neither linked to a source or named the person in question.
The only way I was able to verify it was true was through someone in the comments asking the platform's AI chatbot, and the chatbot providing context that I could research and verify...
> For a long time after the internet arrived on the scene, a lot of online news stories would reference websites, papers, polls, etc. without linking to them. There are still news sources doing this today.
I agree, this drives me crazy. Ironically, one of my favorite uses for Claude is to ask, "What study is this news article talking about?"
It's pretty good at digging up the source and related sources. And most of the time, if you read the source, the article is nonsense and gets everything wrong.
Link rot has accelerated in the past decade. Links are great, when they work. If they are a few years old, they often don't. You certainly can't count on it. If you're referencing a paper or an article, it's probably better to cite the title of the article, the name of the journal or magazine or newspaper it was published in, the author, and the date. Then someone might have a chance of finding it again.
> For a long time...a lot of online news stories would reference websites, papers, polls, etc. without linking to them.
The "$CITY_NAME Business Journal" websites are the absolute worst with this. They'll refer to something specific, for example "$BIGCO's 2025 10-K filing" and it will be a link. That link will go to the 10-K, right? Nope! It goes to another page at the same business journal. Maybe that page is a summary of the 10-K, but probably not. Maybe it's just the general index page for all the articles about $BIGCO at that journal. What it links to, it definitely won't be the specific thing described by the text of that link.
It isn't the transfer of information at all. What's actually happening is you're prompting experiences in a human instead of an AI using text.
Communication only works if you have multiple levels of representation and abstraction, including but not limited to - letter shapes, grammatical structures, style and register, stylometry, and subtext.
All of that is learned, and writers usually assume they can rely on that learning as the context for the text.
So you don't write to 'transfer information' like a network cable, you write to trigger experiences in the human version of latent space.
Factual information is one kind of experience. But even when that's the goal, there are always layers of implied relationship, social register, role, status, and other implications in everything that's written.
In normal communications the context - business emails, personal messages, mainstream journalism, fiction, and the rest - defines what acceptable language looks like.
The content fits inside that. But it has to fit the context, otherwise it lands in a semantic and psychological uncanny valley - like sending LinkedIn speak to a spouse on a wedding anniversary.
The real problem with LLM writing is that it's good at the technical layer - the grammar and spelling - and has some insights into the rest.
But the default content style is marketing and ad speak. And recently it's developed a weird and unique hybrid style which applies marketing fluff and pretension to technical content like code comments.
So you get one register instead of all of them. It can attempt others, but it's still too limited to generate them fluently. Sometimes the results are outstanding, but often it defaults to mechanical clichés.
So that's why it sucks and sounds so hollow.
Can it be fixed? Yes, but it's very hard work, most people don't have the skills, and it takes time - often too much time to be worth the effort.
I'm definitely being slightly too emphatic when I say it's "fundamentally the transfer of information", but I don't think that nuance is important.
When LLMs eventually get good at writing in the correct style for a given context, I'll admit that they have value in that way. But they aren't good at that yet. And even when they do get that good, I'll still dislike it for reasons that are more emotional than rational.
I don't care how good the LLM gets. If I know some text was written by LLM I'd much rather know the prompt - the seed of intent.
If the seed of intent is "convey XYZ details so they know them" then I can choose to go and learn those details any way I see fit - maybe even ask an LLM to summarise some data for me! - rather than having to ingest whatever their LLM use poops out and trying to digest the intent and content and figure it out.
When sharing summaries from calls that I attended but my team has not but I think they would find interesting and I simply don't have time to handcraft it myself I use LLM and attach my prompt and the transcript.
Attaching the prompt initially threw some people..."wait, your admitting to using AI..."..."err yea, unlike you with that PowerPoint you sent me last week". I sense this is the right way to go imho.
Output is not interchangeable with the prompt. In many cases, the prompt does not have the information the sender wanted to give you, and there is no guarantee that your LLM will give those information - or do it correctly - if you use the prompt yourself.
The entire value of here is that sender read the output and is vouching for it. This is where "bits of information" come from. If the sender cannot be trusted to verify and vouch for the LLM text they're sending to you, well, they're an asshole and you should rebuke them or find someone more considerate of others to talk with. Them giving you their prompt doesn't help you with anything.
I am focusing on the thing that is usually hidden. You are correct that for a full picture I'd need the information the LLM operates on.
But my point is that the prompt is the nearest encapsulation of the intent of the originator. Give me the data and the prompt. Vouch for the output if you like, but I want the source.
This is what's so amusing to me. People say they're not interested in the output of an LLM, only what a human has to say. But then when a human says "These words from the LLM are good, I vouch for them" the very same people say "If I wanted those words, I'd get them from the LLM myself, what's the point of this human at all?"
While I agree with some of what you've said and the conclusion you've arrived at, in the end, I think you've missed part of the picture here with regard to "prompting experiences".
> Communication only works if you have multiple levels of representation and abstraction, including but not limited to - letter shapes, grammatical structures, style and register, stylometry
These are methods of encoding, there's no reason all of these can't be represented in an LLM from a technical point of view.
> and subtext.
This is the other half of the equation to me. Humans communicate by relating shared experiences, an LLM cannot have shared experiences. While it might be able to encode subtext that has been specifically called out and explained, it will never be able to encode the breadth of human subtext, especially that which is reliant on emotion.
I don't believe it is possible to change this until the point mankind truly develops a "wetware interface" to the digital world (and I personally don't want such a thing to exist).
That "marketing style" isn't a style, it's the lack of content itself. You just restated their original point despite trying to object to it, because it was correct and there is no way around that.
Sometimes Claude's problem, such as when I ask it to summarize a long, complex session back to me, is it's too information dense. It uses weird invented terms to gloss over complex parts of the architecture instead of explaining them.
But no matter what - too dense or too sparse - it always sounds like Claude.
It misses the forest for the trees. It feels the need to highlight details not understanding what details are most relevant to a human reader and how to survey the larger problems in a cohesive way that emphasizes the right parts without cliche and undue emphasis.
What already happens:
- People give an LLM a bulleted list of points that they want expanded into a professional sounding document.
- The receiver doesn't wanna read all that. They put the full document into an LLM and ask it to summarize it into succinct bullet points.
This is exactly what I want: To communicate with me, have your LLM expand your message to include relevant parts of _your_ context, then I will have my LLM summarize it as briefly as possible with respect to _my_ context.
I first noticed this a year ago when I read a gmail AI summary, then glanced at the main text and saw it was AI generated. I feel like there's good fodder in there for a dystopian sci-fi story about a future where nobody communicates directly with one another, it's all AIs translating, but the AIs slowly start to drift.
It's essentially the tower of babel. Each person will devolve to speak their own internal language only they understand. Each language will need to be encoded down to its meaning to be reinterpreted. None of us will know if the transformers are accurately decoding, or if the other person is accurately interpreting the decoding (which is arguably already a feature of human language without the computers in-between.)
I feel like a lot of this is a problem when someone technical is attempting to communicate a complicated technical subject to a less-technical audience.
I can only dumb a thing down so much before the description is useless (when you zoom out too much you lose the details). Even technical people who could understand it but are lazy / "in a hurry" use the summary, without thinking about what detail they are losing.
Even more infuriating is when they then reply to my email, having only read the AI summary, and ask a question that was already answered by my message.
This is the exact same thing that happened pre-AI, with the added step of wasting energy/resources on the AI summary in the middle.
It was in a gmail promotional video years ago already. Person A would use AI to turn their summary into email and person B would ask the AI to go back to the summary.
They knew it was going to be like that from the beginning.
I’ve been saying the same thing. I’m not saying all, but a significant amount of comms could be bullet points to the benefit of both sender and receiver.
It's worse than that. If it takes others longer for others to consume and understand what you're producing than it does for you to produce it, you'll never be able to communicate with someone efficiently. Communication breaks down at a fundamental level if if you can't keep up with the other side if outputting and they won't slow to allow you do do so.
I see this all the time now with LLM generated output. It's easy to have an LLM generate a chunk of content that can be dropped into a chat or comment, and when it took you 20 seconds to have something written up based on the shared understanding you and an LLM have about the context of the situation, but it takes other people 3-5 minutes to read and understand that content, that fundamentally doesn't scale. It's bad enough when one or two people are doing it, but if the whole team is doing it, the only way to keep up with the stream of information is to also consume it through an LLM. At that point you're likely to be missing much of the nuance, and the amount of errors will explode.
This can be alleviated by people reviewing the output of an LLM and making sure it both includes fundamental information that might be assumed by context and reducing it to the parts that are essential for the new context it's in. This takes time, but is extremely important.
Having an LLM write gobs of text to send to other people instead of doing it yourself is the equivalent of a low yield cognitive zip-bomb. Don't do it.
I don’t quite think this tracks. Perhaps you want to communicate 1000 bits that are well known and can be referenced with a 300 bit key. Then the LLM can easily retrieve the remaining information. It’s like sending someone a link to the Wikipedia page instead of explaining something yourself.
No, I don’t want to read LLM writing because it is BAD at it. It doesn’t really understand how humans think (because it thinks differently), and doesn’t seem to understand core principles very well (presumably due to the lack of world model), so it can’t write something humans enjoy yet.
That's kind of my point. If the 1000 bits are well known, then their inclusion isn't new semantic information. By pasting an LLM's output, you're deciding for your reader that they don't already know that information, and deciding that your LLM prompt is better than whatever they would to to obtain that information if they lack it. IMO, it's much better to give your readers the 300-bit key, and let them decide for themselves if they want/need to get more information, and if so, how.
Mmm perhaps, but I think you normally have an idea of who your audience is. If this was actually how we talked then I would have just responded to this comment with “known-audience rebuttal. Example. Audience knowledge clarification. Example absurdity comment” and let you work it out. I think it’s normally known to you and the LLM and not known to the person you’re speaking to. Speaking like that would just cause confusion and misunderstanding.
I disagree. Unless you gave additional information to the LLM yourself, the LLM doesn't know more than your audience does about what the meaning of such a comment would be. An LLM could certainly come up with something plausible, but it wouldn't necessarily be what you intended.
Why not do both though? I often now write my human summary, then paste in also the content you could get by asking a bot (with markers for which is which). You’re perfectly welcome to ignore the bot text, or ask your own bot, but bots are pretty slow, so I also don’t want to wait for it to “decompress” that 300 bit blob to get the detailed page back. But my human summaries are also only intended to be 50 bits — compressed again to just pass the signal I mean to convey, and not the whole prompt needed to fetch the right info from the right place to substantiate the claim.
TLDR the length was the same curtesy of tl;dr before, just now with a different name.
If only that were the HN we comment on. If a post doesn't spoon feed an explanation for an initialism thats the tiniest bit off the beaten path (eg BGP), there's inevitability a comment about how would it kill the author to define BGP?
I’m not sure the 300-bit → 1,000-bit framing applies in all instances. The 300 bits may be a compressed cue to a much fuller idea. The AI can combine that cue with its prior knowledge to help reconstruct what the prompter was trying to express, with the prompter then verifying whether it’s right. Without the relevant prior knowledge for reconstruction, or the prompter for verification, it becomes much harder to know whether you’ve reconstructed the intended idea.
Unless that 700 bit was transferred on a separate occasion the inferred 700 bits is not true information, anyone could have reconstructed it from the 300 bits.
Not anyone, no. From an information theory standpoint, that it's possible at all to complete these 700 bits, only implies that anyone logically omniscient could. It's entirely possible that an LLM is capable enough to infer these 700 bits, and the human reader isn't.
Right but where is this information coming from, it can't come from the sender because an idea that can be conveyed in 300 bits cannot contain 1000 bits of information. You're just using the LLM to translate the idea into something more legible.
Alternatively those 700 bits are information the LLM added, but where is that information coming from? Is it noise? Random facts? Random lies? And who is the receiver even talking with if most of what they read is something the sender didn't know?
Even Markov chain autocorrect tools do better than 50% odds*, and even GPT-2 was significantly better than that kind of autocorrect.
* at the word level; IDK how redundant/efficient language is when it comes to bits-worth-of-fact-claims-per-word. But "your cat is sitting on my" -> [mat, laundry, roof, head, belly, laptop, microwave, …] clearly has many bits of information, and a Markov chain will encode the most likely next word even if the user doesn't know what the most likely next word is. Verifying where the cat is sitting is also very easy, as is correction.
The first sentence in this comment contains 32 characters; from the point of view of a naïve channel with no compression, that's 256 bits (given none require breaking out of the first bytes of UTF-8).
It did not take 2^256 attempts to construct the first sentence in this comment, because the generation process was not flipping coins per bit.
LLMs also do not emit bits chosen with a [0: 0.5, 1: 0.5] probability distribution.
From a compression point of view, the bits-transmitted-per-bits-in-message ratio can be reduced such that more likely messages use fewer bits than less likely messages. However, this requires the receiver to agree with the sender what the probability distribution over tokens is.
Intelligence is, amongst other things, a compression algorithm. If I can predict your next token, and we both know this, we can agree in advance that you don't need to actually send it.
No single human brain is able to predict the output of an LLM anything like well enough to do that.
In entropy terms: LLMs are noisy sources, their output does contain false statements, yet they add more bits of signal than of noise relative to a human alone.
Or at least, they can add more add more bits of signal than of noise relative to a human alone, but humans who blindly copy-paste the output of an LLM without checking are a pain and add zero value to whatever situation they happen to be in.
For some hypothetical scenario, writing software because I know they can do that, asking an LLM to write some code for you may easily give you 10 kilobits of positive information (code that mostly works), and -30 bits of noise (each bit being one binary decision's worth of incorrect choice by the LLM in what to write, i.e. bugs); if you as a user don't know how to handle the -30 noise that could easily be a totally useless app, but if you can filter out 30 bits of noise, either manually because those 30 bits happen to be your skill set, or even in some cases by prompting it again with the failure mode, then you get to benefit from the 10 kilobits of good stuff that you didn't have before.
In many (but not all) cases, LLMs can fix more than 1 bit of mistakes per follow-up prompt.
Far more than 1 bit for most attempts: they need that just to be able to write coherent sentences, and a lot more to be coherent sentences on the right topic.
Some specific conclusions would be far less than 1 bit.
The average will depend on both the question and the AI.
LLM is not a random symbol generator (hint: training data is not random), and no reasonable person is going to just prompt an LLM and send its output without giving it at least cursory check (at the very least so that blatantly stupid hallucinations don't paint the sender as inconsiderate or incompetent).
That check alone can add bits to the final signal.
I'm not sure we're working in the same framing here. For one LLMs are random, sure you can fix the seed but you don't have to, you can even replace the RNG with true random noise.
So there are 2^300 possible ideas, only 2 outcomes from the cursory check, how do you get 2^700 outcomes? Most of those are just random variations the LLM added which is not a transfer of information. You would be lucky to even identify which of the 2^300 ideas was being conferred.
"Random" is too loose a word, but they were responding in a context where it meant coin flips.
LLMs are not even odds on all possible outputs, they are biased towards patterns which are upvoted by the training mechanism (at a minimum: the source material, RLHF, and synthetic data).
The information any trained model transfers to output, is information it gained during its training.
No single human is capable of having consumed all that training data.
> and no reasonable person is going to just prompt an LLM and send its output without giving it at least cursory check
I'm reminded of an old quote:
The reasonable man adapts himself to the world: the unreasonable one persists in trying to adapt the world to himself. Therefore all progress depends on the unreasonable man.
Yup, it helps when the prompter reads and edits the AI-generated text, but usually they just skim it and send it unedited. Worse, it's usually not 300 bits -> 1000 bits (= a concise message covering all the relevant topics), it's more like 300 bits -> 3000 bits (= typical LLM verbal diarrhea hiding the relevant points in a wall of text).
This is true - but also, models are absolutely terrible at writing articles and I don’t want to read them.
The issue isn’t that a 300 bit idea is padded with 15 KB of content. You can take any human-written article and reduce it by 90% with next to no information loss. What you lose is what makes the article a compelling read instead of a fact table.
I think the reality is that we will see quality long form AI-written content at some point. It doesn’t even feel like labs are particularly interested in chasing that now; code sells way more tokens. Right now the trend is that subsequent models degrade in writing quality as long as that pulls them up on coding benchmarks.
> You can take any human-written article and reduce it by 90% with next to no information loss.
You've unintentionally circled the error here. The "purpose" of an article extends beyond "convey this essential information".
By analogy, a textbook contains far more words than a spec sheet, but attempts to train the human to be able to easily interpret spec sheets. The so-called "information" content of both might be equivalent, yet one does a better job of teaching students.
I was going to write a post disagreeing with this on the basis of the fact that the reader lacks the background information the LLM has. For example, if I were to prompt "explain the proof of quadratic reciprocity using Gauss sums" most readers would need the entire LLM's answer (and much more, probably) and not just the prompt.
But then I realized that the reader can prompt the LLM with the same prompt for the same or equivalent expanded text. Most people don't do this as it's extra effort, but it's interesting to imagine a world where this is the default way of engagement with a text, assumed by both writers and readers alike.
> But then I realized that the reader can prompt the LLM with the same prompt for the same or equivalent expanded text.
That's basically what I've been asking my colleagues (so far a losing battle): Please don't send me AI-generated text. Send me your prompt instead. It is highly likely that I will understand it without needing an LLM, and if not, I can do it myself.
No, thanks. Depending on the type of information being transferred and just how uncertain the recipient's knowledge is, I may or may not need to add some amount of information to whatever I have to share. That information would be extra context you may need to have, on top of what new thing I am trying to explain. If I know I'm talking with someone who's all caught up, great, no need for context setting. Though even in those cases, there might be some terminology I want to define, or to provide a list of acronyms or whatever. That is boilerplate work, but is at least useful, if not needed.
I know people tend to view this sort of comms as "technical" - they shouldn't be. So much misunderstanding happens because people are just not aligned on their shared context or medium.
Of course, AI is not fully reliable to generate that context, you as a writer need to put in the work to review and align the result with your own understanding, or drive it thoroughly before.
All that being said, I acknowledge that people (me included) love to be sloppy in their comms (with or without AI) and then blame others for misunderstanding. It's also unlikely we'll change soon. What can you do.
I’ll often put a long stream of consciousness on the page, or jot down rough meeting minutes, then ask ChatGPT to “summarise this for an email”. The result is shorter, clearer and easier to read.
AI amplifies the habits of the person using it. If they’re lazy or dim, then it's like giving a monkey a gun.
I like this example and it made me think, there is an analogy here to spec-driven development and vibe coding.
Rather than send 300+700 bits, like you said, send 300 (or less!) and let the human intelligence on the other side generate the result. Which supports the even older perspective: “If I had more time, I would have written a shorter letter.”
I’m not sure if this lands on anything very profound, but what about a pattern where, instead of codifying agent output at all, the only artifacts we share are the prompts. And the rewards (respect) accrue to those who generate the most generative among people and AI
So I spent a full workday investigating an issue with an agent and then I spent another hour prompting the agent to summarize the findings so that I can send them to a colleague. How does what you're saying come anywhere close to describing my work?
> If you have 1000 bits of semantic information you want to transfer, you can't give 300 bits of semantic information to an LLM and have it fill in the remaining 700, because it doesn't know what those 700 bits are.
LLMs do inference or computation among other things, so the remaining 700 bits can be something like that. The hidden implication in your claim is that computation adds no information content, which leads to an interesting philosophical discussion.
So for example, if I ask an LLM to give a proof or derive a new theorem from a set of axioms, according to your assumption, if it answers correctly, then I haven't learned anything new.
I am not really sure how to resolve this paradox in information theory.
If you input 300 bits into an LLM, and it outputs an additional 700bits of new information, I’d still rather you tell me those 1000bits than read ann llm’s output of 1000bits.
The primary reason is that human language is becoming a proof of work, that speaking out loud or writing directly indicates that the idea is important enough for a human to express. This is more costly than llm output, which is often just botspam.
From a purely information-theory perspective, the simplest solution is to say that yes, any content derived from existing information carries no information itself.
From a realistic perspective in the context of people copy-pasting LLM output, my thoughts are that asking an LLM to research for you is more defensible, but it's still better to read the LLM's research results and write the important parts in your own words (partly because the LLM probably used way more words than necessary for the context).
> Writing is fundamentally the transfer of information from your brain to my brain. If you have 1000 bits of semantic information you want to transfer, you can't give 300 bits of semantic information to an LLM and have it fill in the remaining 700, because it doesn't know what those 700 bits are.
Sure you can. LLM doesn't know what those 700 bits are, but you do. You may not realize it, and may not even know it at the time of prompting, but you do by the time you're sending.
Typical case is like this: you have 500 bits of semantic information to transfer. You give 300 of them to LLM, and get back the 500 bits you knew you have, and extra 500 you can quickly confirm are correct and relevant. Some of them are just dereferences of your input - where you recalled a pointer, but not what it pointed to. Some of it is information you never had before, but are able to easily validate.
You send that to me. I likely immediately realize the message was AI-assisted, but I trust you to be a decent human being, and not an asshole that lobs unverified LLM vomit over the fence for others to deal with. End result: you communicate 1000 bits of information to me, instead of planned 500, and you yourself learn extra 500 bits.
This is the optimistic scenario, but it does happen when LLM operator is not an asshole.
(Excuse the strong language, but I spent a lot of effort every day on both dealing with inconsiderate people lobbing LLM output at me, and making sure never to act like one myself, so it's a topic close to my heart.)
I think you missunderstand. If the llm can "add back" 200 bits of information, these 200 bits are superfluous by definition.
They can quiet literally be inferred from the starting 300 bits. And the llm has proved it.
If the receiver wanted these 200 added bits, she could infer them either herself or even use an llm to do it.
In principle, all of mathematics can be inferred from the axioms. The field of mathematics is not superfluous and something a receiver could infer themselves if they wanted it.
My brain does not contain all the information that can be added by an LLM; a human brain could contain it, even the biggest LLMs are about 1% of the (if you approximate synaptic count ~= parameters) parameter count of a human brain, but none actually will.
What my brain may actually contain is the information necessary to verify the (in this example) more than 200 bits the LLM claims to have added and trim out the parts which are false, retaining the (in this example) 200 "new" bits of new information added by the LLM.*
Concrete example: I am a software developer by training, though not a web developer. If someone who does not have any developer experience asks me to make a web app, I am forced to use an LLM as I do not know enough JS etc syntax to get it done myself. But as we all know, LLMs are only "ok" but not "good" at making software, so there are a lot of rough edges and outright mistakes. My experience as a software developer extends to detecting such failures and I can usually correct them.
The original person, someone who has no developer experience, can also prompt the LLM. Right now, this would result in something that retains all the errors, because they didn't have someone like me intermediating between them and the LLM.
I add bits by removing noise, the LLM adds bits but they contain noise.
I do not know for how long this will remain true, but today it is true.
* Feels like P versus NP to me. The answers AI generate are at their best when they're easy to verify. Then again, when they're easy to verify, they can be RLed to get good at this quickly and the need to verify goes down, leaving them still pretty bad at things that are hard to verify.
> If the llm can "add back" 200 bits of information, these 200 bits are superfluous by definition.
No, they're not. Getting those bits takes energy.
SOTA LLMs know way more than any individual on approximately anything there is to know (and what they don't, they can look up faster than people can). It's very easy for them to make the "missing" 200 bits explicit, rather than implicit, which in practical terms is the same as adding 200 bits that weren't there before.
Theoretically, an idealized omnipotent mind / AGI could derive the unifying theory from reading your HN comment on a phone screen. There is enough information there, if you were able to extract every bit of evidence available from it. But you are not. Neither am I. It would take us practically infinite work to try, solving this most cruel mathematical riddle.
How often do you modify the LLM output before sending it? If it’s less than 50% of the time, it means that by not modifying it, you have added at most one bit of information to what you originally wrote. (If you don’t understand why, think of it this way: Instead of sending the LLM response, you could send the prompt and one extra bit indicating whether the LLM response to the prompt should be modified, followed by the modifications.)
> How often do you modify the LLM output before sending it?
Me specifically, I never send anyone LLM output I haven't give at least a quick read (not skim, read) to make sure it's reasonable and there is no obvious bullshit there. And then I still mention it's LLM-sourced.
> If it’s less than 50% of the time, it means that by not modifying it, you have added at most one bit of information to what you originally wrote. (...) Instead of sending the LLM response, you could send the prompt and one extra bit indicating whether the LLM response to the prompt should be modified, followed by the modifications.
It's not the case, though. Prompts are not interchangeable with output. There is no guarantee that if you send a prompt, and recipient passes it to their LLM, they'll receive anything similar to what you did. It may have mistakes - different mistakes - or just spend focus differently.
The extra bits I claim LLMs can add to the message hinge strictly on you vouching for the response. Of course, you can just prompt an LLM, learn from the response, and then write your message clean, containing both the bits you originally had, and the bits you gained. But at that point, the LLM already gave you text containing all those bits - if you can vouch for it, you may as well copy it over and save yourself the trouble.
> It's not the case, though. Prompts are not interchangeable with output. There is no guarantee that if you send a prompt, and recipient passes it to their LLM, they'll receive anything similar to what you did. It may have mistakes - different mistakes - or just spend focus differently.
I’m not saying that the response is interchangeable, but that due to the data processing inequality, it cannot convey strictly more information than the prompt.
> The extra bits I claim LLMs can add to the message hinge strictly on you vouching for the response.
My argument is that if you vouch at least 50% of the time, the vouching only adds one bit of useful information – either you vouch or not.
> due to the data processing inequality, it cannot convey strictly more information than the prompt.
Only in the case where the LLM message is not reviewed before sending, and only if we assume reliable LLM (so that the receiver could recreate the same output if given the original prompt). This is not a realistic scenario.
> My argument is that if you vouch at least 50% of the time, the vouching only adds one bit of useful information – either you vouch or not.
The alternative to vouching isn't "not vouching", but "correcting and cutting out wrong bits and vouching for the rest", which means the single "vouched for it" adds all the bits that are in final message but weren't there in the prompt.
> Some of it is information you never had before, but are able to easily validate.
This is exactly the use case an LLM might (huge emphasis on might, depends on workflow, agentic vs. relying on contextual which can hallucinate) be good at and yet humans are notoriously bad at, because we are swayed by emotional responses and it is easy to have an emotional response to text that is programmed to look good for you and you alone.
> you communicate 1000 bits of information to me, instead of planned 500, and you yourself learn extra 500 bits.
Extremely optimistic. If this were the ideal scenario, you would USE the LLM to garner information ABOUT those 500 bits and then reframe them in a way that you yourself would put it. If there is insight, your "word" in your mental register now expands from the original 1000 bits to 1500 or 2000, and then are "processed" by your human brain that includes subconscious choices that are meaningful to the end result. There are tons of hidden semiotic data in your diction and wording (think resource forks in classic MacOS/HFS, only visible to the filesys) that is lost when you rely on another source to put together words for you; it's as if it is a game of Telephone. These are subtleties which you may intend for your recipient to receive and which are crucially important to your recipient and are irretrievable, it is intrinsically lossy. You have an alphabet soup of words, they cannot be put together by an LLM in exactly the way your brain did. We must rely on the fact that we ourselves put this together, the "aha" moment when an LLM does it for you is illusory and does not itself provide meaningfully important confirmation that you indeed say what you mean to say. Of course, humans say things and put things in way we do not intend to all the time. I still fundamentally believe this is more honest than relying on a third party that is not capable of understanding human emotional nuance to put together language for you, when language is and always has been a manner in which to dictate human emotional nuance.
> (Excuse the strong language, but I spent a lot of effort every day on both dealing with inconsiderate people lobbing LLM output at me, and making sure never to act like one myself, so it's a topic close to my heart.)
Does not negate the fact that LLM output itself, at least when used to convey human emotions or thoughts, is lossy. A very highly compressed JPEG with added interpolation from an upscale algorithm might come up with cool details that were never present in the original, and may look cool to both you and the recipient but are not honest to the source material. You receiving 240p JPEGs on a day-to-day basis is irrelevant to this. For the purposes of communication, it is a massive error which has the potential to compound, regardless of whether or not you or the recipient believe this to be the case.
> If this were the ideal scenario, you would USE the LLM to garner information ABOUT those 500 bits and then reframe them in a way that you yourself would put it.
Yes, but at that point in practice we're getting into over-optimizing territory. In this optimistic case I presented, you could learn those 500 bits yourself and formulate a clean message yourself, with all 1000 bits in it, but since LLM already gave you the text, and you feel you vouch for, you may as well send it over and save yourself the effort.
In reality the numbers are probably lower, and writing the message yourself is IMO also a good way to be truly sure you vouch for the "extra" 500 bits, as it forces you to actually pay attention. There's a chance you'll find inconsistency in output, or in your own understanding. I don't begrudge people for eventually cutting the process off here, for practical reasons - it's the fuzzy line between accuracy and perfectionism.
> A very highly compressed JPEG with added interpolation from an upscale algorithm might come up with cool details that were never present in the original, and may look cool to both you and the recipient but are not honest to the source material.
Again, I think it's a wrong take. LLMs aren't pulling the information out of their asses, and you are also not able to express every information directly. LLM can "upscale" information and you can take a look and recognize, "yes, this is exactly as it was", even without being able to write out that "upscaled" version by yourself. Verification is often easier than direct recall.
> I spent a lot of effort every day on both dealing with inconsiderate people lobbing LLM output at me, and making sure never to act like one myself
So why are you so eager to defend your fantastical scenario? It doesn’t matter how considerate you are, truth is the overwhelming majority of people aren’t and won’t be. We’re discussing reality here, not “what could be if we lived in a utopia which will never come to pass”.
The optimistic case is "having 500 bits, giving LLM 300, getting back 1000, and learning extra 500 in the process". Real numbers are lower. People don't vouch thoroughly and don't catch all mistakes.
But reasonable people don't send every output from LLMs to others without giving it a cursory glance (obvious hallucinations or nonsense would paint the sender as incompetent or inconsiderate), and that alone eliminates the worst levels of noise. A cursory read and cutting out obvious bullshit before sending is enough to make the message carry more bits of information than the propmpt.
> the overwhelming majority of people aren’t and won’t be.
In my experience, the "overwhelming majority" are giving something between a cursory glance and cursory edit; whether the resulting message has more or less information than prompt then depends on how much noise LLM added on top. The inconsiderate people I deal with, they often send "net more bits than in prompt" outputs, but those outputs are also verbose and not fully filtered for bullshit, thus it's effortful to tease out the signal from noise.
While I generally agree, the difference is I can also hand an LLM my pile of code and documents which is... a lot more information dense than basically anything I could write. Sure, a one-off prompt in a chat window isn't very helpful.
Summarization—especially of private context—is definitely one of the main exceptions to LLM writing being useless. However, it's still preferable to turn your private context into shared context and then give pointers to that, rather than having an LLM attempt to summarize it.
>If it's able to guess those 700 bits correctly, then they aren't true semantic information, and you really only have 300 bits you want to transfer.
A math teacher only has "this class is about math" to transfer, the rest is known. ;-)
Jokes aside, I think we're in a weird transition now where AI is used to generate text that looks good but is bad. In a few years people will know that and be more critical.
I think we went through a similar phase when DTP had it's breakthrough. Suddenly school papers were laser printed 300ppi times new roman and got more attention than better papers written by hand. But eventually that became the baseline.
I think Ai will make it so that well-written texts with clarity, good layout, correct illustrations, callouts etc become the norm, and will no longer impress anyone unless the information itself is actually good.
And if the information is good, it won't matter if it's AI generated or not.
Maybe things will change in the future, but I know where things stand right now. Would you rather learn about math by interacting with the teacher, by using the internet (including public LLMs), or by having the teacher tell an LLM "teach a math class" and then copy-paste its response to you? Option 1 is the best, but option 2 is at least better than option 3.
Oh it very much depends, your opinion there is certainly not always true and not shared by everyone; you only know where you stand right now with any certainty.
It depends on the teacher. I’ve had good and bad math classes, and LLMs today are quite a bit better than the bad ones. The worst human teacher in my memory didn’t offer interactions. He walked in, turned his back to the class, wrote equations on the board for 45 minutes, and then left. The best math teachers, the ones better than LLMs, are the ones who share the joy and sense of discovery and history of math, and not just the mechanics. But there aren’t that many teachers of that sort.
Learning by using the internet without LLMs is rarely very good, but more often than not in my experience sucks much worse than using LLMs. If you include using public LLMs in internet usage, then it’s not very different from just using LLMs that search the internet. I have heard that a huge swath of today’s high school and college kids are reaching for chatGPT before Google (which is incidentally OpenAI’s goal), and that many of them would rather talk to chatGPT than talk to a teacher. I’m going to refrain from making any claims, but I believe there are a lot of people who disagree with your ranking of the options.
Foreign language learning is one case where I love using LLMs, because it’s not typically an option otherwise. You can practice non-stop and have conversations with someone fluent in a language who will be infinitely patient with your mistakes. This is true of math and other subjects too; using LLMs to practice, so that the human teacher isn’t the bottleneck, to supplement and reinforce the human interactions, is usually better than using the internet without LLMs. The other reason many people prefer talking to LLMs is the lack of judgement. If you aren’t getting it and ask the teacher one too many basic questions, they treat you differently. Sometimes it’s necessary and helpful, and sometimes it’s harmful and takes a long time to change. LLMs don’t do that, they just explain and explain. That lack of judgement is a big reason many people prefer LLM interaction to human interaction.
Okay, sure. The difference between option 1 and option 2 is completely irrelevant to this discussion and I gave what I thought would be an uncontroversial take on it because I thought it didn't matter. The point is at least one of option 1 or 2 is almost always better than option 3. Option 3 is what I and others are arguing against.
Irrelevant? How so? And why did you present two different options if you thought it was irrelevant? I thought you were trying to make a point about human interaction. From my POV, your options 2 and 3 are much closer together than 1 & 2, so if you’re not suggesting that human interaction is preferable to LLM interaction, then what’s your justification for claiming any option is better than any other? I still completely disagree with your clarified point, and suspect many many other people do too, and are demonstrating that preference by using chatGPT for their classes instead of using either the teacher or Google. It’s a simple fact that interacting with an LLM can be preferable to interacting with some teachers some of the time, and LLMs currently often make interacting with the internet more pleasant and more efficient. Your ‘almost always’ still sounds to me like one person’s opinion that is not shared by all.
>Maybe things will change in the future, but I know where things stand right now
It's not an either-or though, LLM generated text already provides lots of value in many situations right now. It's also used for fluff, yes, but much of what humans write is fluff too, reporters often get paid by the word.
The point is that the value in a text has nothing to do with whether it was generated by an LLM or not. What matters is if it's useful or not.
> you can't give 300 bits of semantic information to an LLM and have it fill in the remaining 700, because it doesn't know what those 700 bits are. If it's able to guess those 700 bits correctly, then they aren't true semantic information, and you really only have 300 bits you want to transfer.
It doesn't actually follow, because maybe the LLM is smarter than the original writer (at least in the domain the writing is about) and hence really is able to complete the ideas in a way the writer can't. As an existing example, consider formulating a conjecture and having an LLM prove it. But I agree; if I wanted to read an LLM's output I'd simply ask it myself rather than read someone's supposedly-human writing.
You are stranded in a desert island. You start writing a message "Help, I am..." and pass at that point.
Somebody finds the message. They can no doubt come up with plausible continuations like "Help, I am Robinson Crusoe" or "Help, I am hungry" but they cannot create information. No matter how smart and how long you stare at the message, that is not going to tell you what the original person would have written.
Isn't it from Claude Shannon that information lowers uncertainty? Infinite regurgitation or massaging of data does not create new information. You will get the information form the LLM, not from that original person.
Once you've determined what the information-content of a message is, then you can apply information theory to it. But different receivers can derive a different amount of information from the same message.
Consider, for example, that if somebody doesn't know English at all, then before receiving the message, their best guess at what it is is some probability distribution over all English characters (or sounds, depending on what we assume them to know), and after knowing the first part is "Help, I am" that distribution might not change much at all. Therefore, they derived very little information from this message.
Going in the opposite direction: keeping fixed the knowledge someone starts with, there is an upper limit to how sure they could be (even if they are logically omniscient) in completing the message (that is, a lower limit on the entropy of their probability distribution) - this is what you're talking about in your example. But this limit only becomes important under these constraints - for example, knowing more about the person who wrote the message can let you predict it better, and if predictor A isn't logically omniscient, predictor B can do better than it with the same prior knowledge, just by being smarter than A.
>It doesn't actually follow, because maybe the LLM is smarter than the original writer (at least in the domain the writing is about) and hence really is able to complete the ideas in a way the writer can't.
I think you missed the import bit: "transfer of information from your brain to my brain." Including some third-party necessarily adds more that was not in your brain. So by using an LLM, you are transferring what is in your brain + some sludge that may or may not be useful for the other person's brain.
Depends. If the 700 bits were arrived at by the LLM while spending a lot of tokens, and the result is "good", I may want it through you as a middleman because it used up your tokens and won't eat my subscription usage limit to ask the AI to supply those 700. If you spend the tokens and put the result online, plenty of people can spare their tokens because they don't have to ask the AI to derive it. Bonus if that result was run through some kind of testing and verification.
Obviously this doesn't really apply to super simple questions that the LLM can just spit out the answer to right away.
True. There is at least one more case: when LLM the other person is using has access to their context and information repositories that they don't want to share directly, and so the LLM text gives a peek at a slice of that, and that's not something I can recreate from thin air. In that case the LLM text may have utility for me.
Bits of information depend on the readers prior knowledge, those bits are not absolute numbers. That makes communication not just information exchange but also syncing of priors. So some amount of redundant information maybe needed/wanted.
The value of writing isn’t always to communicate new information. It’s often to align everyone’s assumptions. For instance, when I say casually to a colleague or an agent “this change will require a db migration” they understand it’s to my teams primary application database. If I submit a design doc to a company wide review which database is changing is critical information. If the 300 bits were truly enough your agent or junior engineer would implement the wrong thing correctly as they often do.
Clarifying which database you're referring to is exactly the sort of thing I'm talking about when I say "new semantic information". It should be communicated, and you can't trust an LLM to choose the correct database, you need to specify that yourself.
What I'm talking about is if you add "database foo" to your prompt, the LLM may then add text describing what that database is, where it is, etc. But that's not new information, it (hopefully) already exists in your team's public docs, slack convos, etc. You should just say "database foo" directly to your reader, and if they want to learn more about that database, they can do that themselves, or you can give pointers to them based on what you consider important.
> say "new semantic information". It should be communicated, and you can't trust an LLM to choose the correct database, you need to specify that yourself.
But I can most of the time, and correct it if it chooses the wrong thing. That's the whole reason LLMs are faster. Its the reason we can give a paragraph prompt and get a kLOC PR back but only need to correct about 5% of it.
I have 300 bits of information in rude direct form with some obscene language as well. I ask LLM to wrap it into a nice polite message. Resulted 1000 bits are sent in the email.
I bet no one would like to get direct rude "source" instead.
> I bet no one would like to get direct rude "source" instead.
You bet wrongly. Rudeness carries information.
“Fucking hell, how many times have I asked you to XYZ” is different from “G’day gov’nor, terribly sorry to bother you. May I remind you to XYZ? Would you mind doing so at your earliest convenience? My deepest regards, toodeloo”.
The former conveys urgency and annoyance while the former conveys that you can keep ignoring it (and straining the relationship).
The latter (with its twisted mix of Australian, Cockney and Kings English) carries a calm sarcastic tone which indicates ones displeasure far more than the former, more vulgar statement, could ever hope to achieve.
One is however, reminded that Americans simply don't get our sarcasm, frequently leading to some amusing cultural clashes.
If you equate sarcasm with ease of offence, you are very much mistaken.
Regardless, in Blighty, such sarcasm is by no means twisted and indirect. It is as plain as the nose on ones face. As I said earlier, many Americans simply don't get such sarcasm, leading to comic clashes of culture.
I agree. But there's still a valid use here for LLMs: If I have 1 Million bits of information and I want to communicate a synthesis of 500 bits to you, LLMs can be a viable helper. It is rarely done that way - I agree, but the way I use LLMs: Drop in all the relevant context information (PDFs, HTML, Markdown, Pictures etc. - up to usually 200-300k tokens), then compile a synthesis/summary prompt (usually 1-2 A4 pages of handwritten text. Then copy the output (1-2 A4 Pages), manually edit and send off (often, including an archive of the full original conversation, so the human on the other side can consult the unfiltered original prompt + inferencing, if wanted).
In other words: If I didn't reduce the 500 pages of text for you, you wouldn't know what I mean or what is relevant, or how to filter it yourself.
If only it were 300:700... it is more like 300:100000 and up. But then you use AI to summarize to hopefully retrieve the 300 bits. We should try doing this in a loop with a simple enough sentence to start with and see what pops out after every iteration. Fully automatic Chinese whispers.
"Chinese whispers" - a Chinese colleague teased me about using that phrase in a meeting, which made me question whether I needed to use that phrase in other contexts (e.g. conversation with family, in my home town, or with close friends).
I am not ready to use the Americanism, 'teh telephone game', since nobody knows what that means, yet everyone over a certain age knows what 'Chinese whispers' means.
Using an LLM to write something persuasive is a good way to persuade me that you don't care about the topic personally, and therefore I probably shouldn't care what you or your LLM think.
> You might as well transfer those bits to me directly, rather than having the LLM add on an extra superfluous 700 bits that I then have to filter out.
I’m not disagreeing with your preference for human authored writing, but I don’t know that I agree with the content of your argument.
If you consider that what humans are doing during conversation is a form of compressed encoding / decoding from some latent representation through a quantized signal then if you interpret it that as a compressed sensing problem you absolutely can infer to a very close approximation the original latent representation using far fewer than those 1k bits.
From an information-theoretic perspective, if I can effectively transfer a latent concept to you in 300 bits of quantized communication, then by definition the latent concept itself isn't more than 300 bits. Any bit over 300 in the communication I send to you is fluff.
I love this example, partially because it jives with my conviction that LLMs are the ultimate translation machine. Ever since the embedding model days, it is clear that these models are amazing at representing meaning as math. The fact that LLM's most salient use is for coding somewhat agrees with that. After all, what is a programming language but another language? We instruct people with words and machines with code.
Saying the hard part out loud: we give more value to the people with 1000 bit brains than we do to the 300 bit people. There’s a strong incentive to write like you’re the former and not the latter. Freely available LLMs give people the means to match the incentive.
I notice this in the attitudes of students towards reading. They will take a 10-page journal article and ask the AI for a summary and then read a summary that's the equivalent of maybe half a page. But if the article could have been half a page, why is it actually 10 pages? It's true that there is some boilerplate, but it's strange to me that people could think that 90% of what they're reading is (to use your phrase) "not true semantic information". It's like if you went to a restaurant and ordered a 10-oz steak and they brought you a little teeny bite of steak and said "Oh, other places will give you a bigger one, but most of that is just filler, we just took out all the superfluous parts." It's a worrying sign for our future if things like this are not tripping people's skeptic sensors and making them wonder if they might possibly be missing something.
I think there are a lot of situations unaccounted for here
- people writing in a non native language
- people insecure in their writing
- people not used to writing in industry terms
- people with the curse of knowledge that are aware that they can’t write for a general audience well
Surely others too. None of those mean you have to read it, but I have gotten immense value from reading some things people have had ai write (and I’ve seen a ton of junk as well)
> people writing in a non native language
This category of people are going to be most betrayed by LLM output: 1) the receiver loses the signal that the sender might need to be queried to find their real intent and 2) the sender doesn't have the ability to determine if what they are sending is what they mean
> people insecure in their writing
These people can grow up, I don't care. Not a good enough reason to send a slop grenade.
>people not used to writing in industry terms
Similar to the non-native speakers, but slightly less in magnitude. They can educate themselves though.
- people with the curse of knowledge that are aware that they can’t write for a general audience well
These people probably can get some value out of it but they should take care
If you have 1000 bits of semantic information that you want to transfer but your default communication combines it with 10,000 bits of noise. Giving it all to an LLM and iterate on reducing that noise while making sure the 1000 bits is still present would enable you to communicate more effectively.
Overall, ideas are ideas. I'm not overly concerned with the fact that it was you who had the idea, as long as the idea is interesting. I don't know most of the people who write the things I read, so it seems to be of no consequence to me at all if they wrote it, as long as it is interesting. LLMs are notorious at creating things that are bland and vacuous, but they by no means have a monopoly on it.
Be the source human, machine, or dolphin, if they write a good article, I'm prepared to read it.
LLMs are wonderful at adding noise and okay at removing noise. My point is that they're not very useful at adding signal. If you wrote 10000 bits of noise and 1000 bits of signal, I would rather receive those 11000 bits from you, and if necessary ask an LLM to remove noise based on what I consider noise. If you can point out to the LLM what it should consider signal vs noise regardless of context, it should have been easy to not write that noise in the first place.
I don't think that necessarily holds, but it needs nuance. I tried to convey this within my company by giving a "guide" as to how to use LLM's for writing, across three modes:
1. Transliteration - roughly keeping the number of characters or bits, but translating to a different lingo, language or mental model (e.g. metaphors). Roughly the safest mode, but can still yield catastrophic results - it's safest if the author still provides taste and editing.
2. Compression - taking out redundancy to make the text more dense and more salient. The LLM chooses what to take out - and might take out the wrong things. More dangerous - but if you're happy with the salience and you believe the reader won't have time to read the uncompressed - it's probably safer than having the reader LLM compress without the benefit of your editing process.
3. Decompression - using the salience of your idea to add detail to the reader who wants to understand it fully, by utilising knowledge that is common to you and not common to the reader. This can be very powerful when there's no time to fully write the thing by a human - but it's the easiest to get wrong and to create slop. As an example - you could try explaining concept X + illustrate it through 3 examples. You know the examples are in public memory and easily retrievable - so you write your explanation of concept X, list the examples you want - and the LLM can take all of them, synthesise and bring the full package from your 300 bits to 1000 bits.
You are right that those are not the exact 1000 bits from the original brain, but they could contain 900 of the 1000 - which is still better communication efficiency than transferring 300.
I am however, more and more in the camp of fleshy brains writing everything, as my slop allergy rises.
Professional speeches are usually equivalent to LLM slop in terms of information content. They're written to sound nice first and foremost, and communicating information is a secondary goal.
Yes, but specifically compress the concepts, not necessarily the text encoding. "100 commas" and ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,," may be different amounts of ascii data, but they both represent the same concept that's worth around 20-30 bits of semantic information.
The real benefit is that the use of an LLM allows me to convey to it 1747 bits of scrambled information in an order that fits what's actually inside my head, and then have it unscramble that and convert it down to the 1000 you need. It can do that better and faster than I can.
It massively reduces the time it takes to make a short letter.
That's assumes a couple of things that are trivially not true:
- The assumption that both parties know about the same as an LLM does. An LLM know orders of magnitude more.
- The assumption that the output of the LLM is not refined over a few cycles.
The point is that you might give 300 bits of semantic information to an LLM, it fills it to a 1000 with perhaps 400 wrong bits. You correct it half a dozen times. It's now 950. You do the final touch ups. It's now at 1000. And it still took you 20% of the time to do it.
- An LLM has more knowledge, but it doesn't have information about what you specifically want to convey. Anything you give to the LLM may be good information. Anything the LLM adds is not additional real information, because whatever it adds can be inferred based on whatever you wrote. Or the LLM adds additional information that can't be inferred based on what you wrote, which is even worse because that's basically just misinformation.
- If you're giving additional prompts to the LLM to refine its output, then you're the one adding real information, not the LLM. The LLM is just rephrasing the information and adding noise.
- That's outside the scope of what I claim. The claim was that if you give an LLM 300-bits of information it cannot add more things that are relevant that the reader could not also do themselves. It certainly can. The LLM may add information you do not want to convey. Which is easily mediated by removing it or additional prompting.
- You are adding real information. The LLM is also adding real information. That's the entire point. It happens very often that an LLM suggest something to me that I did not know or simply did not think about. An LLM solved Navier-Stokes recently. That was most certainly not just adding noise. That's real information purely generated by an LLM. Information that was worth a million dollar price. Information that man centuries of mathematicians were not able to do generate.
Keep in mind that I'm using "information" in the information-theoretic sense. One could argue that because that particular solution to NS is provable, it is therefore implied by the propositions they started with and adds no new information.
And I'd still rather read a human's interpretation of the solution to NS than read whatever the LLM wrote.
That means folks using LLMs starting off with 300 bits KNOW that they lack the full payload of information to transfer to you. IOW, they know they need to transfer much more than 300, so they use LLM to fill those gaps. That's the crux of the slop universe out there. Folks are using LLMs for the 700 bits on top of their 300 bits and passing off the full 1000 bits as their own.
I echo the writer's sentiment. "I don't want to read the clanker's 700 bits. I want only your synthesis." (I can get the clanker to generate those 700 myself. Unless ... unless this whole LLM slop market is all about saving you the time to get an LLM to generate those 700!)
isn't it more like: i have 1000 bits, i transfer 1000 bits but depending on the person, it might be lossy, so they only understand about 700. they then come up with the 300+- on their own, potentially putting them over 1000 or they come back and ask questions to fill in the blank. the bits don't ever have to be bit identical.
Sure, but that seems orthogonal to what I'm saying, no? We could continue clarifying the definitions of information in this context, but my fundamental argument is LLMs don't add information to my writing that couldn't be added by the reader themself.
This of course has the potential to change with personal LLMs that can have shared private context with me. However, that isn't a defense for sending people AI slop, it just turns it from "LLMs don't add value" to "LLMs may add value when used judiciously."
They're lossy. I might make an esoteric reference that no one gets, or an analogy that doesn't quite land, and then have the LLM help me come up with something a bit more understandable to a general audience. There's a difference between dumping the output that you spent 5 minutes with, and taking an hour to craft that perfect analogy.
you think like such a techie. consciousness is not analogous to IO. if writing were only the transference of something from point A to B then what is the technical function of poetry, a question (the open ended kind), a pondering, a wondering, and so on? Furthermore, language is lossy and introduces a large degree of subjectivity, mystery, and uncertainty nomatter which words you choose. Words are by nature lower res/on a lower ontological domain than thought. So your premise is preposterous on both a practical and theoretical level.
Furthermore not all writing is for another to consume; nor even for the author themselves to consume. That is to say it has meaning ipso facto, not dependent on transference, as ritual.
Art is an abstract method of communication. When you choose the words of a poem, you're (hopefully) doing it to convey a feeling within you to the reader. If you write a poem about a beautiful spring day, it's probably because you experienced one, or you're remembering one, or someone was telling you about one, and that evokes feelings within you that you want to put into words, right? Surely you wouldn't write a poem about something you don't care about in any way?
When you talk about something you're wondering about, you're saying that you're missing information. Your ponderings are dancing around the void in your knowledge, defining its boundaries, and maybe imagining what answers might be able to fill that void.
When you put your thoughts into words, they're insufficient. You have so many ideas swirling around in your head, and you can never put them all on a page in the fidelity at which they exist internally. But words are the best we have. Whatever words you write are your best attempt to convey your thoughts to me (barring other media). You're distilling your inner voice that speaks a language only you can understand, into an outer voice that others can understand.
I don't think I'm exactly refuting you here. I think what you've written makes sense, and caused me to think about many things, more so than any other reply to me today. But I also don't think your comment is refuting the point I was trying to make, mainly that LLMs rarely add value in human-to-human communication.
I could probably have pasted my comment and yours into an LLM, and it would have come up with a clearer thread connecting my words to yours. But that thread probably wouldn't have been any of the ones either of us saw, would it?
Thanks for adding a new perspective to the conversation :)
To zero in on your last question, I do think a very good LLM can surface the unconscious (even completely unstated or hitherto nonexistent) "connecting thread" linking two subjective thoughts and make them explicitly known. And so what if they were not something either of us had in mind?
Surely the ability to do that is worth taking note of.
I am not advocating for letting LLM's write for you, to be clear. Sentiment wise, I largely agree with you. Just not with your total writing off of the possibility that it could serve.
It's easy to imagine an LLM aiding the communication between a mentally disabled person and their parent/caretaker.
Or, perhaps, some day, between animal and man. Who cares if the mediating component "hallucinates" some particulars of expression if it achieves the goals both want, which were previously impossible?
I will give you a use case where this is absolutely not the case.
I have a bunch of CLI utils I run for various clients and their peculiar setups. They now have man pages with descriptions and examples in them because the LLM went and read my code and did the needful.
I no longer have to re read my own code, rather I can just use the manual page.
Format and description came from semantics and context that (barely) existed elsewhere and I was not going to retain or transmit, but I have now.
But then the LLMs aren't adding new information, they're just reading your code and translating that information from e.g. python to english. Rather than sending someone an LLM-generated doc to someone, send the code and your own personal thoughts on the code. If your reader doesn't want to read and understand your code, they can ask an LLM to analyze it, within the context of their specific use case and your personal thoughts if any.
You're making a big assumption that the code is what is being executed, and not a compiled binary.
Where is the code: My repo? the clients? If it's in mine, the client does not have access and the CLI is a first stop to debugging. They arent in the context of written docs, more likely a production error from a log (thats now spitting out a message to check the CLI).
Less steps, less tools, more context in line and available in an interface your already using.
> they can ask an LLM to analyze it, within the context of their specific use case and your personal thoughts if any.
Or I can skim the man page it generated and make sure it looks good. The "work" (the tokens) dont have get spent over and over again.
I never liked information theory because information theory as Shannon envisioned it fundamentally did not deal with semantics.
AIT tried solving it? But AFAIK it's a lot of pretty results with not much real application.
A better approximation is something of a "shared model"; then you can actually state things like, the transfer of information sometimes is "trivial" because, well, it's right there in your compressor/decompressor.
My understanding is that to unambiguously quantify information, you need to have a known model within which that information fits. In this context, talking about LLMs adding value (or not) via inserting new semantic information, we're assuming that public knowledge on the internet is not new semantic information, and the quantity of information of contained in text talking about public knowledge is equal to the ~32 bits needed to point to that knowledge.
Saying something for years doesn't make it right. Has nobody ever challenged you on that? Our brains prefer to read enjoyable bits, not raw information. If you only got those 300 bits, you aren't going to bother putting them into your LLM to generate 700 bits to make it a more enjoyable read, you're just going to struggle through the 300. Maybe when browsers come with built-in automatic text puff-uppers, then you'll have a case but almost nobody does that now.
Here's a clearer example - would you rather learn a concept from a research paper or a textbook or blog? You say the research paper but they're dense and hard to wade through where-as blogs and textbooks are more wordy but hold your hand, which is something that helps humans learn.
Plenty of people have challenged me on this! There are a lot of exceptions we can come up with (summarization and private context are the biggest examples), and everything's changing in various ways as AIs get more capable. Usually, conversations about people pasting LLM output focus on things like AI generated blog posts, people who have no taste in writing and want an AI to rewrite everything for them, people who are just plain lazy, etc.
> would you rather learn a concept from a research paper or a textbook or blog?
I pretty much always read blogs first, and then move to a research paper only if I want more details or care enough about the subject to verify with the original source. Typically this is because research papers have too much information to be approachable.
"Writing is fundamentally the transfer of information from your brain to my brain. If you have 1000 bits of semantic information you want to transfer, you can't give 300 bits of semantic information to an LLM and have it fill in the remaining 700,"
You absolutely can if that information is in the code, which it often is.
There should not be that much in the code that needs further elucidation.
Some stuff definitely - but not much.
Usually you need the code and architectural summary + that stuff.
The AI is not very good at it but it will get better.
I think the debate here is about a few different things.
That would be true if there is only one receiver and if i could exactly predict how the receiver would interpret these 300 bits. But i could use LLM to expand these 300 bits to 1000 bits of more explicit information that gives enough redundancy and context that minimizes risk of misunderstanding and increase easiness of understanding, then validate such expanded message (and possibly do iterative corrections if LLM did not understand those 300 bits correctly), and finally release expanded message to (potentially large and heterogenous) group of receivers.
If there's a significant risk of misinterpretation when writing only 300 bits, then the idea you're trying to convey is worth more than 300 bits. In the process of prompting the LLM to add information, you're giving information to the LLM that you could instead just give to your readers directly. If you're just using an LLM to expand your thoughts in the hope that you get a better piece of writing, maybe you should put more effort into your own writing.
(This is all under an information model that assumes the LLM and your readers have equal access to knowledge, which I probably should have made more explicit in my original comment.)
It's funny, i push back on pull requests because there is too much description now - a 20 line change has pages and pages of generated description, rationalisation for why it is safe, defense of each design decision, analysis of risks and side effects. People are indignant, you're rejecting my change because there is too much documentation? And my response is, I don't have time to read it and you put me in the position where I can't afford not to - because approving the PR implies I did and accepted it. The investment to read all that for the value of a code change that I'm one prompt away from doing myself if I cared is just not high enough. So it's rejected.
You might be surprised how many organisations/ teams now don’t even review pull requests. Claude does coding, Codex do code review, and if developer did 5 round of this before pull request, we could very well just let CI merge.
I'm very curious how this goes long term. I guess we will find out.
My instinct says that these systems will expand their complexity to fully fit the cognitive budget of the agents that coded them and then atrophy the same way human-built systems do at lower cognitive budget. Only this time, because of the larger up front budget, the complexity ceiling will be higher, and the potential depth of the problem may be much much larger. It may mostly manifest as increasing cost over time - the agents grind for longer and longer, iterating over and over to fix all the failing tests, and the breaking point will be where it never converges and you come back to millions of dollars in budget spent and still tests are failing and effective gridlock on system changes.
But this may be all my human-biased fantasy that justifies still taking a role in software development.
> ... these systems will expand their complexity to fully fit the cognitive budget of the agents that coded them and then atrophy the same way human-built systems do at lower cognitive budget. Only this time, because of the larger up front budget, the complexity ceiling will be higher ...
Well articulated. I tend to think that the outcome will vary hugely by project. We already observe this with some reporting certain models behave in a certain way, whilst others see the opposite.
To use the wooley term “quality”, the top 20% might stand a good chance of making huge strides. But the remaining 80% of projects (in particular the bottom 20%) will atrophy extremely quickly. Yet, these will be the project that many push LLM’s too as their domain/technology is complex and/or outdated. Digital transformations that can be done quickly will be tantalising but ultimately unsatisfactory long term (as you describe).
IMO, PRs were never the right process to use in tightly-collaborating teams such as most companies. PRs got popular because orgs started using GitHub, and GitHub had made that workflow to fit the needs of open source (and modeled it after what open source was already doing in the days when patches got sent around via mailing lists).
I'm old enough to remember using CVS and then subversion in companies. People would commit straight to main (which was then called "trunk"), because making feature branches and merging them was cumbersome. And, on regular intervals, the person responsible for some corner of the codebase would do a show-and-tell presenting it to peers, but without the sharply defined boundaries of what the code looked like before vs. after some recent set of changes. People might remember some things from the previous show and tell or from first hand experience with that code, but that kind of memory is necessarily fuzzy, and diffs weren't an artefact that was typical to look at. So, these reviews didn't block people, and any comments that came from reviews defined a direction that things should go from here on out. If a corner of the codebase was deemed to be in a bad shape, the blame around that was equally fuzzy.
Others are just implicitly doing this but pretending review still exists.
Everyone is fatigued by endless code review which you get no credit for and has become massively more of a burden.
All PRs are superficially fine now. There are no typos, there is unit test coverage, but there are deeper issues that require massive amounts of effort and time to spot.
Unfortunately, a lot of previous PR reviews already were just gatekeeping, or "presenteeism". People would leave comments about class names or method names, like they couldn't understand what an `apply` method, the only method, meant on a class with that was already named appropriately and did one thing only (to give an oop example). No, the method needed to be renamed `PriceChecks.applyPriceChecksWithTimeConstraints`.
Lots of review comments about various conditions that wouldn't feasibly happen (same shit with claude now).
But then I'd see these same reviewers approving PRs where the bigger design was just fundamentally broken. Oh, we're adding a blocking call on our hot path, but at least the method name makes it very clear that it is blocking.
In general I agree that the current AI reviews are creating too much noise and it is masking these bigger design issues.
and this is such a liability imo. I recently got fired for what I think was apparently merging a change to Kubeconfig that both the CTO and DevOps person approved. I made as damn sure as I could that the proposed change was good, but I needed their eyes on it and they clearly weren't, possibly be cause every other repo had a review bot and no special automatic deployment automations. The DevOps person literally commented and said "It's good to merge", who to trust
I also have ADHD-driven blah blah blah. I find achieving brevity laborious, but find LLM results to be unacceptable. A journalism class taught by a well-known professional food critic helped most, but consistently using the free “Hemingway App” editor was almost as beneficial. Blindly following every suggestion pares your writing down to a nub, but it’s great at highlighting sentences you need to rethink without blurring your voice and ideas with bland suggestions.
That's a solid recommendation, thanks. I also did journalism. So I can write succinctly, I just have to put effort in like you. Which is much more than I'm willing to put in after I just wrote a treatise on the human condition to send to the lead developer at 07:30 on Monday.
I just have my LLM read the PR, then give me a summary of what the PR is about, how important it is, and how good the PR itself actually is. Then I have a conversation with the LLM about specific points, especially things where I get that feeling that I don't have a 100% understanding.
Until I fully understand what's going on, the PR doesn't move and my interrogation of the LLM doesn't end. My interaction is littered with "Explain X" and "How does this square with Y?" and "What if Z happens?"
The interrogation is the point, without me having to wade through hundreds of lines of irrelevant code to get at the meat of the matter.
Personally I'm starting to see less and less "smart" people here, given how easily and quickly everyone jumps on random hype trains with very little investigation or thought put into it. And those who do investigate and put thought into it are then quickly down voted for being a hater. This forum has turned into almost like a FOMO echo chamber in a way.
I can read the room. Coding is going to go the way of some other disciplines where machines do most of the detail work and and we know stuff works by verification. There are other fields like this.
Do I like it or not? That doesn't really matter. I need a job, so I'm going to get good at the new way to ensure I continue to have a job. I consider this to be a smart thing to do for myself and my family.
I agree with the work changing more towards verification, but if the verification is one LLM reviewing another one, then that isn't really verification. At best you "assume", but you don't "verify".
Unfortunately I see a lot of (senior as well) engineers who think that just a vanilla LLM reviewing another LLM is sufficient, and my comment was directed towards those. If however you see the LLM era as needing more test support and systems than ever before in the form E2E tests and so forth, where "code review" as such becomes mostly irrelevant as you have such a strong test system in place that if that passes you can be sure it doesn't break anything for users, then yes, that's good.
What I mean is verification through testing, and perhaps other more formal means. This has happened to other engineering domains, yet us programmers have so much hubris that we think it can't happen to us. And everyone is throwing their toys out the cot because it looks like it is actually going to happen.
Code reviews may not even happen, or if they do, it'll be all automated, and the verification will be the key.
Ask anyone in the semiconductor industry when last they understood the design of those things.
At this point, what is a human dev even there for?
We have this at work : fully AI-generated code and description. People will give review comments generated by AI which the "author" replies with an AI-generated response, all with LLM wording full of jargons no one understands not even the person who sent it. When you ask them what they meant, yeah idk Claude said so
Lately on my hobby projects I've been lazily committing work with the prompt 'add the relevant files to git and commit with a message that explains why' - only because I remember 15 years ago reading an HN comment from someone complaining that too many commit messages answer 'what' but not 'why' haha
So far I haven't had a reason to go back through commits to isolate any issues but if I do hoping the 'why' messages may come in handy for my LLM lol
I’ve started just asking it to touch up text to make it sound more human and follow technical writing guidelines and it converges on convincing pretty quickly. It also makes it easier for me to verify that it’s correct because all of the stupid LLM noise and formatting goes away and it gets distilled down to the important bits.
I'm going to be a bit contrarian here. In general, I 100% agree, just send me the prompt; however sometimes I feel an LLM can help structure thoughts. It can be a good sparring partner. If you proofread your LLM content, re-prompt until it sounds right and no longer like AI content, and feel that it actually conveys the information incredibly well, then I'm happy.
One of the instructions I've hammered into Claude is "Write like a human. I don't want this to sound like AI content. Your standard style of writing would fail miserably if it was reviewed by an English teacher. I want you to write prose that is nice to read. For example, write full sentences instead of bullet points."
It works wonders. Suddenly, my plan documents are something I actually understand, and something I'd be happy to share as an RFC. They still need iterating every time before they're ready to share, but I no longer have "the honest truth is" and other BS in my output. It's so refreshing for my brain to be able to actually focus on the content.
I've noticed this with my partner, too. She used Claude to draft a strategy document and felt completely overwhelmed. A classic moment of "AI did all the thinking for me, but now I don't know what I'm presenting". Once I helped her prompt Claude into writing the strategy in proper English, she understood what the AI was proposing, rejected large parts of it, iterated several times and ended up with a draft document where she edited the finishing touches herself and felt that it was truly hers. The AI was still incredibly useful: it helped her with the blank slate problem, and tremendously sped up her workflow.
So yeah, if you proofread and iterate on your AI's output until you feel you'd be proud if you had written it yourself, I'm happy to read it, too.
I'll be even more contrarian: I don't care if someone didn't write it as long as it's good (that's the major problem). I don't mind people using quotes when writing either. It's just that typically people (used to) curate their words before presenting them as writing. Although I'd prefer it if they also credited/explained what they used to write it in the case of heavy AI generation.
This, really. If I am entertained and learning then I do not care where the content came from.
Of course Claudeish or GPTish ("unusually") will easily push me away. Just like any repetitive or obnoxious tendencies that might appear in human writing.
> One of the instructions I've hammered into Claude is "Write like a human. …" It works wonders.
If this actually works, it's absurd. The implication is that Anthropic could trivially make Claude sound less obnoxious, but chooses not to. I don't see a way this could be justified as a safety feature or anything, so.
I tried a lot to get rid of Claude vibes and I failed. I made blacklists, I used prompt evolution to test against good human writing samples, even created a dynamic policy for dialogue, a cli tool used to steer agents dynamically by well chosen questions. The poor writing style made even models like Fable look worse than Sonnet 3.7 to me.
> So yeah, if you proofread and iterate on your AI's output until you feel you'd be proud if you had written it yourself, I'm happy to read it, too.
I agree, using LLMs is disproportionately frowned upon, but what matters is if you invested your own attention in the process. I use LLMs a lot for sparing, usually ask it to assume some opposing persona or use web search, not relying on its defaults.
I appreciate the direction of this article, and commend the author on publishing work, but the first sentence in the first paragraph is exactly what they lament.
> A pattern I see is that people use AI to build something new, then they use AI to retrospectively summarize what they have already built into a design document. Reading a document like this isn’t just difficult—it is punishing.
FWIW Pangram [0] "We believe that this entire text is human-written."
I've only just discovered pangram, but I've seen it referred to a few times in HN recently wit nothing obviously pejorative about it. Take it with all required grains of salt, though
That's a classic calling card of LLM prose but that doesn't mean they actually used an AI. Though sure, it would be nice if we gave this style a name and then abolished it.
I call it the “it’s not delivery—it’s DiGiorno” construct. I generally tune out when I see this construct, which is so prevalent on LinkedIn these days, which is inundated with clichéd writing.
No, usually humans write more than just a short unsubstantiated punchline in their "it's not x, it's y's", unless they're trying to sell you something.
Well, not quite the classic calling card - LLMs generally use a contrast of opposites. (e.g. "it wasn't a bug -- it was a feature", although that predates LLMs :)
TFA's use is more common in "normal" language: "it's not just [minor], it's [major]". (But, as others have pointed out, it was probably deliberately parodic anyway.)
I was about to comment on the exact same thing. Stopped reading after this. AI writing patterns has started to show itself in my AI allergy. Don’t get me wrong I like AI but it’s a pain to read the same style over and over again.
> People who rarely produced original writing are suddenly producing
these posts are beginning to make me wince. ai is giving voice to a lot of folks we probably wouldn't even be hearing from otherwise - because not everyone does their thinking in articulate prose, and extracting their realizations as shareable language takes effort, or long, embarassing iterations with ai "workshopping" to get to something they can read back and go yes this is what I am trying to say.
now the sentiment is that whatever these people had to say they could have just come out and said it - and that would be more passionate and less disjointed - no, not necessarily, and the more it gets repeated the more it's sounding elite and pompous to me. for example this:
> I had someone write me a personal message about a sensitive topic that was clearly workshopped with AI in an effort to nuance the conversation and not offend me. But the message became impersonal, dispassionate, and disjointed. It had all of the parts, but it didn’t make sense as a whole. I wasn’t interested in reading it, or responding.
this reads as someone choosing to ignore a personal message about a sensitive topic - because of an assumption they made about their writing: that they workshopped it when they should not have - because whatever it was they were struggling to say they should have isntead trusted the author to be able to understand the precise meaning of their raw, inelegant thoughts no matter how embarrassed they might be to share them in that form. they're complaining about a message they probably weren't otherwise going to see.
But how is that a positive that people that cannot give a fuck about writing coherent thoughts down to share them can now hog your attention with something that has no clever meaning whatsoever? Is talking just for the sake of talking a good thing in your book?
> people that cannot give a fuck about writing coherent thoughts down
That's clearly not the type of people they are talking about. For some people expressing their thoughts in words is really, really hard. It doesn't mean they don't give a damn. Just like being in a wheelchair doesn't necessarily make you lazy.
At the risk of sounding extremely insensitive, I am not interested in hearing the thoughts of someone if for them expressing them into words is really, really hard. I would be interested in reading the 30% or so that comes from them, but not at the cost of having to skim through the rest of the slop.
Do you realize there is no comment on this entire website of any real importance, including yours? Someone saying their reasons for agreeing with the article isn't something worthy of shut-downs and insults.
Everyone is an elitist, when you've enplaned do you assume the pilot is qualified to fly it or do you hope everyone gets a turn?
Admittance to the canon is no less a responsibility, would you trust the yoke of human culture to the artless, craftless, and naive?
It should be obvious to you, as a great epicurian of thinking, that the ability to spit out prose is orthogonal to the ability to think thoughts worth prosing about.
I strongly disagree with this statement. The careful selection of words that results in elegant prose is one and the same with the thinking. Sure, there are many ideas worth prosing about, but those are just a premise until someone goes through the effort to intentionally shape words around the premise.
>Orthogonal means intersecting or lying at right angles (perpendicular) in geometry, or mathematically independent and non-interacting in advanced fields
Because while I dislike AI writing, it matters in context. If someone is trying to formulate a coherent thought of work tasks required of me then I don't really care if it reads like AI as long as the point they're trying to get to me is clear.
If there's an attempt at something human, or emotionally important I will quickly balk at anything spat out through an LLM. If you're trying to communicate at the human level with me I want your input, flaws warts and all.
A card with just...nothing personal on it has as much value as machine text. I understand that you took the effort to think of me, but that's the bare minimum. If trying to express yourself isn't worth the time to make it personal then I'd almost appreciate if you didn't try at all because it hurts more to believe I'm not worth the effort.
How is ai giving them voice? All that generated slop does that you stop to want to read anything. And it is not their voice anyway. It is not their words, their ideas are mixed with ai generated points.
As for second paragraph, the writing "had no overall meaning". There is nothing to respond to if there is no overall meaning.
What we are seeing here is a Luddite reaction from artisans. A skill has been automated and those that have the skill are not happy about it.
For many writing is painful and the machines ease the pain. And like all machines it takes a while to use them skilfully. In particular remembering that like code it is write once and read many. What the reader needs should be at the forefront of a writers mind - both for code and prose.
Nobody gets on a motorbike and wins the TT in the first weekend.
Writing can be painful because putting pen to paper forces you to think through every word and sentence. You immediately realize that the first thought that came to mind, wasn't a 100% correct idea, and it'll be more accurate and better formulated this way or that way. Writing is a sort of extended thinking, in which you're being held accountable for every word (by yourself). Some people argue that slowing down the writing process even more, for example by going from keyboard to pen and paper, gives you more time to process your thoughts, and leads to better and more articulate writing.
When I read AI-enhanced writing, I always feel like I have to peer through the blinds to see the writer's true intention. This feels very dishonest and partly offensive, because it can sometimes take several paragraphs before I realize that I'm reading AI output, rather than someone's processed thoughts.
Not the person you're addressing, but writing by hand remains a very valuable tool. It's eminently practical and far from obsolete. [1]
[1] "There's actually some very important things going on during the embodied experience of writing by hand," says Ramesh Balasubramaniam, a neuroscientist at the University of California, Merced. "It has important cognitive benefits."
You may think this is silly, but you actually think more when you write by hand than when you type on a computer, and especially more than when you just listen. When attending a lecture you should write notes on paper even if you just throw them away.
Every person who has ever significantly won has done it by getting others to run for them. Tell me the billionaire that earned their fortune through the work of their own hands.
> What we are seeing here is a Luddite reaction from artisans.
They are hardly artisans. They are simply the vocal few who repeatedly conflate communication with art simply because the written word can be used for both.
"Easing the pain" is the foundation of many of today's problems. It's gotten too easy to avoid discomfort and effort. Also, if we are talking about disabilities, maybe we should start respecting different forms of communication instead of enabling people to "write" in a way that is not authentic to them at all, just so they can seem "normal".
Actually, yes. Even though you tried to make it sound ridiculous. I want to move towards a world where the worth of a person (and their thoughts, feelings, etc.) is not measured by an exclusive standard of "intelligence", especially if that standard is created/enforced by mega-corporations.
I really, really despise this sentiment and the people that make it.
Writing has always been incredibly difficult for me, both due to incredible difficulties with handwriting and finding my own voice. That difficulty I find in writing is the friction that identifies what my voice is, the words I choose and the meaning I attempt to convey through my own understanding.
I cannot read AI-generated crap for the life of me because all of it simply has no meaning. I will go back and read through paragraphs, trying to grasp at the meaning of something that does not exist. If your argument is that AI-generated writing eases the pain then I will make the opposite argument: reading AI-generated slop increases the pain 10x. I would sooner read through an essay with endless punctuation issues and spelling mistakes than one passed through an LLM because one is an earnest attempt to communicate while the other is offloading all of the mental burden to me. Which gets to the root issue of a lot of the complaints around people slinging slop: you make it marginally easier for you, and then far worse for everyone around you.
Nobody cares what your English grade is. Everyone would prefer to just hear you speak in your own voice. This isn't Luddism, this is the equivalent of people pushing their selfie through five Instagram filters until they look like an Oblivion character because they have no confidence in their own appearance.
This is not done in the service of readers, nobody wants the slop, it's writing dysphoria
the article we're commenting on quite literally cites a survey on reader behavior:
"Cynthia Dunlop recently shared survey results on reading AI-written articles entitled Report: How developers react to AI-scented blog posts. For the majority of readers, if they think an article is AI-assisted or AI-authored, they will stop reading (78%) and avoid the author in the future (71%)[...] the strongest result in the survey was people overwhelmingly preferred the author’s own writing (98%), with all its flaws and idiosyncrasies,[...]*
there's a say-do gap in play - similar to makeup/no-makeup studies: people claim to prefer natural/no-makeup while behavioral research reveals a preference for subtle/light/professionally-applied makeup
when it comes to ai people love to claim to hate it outright - this thread is full of descriptions of ai writing being impossible to read, totally incoherent, and superlatively bad/worse than average human writing. meanwhile evidence suggests that we might actually kind of like ai writing - with a growing portion of samples making hn's frontpage (https://www.salahadawi.com/hacker-news-ai-detector/state-of-...)
I've seen my share of ai slop and I get where the rage is coming from - I think folks could do a lot better than shouting down every use of ai as slop, especially while hypocritically using ai in their own writing. the pompous elitism on display in this thread alone is gross even for hn
the fact that the HN (or other) frontpages feature a non-trivial amount of AI content is to a large extent owed to the fact that many commenters don't even read the content they're commenting on, another problem in and of itself.
There is no elitism in demanding that people use their own words, or if you want to call it elitism then elitism is healthy and good. Imagine ten years ago you talk to someone online and they told you they outsourced their posting to a call center guy in the third world, that's what this is. Literacy and agency aren't gross, it's what we ought to expect of an educated adult. It's fundamentally a sign of honesty that if I talk to you or read what you write I want to hear words out of your mouth, not some meat puppet version of you hooked up to a chatbot.
On the note of imperfections... My preference is now strongly towards reading someone's stream of thought thrown down at speed over than that same stream of thought shat through the digestive track of an AI agent. At least that feels human.
To me, it also seems like they AI digestion is getting actively worse? As best as I can tell, all the agentic nature and reasoning for code is now making writing actively worse, as the agent pulls across your whole knowledge base and will take that one thought and eagerly join and context it thinks is relevant, with the reasoning spread throughout the page.
For better or for worse I’ve always preferred this, and IME a large portion of people (especially younger or who don’t write/communicate much for recreation or their work) will not read it or assume it was AI.
IMO it is a costly/goodhart-resistant way to “show your work” and help other people understand or challenge your mental model. (IE a justification for something you believe to be true). Overly polished writing is performative, it’s hard to take seriously once you’ve read The Economist/LW enough to see how poorly “well written” correlates with truth
To a certain extent are all wrong or ignorant about almost everything because our knowledge/time are very limited. But it is really important to understand what other people think in order to coordinate with them/align human goals and understanding.
It’s good that the average person taking the lowest-friction path to using an LLM in bad faith is easy to identify now. The more obvious and disliked it becomes, the more they’ll be hit with the stick to actually know things and not bother people. It’s so much worse to be “bad and stupid and not care” than “possibly cringe or wrong”
I feel the same. I've actually started to appreciate things I used to dislike. Like typos, or grammatical errors. I used to see it as a lack of attention to detail. But more and more it now feels like "hey, something written by a fellow human!".
Same thing for video voice overs. Things like a bad quality microphone, or someone who doesn't pronounce things very clearly. Now I go: for sure a human!
There's a lot of misunderstanding here. The quality of LLM writing has not plateaued, it has dropped significantly. They've cut it to the point where people start noticing.
Do a search on "Claude Sonnet 4.5" on Reddit and you'll see lots of disappointed users [1]
I'm curious if there have been any formal studies on this. I suspect that people have become better over time at identifying LLM writing styles, and are also combining that with selection/confirmation bias. Dunno what fraction of the perceived drop in quality comes from that.
Have you seen the episode "Darmok" of Star Trek The New Generation series? It is about a people with an impenetrable language aimed at packing as much information as possible by using references and metaphors, not descriptions. I guess this is where LLM language output is heading.
I feel like this is more of a curation issue than an issue with using AI itself. If someone is using AI to distill information to the point where important context gets lost, then yeah, I think they are definitely doing it wrong.
What I often notice in pull request descriptions isn't necessarily a problem with them being written by AI. It's the lack of nuance. They don't tell me exactly what the author expects the reviewer to pay attention to, what might be risky, or where they would particularly value another set of eyes. Being able to articulate those things is far more important to me than whether AI was involved in writing the text.
I would also push back a little on the author's point about imperfections. Being able to say that imperfections are fine somewhat ignores the elephant in the room: bias. Bias can have very real consequences. The author may personally be more tolerant of variations or imperfections in writing, but that's not something we can realistically expect from everyone. I'm also not sure it's fair to put the burden of dealing with those biases entirely on the person doing the writing.
That's also why I keep coming back to the expectation of having a distinctive "voice." Not everyone has the same attachment to their writing voice in the first place. For some people, getting their ideas across clearly and being understood matters much more. If voice isn't something they particularly value or identify with, then insisting that they preserve it feels somewhat moot.
“Language can become a screen which stands between the thinker and reality. This is the reason why true creativity often starts where language ends.”
— Arthur Koestler
We do not experience the world as words. Words and language are already an hallucination on top of reality and LLMs are an hallucination on top of language.
There are many use cases for LLMs that will help humanity progress. Putting AI in between yourself and the people you want to build productive relationships with is not one of them.
My take here is, I think the real issue is AI pretending to be human by mocking a human style.
I'd always rather read an actual person's thoughts, clearly communicated and written down in their voice. But, there's clearly a subsection of people who don't want to (or aren't confident in / etc) writing their ideas fully.
If they have a bullet point list, and they want to use AI to neaten up the bullet point list and present it back, that's fine by me. I wouldn't have minded reading their thoughts in a messy list, but AI is serving to make it more readable.
What I think everyone hates, is AI being used to convert the list into pseudo-thought-out-prose. It just contains a bunch of bloat that wastes people's time.
So my advice if you use AI to write and don't want to change: keep going, just stop seeing the desired output as human prose. Just neaten up your bullet point list / promot / whatever, and send me that.
A good code or commit comment should succinctly explain the subtle details or motivation not obvious from the content itself. AI comments tend to wax poetic about surface details.
The same goes for articles. I want the author to distill the material using their own voice and hard earned expertise. I don't want a generic text which sort of tells the tale.
> I would much rather someone be themselves and write with their own voice, or with some passion
I recently had a fairly successful article here on HN. The common thread among comments was appreciation for the passion and it being "hand-crafted". I am thrilled my passion shone through despite the dry subject. Still, it's sad that something simply being written by a human has become such a rare compliment.
Nobody is going to read anything regardless of whether you used a LLM to write it or not (but if you use a LLM they'll use that as their excuse for the virtue signalling aura farm), so honestly why bother writing.
If it's written down in a form that can't be repudiated, a succinct message to any stakeholders covering any decision of substance is good practice. They might not read it, but they cannot pretend I did not make a good faith attempt to inform them.
(And I try to be judicious about what counts as "of substance" so they can't accuse me of flooding them with messages)
I think AI can write good already. It’s all up to the prompt. Strong guidelines. Without them, it will sound like an AI because, just as the article said, the language reflects the writer.
this is intended as a silly observation and not an actual accusation, but for an article about not writing with AI, the following is one AI-ass sentence:
“Reading a document like this isn’t just difficult—it is punishing.”
LLM didn't invent it themselves, they are plagiarism machines. Obviously these kind of patterns had to be popular in the first place to be overused by LLMs
Also we have reached the point where LLMs are pubkishing more contents than humans so we have probably reached a point where humans will inconciously start copying and using LLMs style too.
c'mon, it's literally the top 2 AI writing tropes in a single sentence. i googled "AI tropes", and here are the first 2 tropes listed in the first search result:
1) negative parallelism: "It's not bold. It's backwards."
2) em-dash addiction: "The problem -- and this is the part nobody talks about -- is systemic."
again, i'm not saying the author wrote this sentence using AI. i was just tickled by how stereotypically AI it sounded. for better or for worse, those characteristics are just going to set off the AI detector in people's heads now, as evidenced by the fact that there are like 10 separate comments on this post that picked up on the exact same thing.
I kind of agree with this, but I think it’s model specific and a temporary problem.
Opus 5 and Fable 5.1 commit/PR messages are incomprehensible garbage.
However, Astra messages are nearly perfect for a copy/paste to less-technical stakeholders. I maybe fix a line or two.
Give it a year, and I suspect that I won’t even need to make those fixes.
The only problem right now is I can’t get my team 20x OpenAI accounts due to supply constraints. We’re all stuck waiting, hoping Anthropic either ups their game or OpenAI gets more capacity.
We would easily pay $1000 a month per developer/PM for a business tier ~30x account or similar that let everyone use Astra all week without running out of tokens. And that is entirely because of the writing and communication improvements.
These last few days, there has been a small stream of blogposts on HN expressing similar things, and I have enjoyed them all.
My question recently has been how to broach this subject with colleagues who really enjoy producing prose with AI. There is not yet a better cultural shorthand for this sort of thing than "slop" which is a harsh-sounding word and itself sort of a thought-terminating cliché. "I don't want to read what you didn't write" is maybe closer — but it needs a pithier and somewhat more encouraging encapsulation, like "I want to hear it from you".
Has anyone had good experiences setting up professional boundaries or team norms around AI-written docs?
It's something that my company has been struggling with. Developers are generating mountains of code, documentation, and Jira ticket comments. It's incomprehensible and overwhelming.
For the team I lead, my guideline is AI generated is fine but it needs to be human-edited and/or summarized. You want me to read what you're offering? Put some effort into it and meet me halfway. I don't want AI generated gibberish with made-up terms. You'd better also understand what you are presenting as your work. It's been fairly well-received though we're still working on it.
I have some co-workers on other teams who use AI to generate responses to literally everything. Ask a simple question? Get pages of AI generated nonsense in response. They are proving a tougher nut to crack.
There is one person who is so good at writing with AI assistance.
Everyone else, including me, not so much. I envy his talent and ability to so successfully use the new tools. It's just something to think and talk about regularly. Maybe one day we will all be able to use the tools as well as that guy.
Ask questions about how the audience is taking the blog posts. Say that many people seem to really not like LLM writing and ask how are they listening to their readers on this point.
You may find that brings a better and more human focused discussion. Who is the audience and what is best for them? Don't frame it about your preferences but about the real persons that your colleagues seek to talk to.
In my current position(DevOps) we are getting this kind of messages or issues a lot recently. While forwarding these issues to my team lead, i don't have to deal with them which feels great but once its assigned to me... Its another long process to understand what issue actually means and what i need to do.
Thats said, i don't know any professional way for this but currenlty trying to communicate with peers in order to work without a problem.
im sensing about a 20 month lag between hn sentiment and normy-coworker sentiment. managers and business-types pointing out the benefits and challenges of working with an llm i told them about early 2025. im just politely nodding along waiting for them to catch up. theres some emperors new clothes happening at my company so maybe im complicit
At a broad level, the issue isn’t generated content, it is the ratio of verification capacity to generation capacity (V/G). Your pain is because generation capacity has increased significantly, while verification is laborious and capacity has not (and can not) catch up.
Unlike spam which is from external sources and can be ignored, messages from other employees have to be responded to. I guarantee this is creating bottlenecks all across the firm, outside of the individuals who are feeling productive.
For fixes, theres theoretical approaches that might work?
If you need leadership to help you, then this issue has to become something that is on their radar, which means that something needs to go wrong or costs need to be registered.
The shortest conversation for that is to make people aware that generation has improved individual productivity, while moving the costs of that production to the rest of the firm.
If leadership is not at the stage to listen, then you need to move the costs you are incurring to the people who are sending them to you. Maybe set time aside to sit down with whoever sent a PR and then read what they sent together, to understand it.
It also makes a difference if tokens are being subsidized or not. If the firm doesn’t care how many tokens are being used, then you are naturally going to have over production.
AI-generated writing does not bother me, unless it is intended as prose writing, where the flow communicates as much as the specific facts that are described. For technical writing, I usually just scan for important information anyway, and in that sense the only thing that bothers me about AI-generated writing is if it's too verbose and obscures the important information by leaning ineffectively on jargon and high-minded verbiage. I find that Claude Opus is very guilty of this, whereas GPT 5.6 variants are not.
>the only thing that bothers me about AI-generated writing is if it's too verbose and obscures the important information by leaning ineffectively on jargon and high-minded verbiage
But so many READMEs and design docs are degraded to this level by uncritical/unmonitored use of Claude. You make it sound like it’s an exception, not a rule.
In my experience, that brain damage inducing Claude-style technical writing is everywhere. It sucks not only because of its verbosity, but because it includes completely pointless tidbits extracted from random extended LLM sessions, with no attempt to prune redundant or gratuitous details.
It’s funny, this pattern is very similar to how I write code with AI these days.
I write almost every line but I have one or more agents always running against the changes alongside me evaluating correctness, suggesting approaches, auditing test coverage, etc.
I’ll also use it for grunt work refactors, scaffolding, resolving merge/rebase conflicts, etc.
If you don't get to the point and continue to go on and on, I will use a LLM to summarize what you wrote and then decide whether to spend time listening/reading what you wrote.
We're at a point where I assume that whatever I read is something the "author" endorses, but may not have written. And that may be more important than the specific words an author may have chosen themselves. I doubt we can ever go back.
What should worry us is that AI may introduce a slant that the prompter assimilates during their conversation with the tool, which could introduce a subtle control lever over public opinion.
I don't disagree with the main points of the article. But I feel like soon with all the writing that's been hating on AI writing recently on HN, the LLMs are going to be really good at writing articles about how bad AI is for writing...
>I had someone write me a personal message about a sensitive topic that was clearly workshopped with AI in an effort to nuance the conversation and not offend me. But the message became impersonal, dispassionate, and disjointed. It had all of the parts, but it didn’t make sense as a whole. I wasn’t interested in reading it, or responding. I would much rather someone be themselves and write with their own voice, or with some passion, even if they can’t quite find the words, or risk offending me.
Only a privileged person can say this. Not everybody is a native speaker or good communicator. As an IT guy somewhere on the spectrum, I have damaged my career on multiple occasions by clumsy writing.
The most sensible solution that I've seen so far is to delineate human spaces from LLM spaces. Technical mumbo-jumbo can be LLM as an artifact, and human thought can just point to them.
To me it is just a bittersweet irony: people who don’t know what a book is for 30-40 years or in earlier days got the RTFM (gif smiley not included) treatment dump one liners to produce a giant load of text corpus that needs to be processed by an AI to get the gist out of it with sometimes quite some refinement by asking many questions.
So there it is: cruel people who don’t even read their own prompts while happily believing others read their text load and pat the genius on the shoulder for being such a great guy in writing, while in essence it turned the RTFM movement on its head: UUOAI - “Use your own AI.”
People cannot communicate better by using the digital equivalent of logorrhea.
> I also see more and more pull request summaries clearly written by machines for machines. They are rich in detail: this was changed to that, these things were split, those things were merged, this was left untouched, tests were added for this, and so on. But I’m left asking: Why are we doing this? What is the value? How risky or urgent is this work? Where do you want my input? What should I pay attention to?
My own solution for this (for now at least) is to add a preamble section “AI;DR” to PR descriptions, filling what a TL;DR section would have done in the past. I’ve socialized to my peers that no part of such a section will be written by AI, and it should contain exactly this missing high-level content described by the author.
I've always wanted to write this essay, but now I no longer need to. My title would have been pretty similar: If You Want Me to Read It, Please Write It.
One minor addition: when a (conscious) human writes a proposal, a question or a request, they'd instinctively choose words with which they can express their uncertainty in some of the details. I as a reader can detect these uncertainties and I can choose to do more research assert or correct them, if it feels appropriate. AI-generated texts miss those cues, therefore they are much harder to read and connect to.
Writing for yourself is crucial. Writing is in a way thinking and understanding. If you can't write for yourself, then don't write at all if you expect someone else to read it.
I have a large and complicated project at $DAYJOB where I am the solo dev with a large team of domain experts. They AI dump me with suggestions (rather rarely, thank god) for features with complicated implementation details despite having no insight into the codebase. What is the point of that? I can't discuss it with them. If I jump on a call, they'd have to say "oh sorry — I don't know, I'll get back to you".
When I need to send them a message about something domain specific I have no knowledge of, I'll spend hours understanding it as far as I can. And then wording it in a way that makes sense to me. I use AI A LOT in the understanding and proof reading of messages, but the wording and writing is mine.
I'd rather understand something partially, write a bit poorly, and send an imperfect message than to AI dump a perfectly formed message which takes everything into account but not really understand everything that I'm sending. It closes the door to actual communication between two humans.
My org, like every org, is struggling with this as well. We find ourselves writing skills like `/decomment` and often running several "minimize" passes on PR descriptions, etc.
One thing that has markedly helped is tweaking our AGENTS.md prompt and prose-heavy skills like `/pr` or `/ticket` to specify that all technical output needs to be written in ASD-STE100 Simplified Technical English. It has helped us so much. We still have to `/decomment` or minimize to remove unnecessary stuff that's been written in ASD-STE100, but the output is a lot easier to digest.
OTOH, we have a guy in marketing that has taken it upon himself to use Claude to draft these wildly complicated and dense implementation specs that he doesn't understand for various things that he just hands off to my team to implement. He has no context for how our product is built, the various layers and their interactions, etc. so they're often flat out wrong and overly complex, invent systems that do not exist, all of that. Then he gets mad that we're not moving as fast as he wants, because he gave us the implementation spec! I have been struggling to come up with a way to diplomatically tell him that his actions are slowing us down, and if he'd just describe what he needs instead of speccing it himself, we'd be twice as fast...
I feel the same way, and yet I'm also guilty of this. I'm trying to reach out to an expert, and because of the stakes, I want to write it exactly right. I'm using Claude to write it, but I don't like how Claude writes, so I rewrite it in my own words, and I still don't like it. I've iterated over this for several days. Finally I talk to my wife, she gives me some pointers, and I just write it down perfectly.
And yet, every person high up in business that I talk to, including my wife, advises me to use Claude for this sort of thing. I don't get it. Claude text sucks, and it hurts my eyes when I'm forced to read it.
As someone who is usually on the other side of this interaction (software engineer at a company receiving bug reports), it's much easier when people write in their own words (even if they sometimes use the wrong terms or say something which can't be true) than when they use an LLM to make things clear and correct. I want to know what you're trying to do, what you expected to happen, and what actually happened - that doesn't benefit from having an LLM rewrite it.
Sure, I use precise and technical language, but I'm still a human and I can understand what people are getting at when they use imprecise terms. But if the text passes through an LLM I can't tell whether the result actually matches the original problem or whether the LLM has made guesses or filled in blanks.
Bug reports aren't so high-stakes (although poorly written bug reports can often be useless). I've got something to sell, and I want this expert to confirm that what I've built is better than what's currently available. That's a much harder approach. You want people to send bug reports. You don't want everybody to try to sell something to you. But I really think that I've solved the problem that she identified, so I want her to take a look at it instead of dismissing it as most people would probably do when a random nobody claims to have solved something. It doesn't help that this entire issue is way out of my comfort zone.
I ask coworkers point blank to send me a non-llm version of their slop sagas. The initial exchange can be a bit awkward, particularly with new hires, but consistently asking for human prose has worked for me.
On Windows, use wincompose. Space followed by a hyphen gives you an emdash with the cute half-width spaces around it. On Android, I can get an emdash by long pressing hyphen. There's also a keyboard shortcut on the mac for it. No idea about Linux/BSD or iOS, unfortunately.
On Windows I also use QuickAccent, now part of PowerToys. I could never remember the Alt-code for the em dash. Apparently it's Alt-0151. With QuickAccent I hold a letter and right or left arrow and up comes a menu of related symbols. It's pretty neat.
On Mac, it's option-shift-hyphen. Option-hyphen gives the en dash.
On Linux, you may be able to type Ctrl-Shift-U to get unicode entry and type 2014 and Enter. 2013 is the en dash. Again, I will never remember that.
This sentiment gets echoed over and over. While I can certainly see the value in human-created creative works, if the subject is pure facts, there are only so many ways to state the facts without being caught in the weeds and missing your entire point.
Moreover, people are finding it hard to differentiate what is an is not AI-generated with newer models, often attributing original work with those of LLMs. It has just become an easy scapegoat for lazy comprehension and a desire to do less. You are jumping at AI boogeymen.
Just about the only thing here I can level with you on is, yes, AI is far from perfect and will continue to advance. Otherwise, so much of this reads as fruity prose to excuse apathy.
A lot of my problem with AI writing is that the facts aren’t checked. You get a big cheesy dramatic arc but the core story isn’t factually correct. There’s only so much time in a lifetime, too little to waste fact checking someone else’s work because they couldn’t be bothered to.
I have junior devs submitting huge PRs with multi paragraph descriptions about load baring root causes and "why this works". Then I ask them directly what are the steps to test this and they don't know, they have never run the code.
Rather than entertain the idea of reviewing the code I just send it back to them until they work out how to run it. And almost always they submit a new change because the last one didn't actually work, despite how confident and articulate claude was to them.
A change can pass CI without actually making the change required. The tests test that the code does what it does, not that it does what the user wanted.
AI is really good at making tests that pass even if the change itself is wrong.
It's curious that you're pinning the blame of "excusing apathy" on the reader and not the writer. Why doesn't it go both ways? Given that producing LLM text is so cheap, your proposal amounts to an asymmetric war which is ultimately a DDOS on my finite time and attention. Or in other words: if it wasn't worth your time to write it, it's not worth my time to read it.
Let us say someone had few good ideas, and seeded them into a prompt, and after few back and forth, web searches via agent, feedback from the author, a piece of work was produced and the author decided to share it as a blog.
This is not much different than how people are producing original work LLM in areas such as math.
Would you object to reading their work because it was a byproduct of collaboration between AI/Humans? What about songs? movies? math proves? and software produces as such?
In my mind the issue isn't that it was crafted by AI, but that it wasn't crafted with the intention of getting the point across in a way that respects the reader's time or energy. The result is often needlessly wordy with information buried in a sea of flourish. This might be forgivable if the author had a certain entertaining style, but LLMs are aggressively mediocre by their nature, making wading through it painful.
Its certainly possible to get an LLM to write concise prose but thats not how most people are using them. We can talk all day on HN about the ideal LLM user that is conscientious, reviews the output, puts time and effort into stylistic concerns, etc. But none of that actually matters because thats not what people are doing; the entire reason they're using an LLM in the first place is so that they don't have to do that.
On that we agree then. I also don't want to go through mountains of tangled text written by an AI.
But what if the AI models improve their writing style considerably to match that of the best authors? which I think is plausible soon.
I feel in the near future, what really matters the most is the seed and the dialog/back and forth more than then who wrote what. And that the best content would be a by product of human/machine interaction atleast given the current generation of tech.
Until perhaps, in the the not so distant future, the math can encode our taste, judgments and develop imagination, then maybe the best human wouldn't be able to create anything better than the machine. But then again, we want to communicate the inner state of human being, because we want to build relationship, so content for consumption might be dominated by machines, but communication as way to build relationship might stay genuinely human...but even that might change, what if machines eventually become better at relationships than humans? that could be how AI truly "kill us"..I think within few decades we will discover that our intellect was a short period of cosmic evolution, and we will accept our fate as animals with analog organic intelligence that helped to bootstrap what comes next, this might be scary but humbling, and I think it's inevitable.
The lazy ones are the people spending 60s prompting then hitting copy/paste. They dump it on me as is and call it a job well done, which leads to me doing their job for them or calling them out to stop the behavior.
I don’t like doing either of these things. I don't want to do your job and I don't want to lecture about how it isn’t “your work” if you don’t touch the content after an LLM spits it out. It is rude and selfish to put me in that position.
Seconding the author, at my last job a new VP started pushing AI hard, in every use case, and suddenly I had to read AI-generated high-level design documents that were both overly-detailed, and... really difficult to read.
There was just something about them, like they were written by someone obsessed with detail but without a clear idea of the audience they were writing to.
The result always ended up being a document that was way too long, way too detailed, without context for those details, speaking over the head of the reader.
And I don't want to read the millionth iteration of some old man complaining about the kids these days, written by a human or not, yet here we are.
If you don't like AI slop. Don't read it. But wasting your time generating human slop to complain about AI slop is so obviously futile that it immediately identifies the writer as lacking the capacity for reason or emotional clarity, or merely seeking attention for their self-promotion with clickbait.
The problem is, some of us work with colleagues, or even bosses, who throw AI replies at us in nearly every conversation — so we don’t have a choice to NOT read it.
I am sorry, I didn't take into account people in literal slavery. I am very sorry to hear about your condition. Perhaps you should contact the UN or some human rights organization, they may be able to help.
Just throw it into your LLM and de-slopify it. I sometimes just ask "wtf is he on about, 2 paragraphs max". Tends to work. Sometimes you still have slop afterwards but at least it's 2 paragraphs of slop instead of a wall of it. (Not condoning the behaviour, just coping with it :P)
I’m fine with reading something you didn’t write if it has the information I need. I’m not fine with reading something you didn’t read and will just waste my time.
I think the future is people's AI agents evaluating the content generated by other people's AI agents and vice versa.
What the humans would see would be what is filtered up to them after all of this evaluation is done. And content generation would be the reverse. People would provide a seed of an idea and then their agents would turn that into potentially useful work.
I'm finding that executives at work are generating AI slop and sending it at each other and teams that are then meant to digest it, without any critical thought. Often these are sloppy and contain broad mistakes that aren't relevant to our company's setup.
For me the only thing that matters is the content the quality of the writing. If you're primarily interested in something else than content and quality, there is no point in reading at all. So I don't understand these anti-AI posts. If all they say is that most AI written texts are too long and badly written, I agree. However, that's also the case for many texts written by humans.
“Reading a document like this isn’t just difficult—it is punishing.”
I’d be curious whether the author composed this sentence himself or it was the output of AI. Personally I often find myself “it’s not X it’s Y” and then recoiling in disgust and rephrasing it simply because AI has made it so grating from overuse.
As an educator in university I have found the project proposals of students have become longer, to some degree more well written and quite certainly have LLMs involved.
This becomes problematic when it turns out people haven't even read what they sent me.
To clarify: this is about voluntary projects that do not win them extra points.
> I don’t want to live in a world where you use AI to summarize something important into unreadable text, and then I use AI in an attempt to decipher it. I want to hear you, imperfections and all.
It doesn't fundamentally change the equation if I use AI to prepare and then write it myself. If I'm using AI effectively, it's likely that you won't be able to tell.
This sort of post is increasingly coming off as high and mighty, where the user thinks they are being exceptionally creative and other people who are using AI are using it mindlessly.
Also, writing, indeed any making activity, forces you to think clearly. Often that is the deepest purpose of prose, even code - forcing the author to confront the problem and sharpen their own understanding of the world we share. That it then communicates something to another person (or to a machine), is a happy bonus.
> Reading a document like this isn’t just difficult—it is punishing
I stopped reading here. I’ve become so sensitive to LLMism, I don’t want to go on. Tip; omit the needless ”it’s not X, it’s Y” and just write ”it’s Y”.
The problem is that whenever someone writes something they are accused of using AI. I know people who are accused of this because they know how to use em dashes, or even just write bullet-points. Reminds me of the bores who accused me of using Google Translate for something, and I had to show them that the language in question wasn't even on Google Translate.
If you are looking for top-notch corn-fed drivel, look no further than my comedy newsletter. Here’s the latest issue: https://yelluwcomedy.substack.com/p/sweater-weather?r=7c090l...
It’s full of all the questionable wonders only a human mind can think of. Hope you enjoy it.
On Macintosh keyboards it's ⌥⇧-. On the iOS soft keyboard, it's a long-press on the hyphen key. People often point out that various software will turn a double hyphen -- into one by default. We're overflowing with different ways to type an em dash, for those with any curiosity to learn.
As I have been saying for years:
Writing is fundamentally the transfer of information from your brain to my brain. If you have 1000 bits of semantic information you want to transfer, you can't give 300 bits of semantic information to an LLM and have it fill in the remaining 700, because it doesn't know what those 700 bits are. If it's able to guess those 700 bits correctly, then they aren't true semantic information, and you really only have 300 bits you want to transfer. You might as well transfer those bits to me directly, rather than having the LLM add on an extra superfluous 700 bits that I then have to filter out.
For a long time after the internet arrived on the scene, a lot of online news stories would reference websites, papers, polls, etc. without linking to them. There are still news sources doing this today. Sometimes such articles interpret or place context around their hidden references, but a lot of the time they just summarize.
Giving someone the text output of a LLM is very similar to publishing a summary without links to the referenced material. When you were querying your LLM, you could have asked specific questions or asked for a custom focus or point of view. Your intended audience might have questions or different concerns, but they're unable to interact with your LLM. What you have delivered is static and unresponsive. It has all the disadvantages of being machine output without the advantage of being interactive, the way your LLM was for you.
It may have to wait until compute is cheap enough that tokens are essentially free, but we need a system to pass "hyperlinks" to LLM's primed with context, ready to be interactively queried on a chosen context. It's being overly generous to assume that people are putting even 300 bits into a LLM for every 1000 bits of regurgitated writing they try to pass off as their own. When people post LLM output as if it were their own, I have no choice but to assume they had zero knowledge of the subject, but this query taught them what they wanted to learn, and now they're sharing that. That's fine, but please pass an interactive LLM link rather than static text.
Once we have "hyperlinks" for LLM sessions, perhaps we can share LLM output a little more usefully and honestly.
I've seen professional journal pieces refer to science journal articles only to go and read the original article and find that it draws a different conclusion than what is implied by the journalist.
My father used to complain about that 40 years ago, though in his case he was reading newspapers rather than professional journals. But he'd point it out to me often enough that I started to see the pattern. Scientist publishes paper saying "We may have found evidence of X, which suggests the possibility that Y may also be occurring". Journalist: "Scientists find X which proves Y".
This has been happening for decades; I still see it happening today*. My cynical suspicion is that words like "maybe" and "suggests the possibility" don't sell enough papers.
* Worst offender I can remember was actually from the summary of a paper published on the research institution's own website, so I couldn't blame it on "Oh, the journalist misunderstood what the scientist wrote". Summary said "Exposure to X can, on average, cause a 40% higher chance of Y" (where Y was a negative health outcome). I clicked through to the study and read it. Turned out the confidence interval on that chance of Y was so wide, all you could say with 95% confidence was that exposure to X could do anything from reduce your chance of Y by 5 percent, or increase it by 85 percent, or somewhere in between. They had averaged -5 and +85 to get the scarier-sounding 40% number that they published in the summary, but the truth would have been far closer to "this confidence interval is so wide that we really can't conclude anything from this data". But that wouldn't be nearly as likely to get them grants, so they tortured the data in their summary so that it would look better.
Scientist: "My discoveries are useless when taken out of context"
Media: "Scientists claim their discoveries are useless"
In most cases, they are only useful to other scientists in their field, which does mean that they are useless to the general person until they show up in the form of a new commodity.
There is also incentives to adapt the message to the outlet. If you send that data to peer review and say we found 40% increase, the reviewers would reject it, so they have to moderate themselves. But if you send a summary to the university’s outreach outlet saying that we found something or nothing we don’t know, then they would also reject putting it out. So even for the authors, the incentive is to send a careful conclusion to peer review and an overblown one to popsci.
Btw, I also think a 95% confidence interval is just the wrong statistic to look at given that data, and that they could probably have analyzed it better.
> Worst offender I can remember was actually from the summary of a paper published on the research institution's own website, so I couldn't blame it on "Oh, the journalist misunderstood what the scientist wrote".
It's a similar thing. We live in the "attention economy", and research institutions - particularly after the US President openly went and had his minions cut funding to research purely on ideological reasons, but it's been a problem for decades - are just as susceptible to blow stuff out of proportion to make headlines and thus increase the chance someone might throw some money over the fence.
And media does the same, just to manufacture artificial debate. And so do politicians.
And frankly, I'm fed up with that, we will drive ourselves into a wall.
> so I couldn't blame it on "Oh, the journalist misunderstood what the scientist wrote"
It's never the case that someone misunderstood what scientist wrote. Much like the scientific papers, news articles, including those reporting specifically on the discovery, have their own goals, and the paper being cited is used as evidence or argument for article's own "study". Except for press, the standard is rhetorical, not scientific, it's the conclusions and not the methods that are "pre-registered" at the start, and claims are defended by "hey it's just a point of view", not by statistical significance.
In your own example of worst offender: the scientific study was trying to establish and quantify the connection between X and Y. The summary article was trying to push the angle that "this institution is doing important work". It started with that conclusion, and the paper cited was just the first thing the author found that could be easily massaged into supporting that conclusions by rhetorical standards.
Same paper might get cited by journalist trying to push for "X is bad for you", and they'll do roughly the same as the summary article. And, same paper may be cited by someone claiming they have a miracle cure for Y, and they'll make a honest observation that "absence of X reducing Y is a common bullshit claim based on misunderstanding the paper [citation], that actually shows there's no correlation there, I mean look at the confidence intervals, even the author says that in text nobody bothers to read"... - citation may be honest, but the article itself is still using it to prop up a different flavor of bullshit.
TL;DR: don't believe news. It's bad for your mental and physical health (p<00.05).
> It's never the case that someone misunderstood what scientist wrote.
This is wrong. Reporters frequently don’t understand the science or the nuance in the science.
Reporting and science are two very different disciplines. Reporters rarely have a deep background in science and almost never have a background in the specific area that they’re reporting on.
Hell, even scientists have trouble accurately describing the work of a different scientific discipline.
Don’t invent bad faith motivations; they exist but most of the time it’s just two people slightly talking past each other.
> Don’t invent bad faith motivations; they exist but most of the time it’s just two people slightly talking past each other.
I'm not inventing them, but maybe conflating two sources:
1. Malice directly intending to hurt or defraud people. Probably not as common as how I make it seem.
2. Not caring. Well, I subscribe to the view that not expending effort to be accurate when talking to other person is as bad as slashing their tires (paraphrasing an old quip), so I very much consider bullshitting and picking a conclusion and then massaging facts to fit it, to be acting in bad faith too.
Nah, even in middle and high school it was plain that people sometimes misunderstood what the teachers meant. It's not like that goes away in adulthood; even when people "care", they still often misunderstand nuance or details, and sometimes even the bigger picture.
Honestly, it's plain weird to say that people never just make mistakes.
PS - worth adding that "I misunderstood" and "I didn't care enough" are not mutually exclusive. You can do both, so saying "they didn't misunderstand, they just didn't care" isn't a reasonable rebuttal. But even setting that aside, there'll be plenty of folks who care but still don't understand.
I'm gonna invoke Hanlon's handgun here: not attributing stupidity to what's adequately explained by systemic incentives promoting malice.
I normally assume people make mistakes. I don't believe this is a good explanation for news publications, university press releases, politicians, etc. because those are organizations with agenda, and commit "misunderstanding" of this type pretty much in every thing they publish. The pattern here is pretty conclusive, IMO.
> I normally assume people make mistakes. I don't believe this is a good explanation for news publications, university press releases, politicians, etc. because those are organizations with agenda, and commit "misunderstanding" of this type pretty much in every thing they publish. The pattern here is pretty conclusive, IMO.
The pattern of personal motives fits misunderstanding. You need to show that there is an organizational pattern of "malice" (your word, not mine), rather than an organizational pattern of "we are trying to publish quickly, and quality accidentally falls to the wayside". I.e., negligence, not malice.
You haven't provided even a shred of evidence suggesting there's malice at the journalist level. Every science journalist I have met genuinely cared about the science (which is why they were writing on it), but they didn't have time to learn enough about the subjects to understand they were oversimplifying things.
this is such a bad faith Reddit-tier comment.
Try checking yourself on next few articles you read, and see how quickly you realize it's the actual reality.
Did just that with your comment. Seems like an exercise in discernment.
A lot of his takes on here are like this.
I have recently reviewed a paper that referenced my own article… but the conclusion was so off, I actually went back and re-read that entire article just to be sure there was no hint to the conclusion that the author derived. There was none. Uncanny experience.
Not just news stories. This is a huge problem with social media and forums in general too. Lots of people making bold claims about random things, lots of stories they say are based on a third party source, but very few actual links to said sources in question.
I still remember a recent example where one of those trivia accounts on Twitter posted an interesting story about some guy whose life completely changed after an accident, but neither linked to a source or named the person in question.
The only way I was able to verify it was true was through someone in the comments asking the platform's AI chatbot, and the chatbot providing context that I could research and verify...
This sounds a lot like knowledge graphs. Ideally, we can hyperlink to not only articles but concepts, facts, etc.
I’m a big fan of this approach.
> For a long time after the internet arrived on the scene, a lot of online news stories would reference websites, papers, polls, etc. without linking to them. There are still news sources doing this today.
I agree, this drives me crazy. Ironically, one of my favorite uses for Claude is to ask, "What study is this news article talking about?"
It's pretty good at digging up the source and related sources. And most of the time, if you read the source, the article is nonsense and gets everything wrong.
Link rot has accelerated in the past decade. Links are great, when they work. If they are a few years old, they often don't. You certainly can't count on it. If you're referencing a paper or an article, it's probably better to cite the title of the article, the name of the journal or magazine or newspaper it was published in, the author, and the date. Then someone might have a chance of finding it again.
Doi url
That works for pieces that have one, although places that index those aren't always free.
> For a long time...a lot of online news stories would reference websites, papers, polls, etc. without linking to them.
The "$CITY_NAME Business Journal" websites are the absolute worst with this. They'll refer to something specific, for example "$BIGCO's 2025 10-K filing" and it will be a link. That link will go to the 10-K, right? Nope! It goes to another page at the same business journal. Maybe that page is a summary of the 10-K, but probably not. Maybe it's just the general index page for all the articles about $BIGCO at that journal. What it links to, it definitely won't be the specific thing described by the text of that link.
> There are still news sources doing this today.
It's the opposite, all big news websites do this. Fairly sure it's part of the policy.
I would say only small, niche websites link to sources.
It isn't the transfer of information at all. What's actually happening is you're prompting experiences in a human instead of an AI using text.
Communication only works if you have multiple levels of representation and abstraction, including but not limited to - letter shapes, grammatical structures, style and register, stylometry, and subtext.
All of that is learned, and writers usually assume they can rely on that learning as the context for the text.
So you don't write to 'transfer information' like a network cable, you write to trigger experiences in the human version of latent space.
Factual information is one kind of experience. But even when that's the goal, there are always layers of implied relationship, social register, role, status, and other implications in everything that's written.
In normal communications the context - business emails, personal messages, mainstream journalism, fiction, and the rest - defines what acceptable language looks like.
The content fits inside that. But it has to fit the context, otherwise it lands in a semantic and psychological uncanny valley - like sending LinkedIn speak to a spouse on a wedding anniversary.
The real problem with LLM writing is that it's good at the technical layer - the grammar and spelling - and has some insights into the rest.
But the default content style is marketing and ad speak. And recently it's developed a weird and unique hybrid style which applies marketing fluff and pretension to technical content like code comments.
So you get one register instead of all of them. It can attempt others, but it's still too limited to generate them fluently. Sometimes the results are outstanding, but often it defaults to mechanical clichés.
So that's why it sucks and sounds so hollow.
Can it be fixed? Yes, but it's very hard work, most people don't have the skills, and it takes time - often too much time to be worth the effort.
I'm definitely being slightly too emphatic when I say it's "fundamentally the transfer of information", but I don't think that nuance is important.
When LLMs eventually get good at writing in the correct style for a given context, I'll admit that they have value in that way. But they aren't good at that yet. And even when they do get that good, I'll still dislike it for reasons that are more emotional than rational.
I don't care how good the LLM gets. If I know some text was written by LLM I'd much rather know the prompt - the seed of intent.
If the seed of intent is "convey XYZ details so they know them" then I can choose to go and learn those details any way I see fit - maybe even ask an LLM to summarise some data for me! - rather than having to ingest whatever their LLM use poops out and trying to digest the intent and content and figure it out.
It is about empowerment, rather than eating shit.
When sharing summaries from calls that I attended but my team has not but I think they would find interesting and I simply don't have time to handcraft it myself I use LLM and attach my prompt and the transcript.
Attaching the prompt initially threw some people..."wait, your admitting to using AI..."..."err yea, unlike you with that PowerPoint you sent me last week". I sense this is the right way to go imho.
This is one of the weird things about LLM use. It is a handy tool.
If you have found a prompt that gives a good result, then by sharing it with me I now have a tool that gives a good result.
The problem is not the use of LLMs. It is the "passing it off as your own work".
You're focusing on the wrong thing.
Output is not interchangeable with the prompt. In many cases, the prompt does not have the information the sender wanted to give you, and there is no guarantee that your LLM will give those information - or do it correctly - if you use the prompt yourself.
The entire value of here is that sender read the output and is vouching for it. This is where "bits of information" come from. If the sender cannot be trusted to verify and vouch for the LLM text they're sending to you, well, they're an asshole and you should rebuke them or find someone more considerate of others to talk with. Them giving you their prompt doesn't help you with anything.
I am focusing on the thing that is usually hidden. You are correct that for a full picture I'd need the information the LLM operates on.
But my point is that the prompt is the nearest encapsulation of the intent of the originator. Give me the data and the prompt. Vouch for the output if you like, but I want the source.
This is what's so amusing to me. People say they're not interested in the output of an LLM, only what a human has to say. But then when a human says "These words from the LLM are good, I vouch for them" the very same people say "If I wanted those words, I'd get them from the LLM myself, what's the point of this human at all?"
These humans are behaving worse than the LLMs at this point.
> If I know some text was written by LLM I'd much rather know the prompt - the seed of intent.
Here's the proximal prompt "Okay, take everything we've been talking about for 2 hours and apply those edits to the the final draft for publication."
What exactly does that give you?
If you give me that. Then you've given me nothing. Give me the same transcript the LLM has access to instead.
Anyone know of any good research articles on the Information-theoretical aspects of this? I find it a compelling topic.
While I agree with some of what you've said and the conclusion you've arrived at, in the end, I think you've missed part of the picture here with regard to "prompting experiences".
> Communication only works if you have multiple levels of representation and abstraction, including but not limited to - letter shapes, grammatical structures, style and register, stylometry
These are methods of encoding, there's no reason all of these can't be represented in an LLM from a technical point of view.
> and subtext.
This is the other half of the equation to me. Humans communicate by relating shared experiences, an LLM cannot have shared experiences. While it might be able to encode subtext that has been specifically called out and explained, it will never be able to encode the breadth of human subtext, especially that which is reliant on emotion.
I don't believe it is possible to change this until the point mankind truly develops a "wetware interface" to the digital world (and I personally don't want such a thing to exist).
That "marketing style" isn't a style, it's the lack of content itself. You just restated their original point despite trying to object to it, because it was correct and there is no way around that.
I don't agree.
Sometimes Claude's problem, such as when I ask it to summarize a long, complex session back to me, is it's too information dense. It uses weird invented terms to gloss over complex parts of the architecture instead of explaining them.
But no matter what - too dense or too sparse - it always sounds like Claude.
It misses the forest for the trees. It feels the need to highlight details not understanding what details are most relevant to a human reader and how to survey the larger problems in a cohesive way that emphasizes the right parts without cliche and undue emphasis.
What already happens: - People give an LLM a bulleted list of points that they want expanded into a professional sounding document. - The receiver doesn't wanna read all that. They put the full document into an LLM and ask it to summarize it into succinct bullet points.
We've invented the opposite of lossless compression
Lossy expansion or bloat?
Depends on if you are buying tokens or being paid for tokens
Students where doing that for the ages when they have to answer questions like. "Answer the following question in not less than 4 pages"...
Seven bullet points squeezed into five pages of text
A broken telephone.
TAM: $50T
It's been a huge boon to hardware vendors.
This is exactly what I want: To communicate with me, have your LLM expand your message to include relevant parts of _your_ context, then I will have my LLM summarize it as briefly as possible with respect to _my_ context.
Just ask them for their prompt, and then put that in your own LLM with your own context. Why do you need their LLM's fluff?
I first noticed this a year ago when I read a gmail AI summary, then glanced at the main text and saw it was AI generated. I feel like there's good fodder in there for a dystopian sci-fi story about a future where nobody communicates directly with one another, it's all AIs translating, but the AIs slowly start to drift.
>but the AIs slowly start to drift.
It's essentially the tower of babel. Each person will devolve to speak their own internal language only they understand. Each language will need to be encoded down to its meaning to be reinterpreted. None of us will know if the transformers are accurately decoding, or if the other person is accurately interpreting the decoding (which is arguably already a feature of human language without the computers in-between.)
Agreed re: the sci-fi story / trope.
I feel like a lot of this is a problem when someone technical is attempting to communicate a complicated technical subject to a less-technical audience.
I can only dumb a thing down so much before the description is useless (when you zoom out too much you lose the details). Even technical people who could understand it but are lazy / "in a hurry" use the summary, without thinking about what detail they are losing.
Even more infuriating is when they then reply to my email, having only read the AI summary, and ask a question that was already answered by my message.
This is the exact same thing that happened pre-AI, with the added step of wasting energy/resources on the AI summary in the middle.
It was in a gmail promotional video years ago already. Person A would use AI to turn their summary into email and person B would ask the AI to go back to the summary.
They knew it was going to be like that from the beginning.
E. M. Forster wrote that story in 1909.
https://web.cs.ucdavis.edu/~rogaway/classes/188/materials/th...
Adrian Tchaikovsky has a good take on it as well, Human Resources.
I’ve been saying the same thing. I’m not saying all, but a significant amount of comms could be bullet points to the benefit of both sender and receiver.
Axios built a hefty business off this simple idea
It's worse than that. If it takes others longer for others to consume and understand what you're producing than it does for you to produce it, you'll never be able to communicate with someone efficiently. Communication breaks down at a fundamental level if if you can't keep up with the other side if outputting and they won't slow to allow you do do so.
I see this all the time now with LLM generated output. It's easy to have an LLM generate a chunk of content that can be dropped into a chat or comment, and when it took you 20 seconds to have something written up based on the shared understanding you and an LLM have about the context of the situation, but it takes other people 3-5 minutes to read and understand that content, that fundamentally doesn't scale. It's bad enough when one or two people are doing it, but if the whole team is doing it, the only way to keep up with the stream of information is to also consume it through an LLM. At that point you're likely to be missing much of the nuance, and the amount of errors will explode.
This can be alleviated by people reviewing the output of an LLM and making sure it both includes fundamental information that might be assumed by context and reducing it to the parts that are essential for the new context it's in. This takes time, but is extremely important.
Having an LLM write gobs of text to send to other people instead of doing it yourself is the equivalent of a low yield cognitive zip-bomb. Don't do it.
I don’t quite think this tracks. Perhaps you want to communicate 1000 bits that are well known and can be referenced with a 300 bit key. Then the LLM can easily retrieve the remaining information. It’s like sending someone a link to the Wikipedia page instead of explaining something yourself.
No, I don’t want to read LLM writing because it is BAD at it. It doesn’t really understand how humans think (because it thinks differently), and doesn’t seem to understand core principles very well (presumably due to the lack of world model), so it can’t write something humans enjoy yet.
That's kind of my point. If the 1000 bits are well known, then their inclusion isn't new semantic information. By pasting an LLM's output, you're deciding for your reader that they don't already know that information, and deciding that your LLM prompt is better than whatever they would to to obtain that information if they lack it. IMO, it's much better to give your readers the 300-bit key, and let them decide for themselves if they want/need to get more information, and if so, how.
Mmm perhaps, but I think you normally have an idea of who your audience is. If this was actually how we talked then I would have just responded to this comment with “known-audience rebuttal. Example. Audience knowledge clarification. Example absurdity comment” and let you work it out. I think it’s normally known to you and the LLM and not known to the person you’re speaking to. Speaking like that would just cause confusion and misunderstanding.
> I think it’s normally known to you and the LLM
I disagree. Unless you gave additional information to the LLM yourself, the LLM doesn't know more than your audience does about what the meaning of such a comment would be. An LLM could certainly come up with something plausible, but it wouldn't necessarily be what you intended.
Why not do both though? I often now write my human summary, then paste in also the content you could get by asking a bot (with markers for which is which). You’re perfectly welcome to ignore the bot text, or ask your own bot, but bots are pretty slow, so I also don’t want to wait for it to “decompress” that 300 bit blob to get the detailed page back. But my human summaries are also only intended to be 50 bits — compressed again to just pass the signal I mean to convey, and not the whole prompt needed to fetch the right info from the right place to substantiate the claim.
TLDR the length was the same curtesy of tl;dr before, just now with a different name.
If only that were the HN we comment on. If a post doesn't spoon feed an explanation for an initialism thats the tiniest bit off the beaten path (eg BGP), there's inevitability a comment about how would it kill the author to define BGP?
I’m not sure the 300-bit → 1,000-bit framing applies in all instances. The 300 bits may be a compressed cue to a much fuller idea. The AI can combine that cue with its prior knowledge to help reconstruct what the prompter was trying to express, with the prompter then verifying whether it’s right. Without the relevant prior knowledge for reconstruction, or the prompter for verification, it becomes much harder to know whether you’ve reconstructed the intended idea.
Unless that 700 bit was transferred on a separate occasion the inferred 700 bits is not true information, anyone could have reconstructed it from the 300 bits.
Not anyone, no. From an information theory standpoint, that it's possible at all to complete these 700 bits, only implies that anyone logically omniscient could. It's entirely possible that an LLM is capable enough to infer these 700 bits, and the human reader isn't.
Right but where is this information coming from, it can't come from the sender because an idea that can be conveyed in 300 bits cannot contain 1000 bits of information. You're just using the LLM to translate the idea into something more legible.
Alternatively those 700 bits are information the LLM added, but where is that information coming from? Is it noise? Random facts? Random lies? And who is the receiver even talking with if most of what they read is something the sender didn't know?
No, the 700 bits come from the sender verifying and vouching for the information before sending.
Unless that takes 700 attempts on average I don't think that actually works.
That assumes each bit is a coinflip, doesn't it?
Even Markov chain autocorrect tools do better than 50% odds*, and even GPT-2 was significantly better than that kind of autocorrect.
* at the word level; IDK how redundant/efficient language is when it comes to bits-worth-of-fact-claims-per-word. But "your cat is sitting on my" -> [mat, laundry, roof, head, belly, laptop, microwave, …] clearly has many bits of information, and a Markov chain will encode the most likely next word even if the user doesn't know what the most likely next word is. Verifying where the cat is sitting is also very easy, as is correction.
In information theory, each bit is a coin flip by definition
I realise I phrased this poorly.
I will try harder. Consider entropy.
The first sentence in this comment contains 32 characters; from the point of view of a naïve channel with no compression, that's 256 bits (given none require breaking out of the first bytes of UTF-8).
It did not take 2^256 attempts to construct the first sentence in this comment, because the generation process was not flipping coins per bit.
LLMs also do not emit bits chosen with a [0: 0.5, 1: 0.5] probability distribution.
From a compression point of view, the bits-transmitted-per-bits-in-message ratio can be reduced such that more likely messages use fewer bits than less likely messages. However, this requires the receiver to agree with the sender what the probability distribution over tokens is.
Intelligence is, amongst other things, a compression algorithm. If I can predict your next token, and we both know this, we can agree in advance that you don't need to actually send it.
No single human brain is able to predict the output of an LLM anything like well enough to do that.
In entropy terms: LLMs are noisy sources, their output does contain false statements, yet they add more bits of signal than of noise relative to a human alone.
Or at least, they can add more add more bits of signal than of noise relative to a human alone, but humans who blindly copy-paste the output of an LLM without checking are a pain and add zero value to whatever situation they happen to be in.
For some hypothetical scenario, writing software because I know they can do that, asking an LLM to write some code for you may easily give you 10 kilobits of positive information (code that mostly works), and -30 bits of noise (each bit being one binary decision's worth of incorrect choice by the LLM in what to write, i.e. bugs); if you as a user don't know how to handle the -30 noise that could easily be a totally useless app, but if you can filter out 30 bits of noise, either manually because those 30 bits happen to be your skill set, or even in some cases by prompting it again with the failure mode, then you get to benefit from the 10 kilobits of good stuff that you didn't have before.
In many (but not all) cases, LLMs can fix more than 1 bit of mistakes per follow-up prompt.
I assumed a coin flip, indeed in practice you would likely achieve far less than 1 bit per attempt.
Far more than 1 bit for most attempts: they need that just to be able to write coherent sentences, and a lot more to be coherent sentences on the right topic.
Some specific conclusions would be far less than 1 bit.
The average will depend on both the question and the AI.
This is about information from the sender to the recipient. It is fundamentally impossible to transfer more than one bit in one binary decision.
A LLM adds noise, not information. At least in this framing.
Frankly, it's a stupid framing.
LLM is not a random symbol generator (hint: training data is not random), and no reasonable person is going to just prompt an LLM and send its output without giving it at least cursory check (at the very least so that blatantly stupid hallucinations don't paint the sender as inconsiderate or incompetent).
That check alone can add bits to the final signal.
I'm not sure we're working in the same framing here. For one LLMs are random, sure you can fix the seed but you don't have to, you can even replace the RNG with true random noise.
So there are 2^300 possible ideas, only 2 outcomes from the cursory check, how do you get 2^700 outcomes? Most of those are just random variations the LLM added which is not a transfer of information. You would be lucky to even identify which of the 2^300 ideas was being conferred.
"Random" is too loose a word, but they were responding in a context where it meant coin flips.
LLMs are not even odds on all possible outputs, they are biased towards patterns which are upvoted by the training mechanism (at a minimum: the source material, RLHF, and synthetic data).
The information any trained model transfers to output, is information it gained during its training.
No single human is capable of having consumed all that training data.
> and no reasonable person is going to just prompt an LLM and send its output without giving it at least cursory check
I'm reminded of an old quote:
Touché.
Yup, it helps when the prompter reads and edits the AI-generated text, but usually they just skim it and send it unedited. Worse, it's usually not 300 bits -> 1000 bits (= a concise message covering all the relevant topics), it's more like 300 bits -> 3000 bits (= typical LLM verbal diarrhea hiding the relevant points in a wall of text).
This is true - but also, models are absolutely terrible at writing articles and I don’t want to read them.
The issue isn’t that a 300 bit idea is padded with 15 KB of content. You can take any human-written article and reduce it by 90% with next to no information loss. What you lose is what makes the article a compelling read instead of a fact table.
I think the reality is that we will see quality long form AI-written content at some point. It doesn’t even feel like labs are particularly interested in chasing that now; code sells way more tokens. Right now the trend is that subsequent models degrade in writing quality as long as that pulls them up on coding benchmarks.
> You can take any human-written article and reduce it by 90% with next to no information loss.
You've unintentionally circled the error here. The "purpose" of an article extends beyond "convey this essential information".
By analogy, a textbook contains far more words than a spec sheet, but attempts to train the human to be able to easily interpret spec sheets. The so-called "information" content of both might be equivalent, yet one does a better job of teaching students.
I was going to write a post disagreeing with this on the basis of the fact that the reader lacks the background information the LLM has. For example, if I were to prompt "explain the proof of quadratic reciprocity using Gauss sums" most readers would need the entire LLM's answer (and much more, probably) and not just the prompt.
But then I realized that the reader can prompt the LLM with the same prompt for the same or equivalent expanded text. Most people don't do this as it's extra effort, but it's interesting to imagine a world where this is the default way of engagement with a text, assumed by both writers and readers alike.
> But then I realized that the reader can prompt the LLM with the same prompt for the same or equivalent expanded text.
That's basically what I've been asking my colleagues (so far a losing battle): Please don't send me AI-generated text. Send me your prompt instead. It is highly likely that I will understand it without needing an LLM, and if not, I can do it myself.
No, thanks. Depending on the type of information being transferred and just how uncertain the recipient's knowledge is, I may or may not need to add some amount of information to whatever I have to share. That information would be extra context you may need to have, on top of what new thing I am trying to explain. If I know I'm talking with someone who's all caught up, great, no need for context setting. Though even in those cases, there might be some terminology I want to define, or to provide a list of acronyms or whatever. That is boilerplate work, but is at least useful, if not needed. I know people tend to view this sort of comms as "technical" - they shouldn't be. So much misunderstanding happens because people are just not aligned on their shared context or medium. Of course, AI is not fully reliable to generate that context, you as a writer need to put in the work to review and align the result with your own understanding, or drive it thoroughly before.
All that being said, I acknowledge that people (me included) love to be sloppy in their comms (with or without AI) and then blame others for misunderstanding. It's also unlikely we'll change soon. What can you do.
As with everything, it depends how you use it.
I’ll often put a long stream of consciousness on the page, or jot down rough meeting minutes, then ask ChatGPT to “summarise this for an email”. The result is shorter, clearer and easier to read.
AI amplifies the habits of the person using it. If they’re lazy or dim, then it's like giving a monkey a gun.
I like this example and it made me think, there is an analogy here to spec-driven development and vibe coding.
Rather than send 300+700 bits, like you said, send 300 (or less!) and let the human intelligence on the other side generate the result. Which supports the even older perspective: “If I had more time, I would have written a shorter letter.”
I’m not sure if this lands on anything very profound, but what about a pattern where, instead of codifying agent output at all, the only artifacts we share are the prompts. And the rewards (respect) accrue to those who generate the most generative among people and AI
So I spent a full workday investigating an issue with an agent and then I spent another hour prompting the agent to summarize the findings so that I can send them to a colleague. How does what you're saying come anywhere close to describing my work?
> If you have 1000 bits of semantic information you want to transfer, you can't give 300 bits of semantic information to an LLM and have it fill in the remaining 700, because it doesn't know what those 700 bits are.
LLMs do inference or computation among other things, so the remaining 700 bits can be something like that. The hidden implication in your claim is that computation adds no information content, which leads to an interesting philosophical discussion.
So for example, if I ask an LLM to give a proof or derive a new theorem from a set of axioms, according to your assumption, if it answers correctly, then I haven't learned anything new.
I am not really sure how to resolve this paradox in information theory.
If you input 300 bits into an LLM, and it outputs an additional 700bits of new information, I’d still rather you tell me those 1000bits than read ann llm’s output of 1000bits.
The primary reason is that human language is becoming a proof of work, that speaking out loud or writing directly indicates that the idea is important enough for a human to express. This is more costly than llm output, which is often just botspam.
Someone mentioned a similar thing elsewhere in the thread, that by communicating those extra 700 bits, the sender also implicitly vouches for them.
From a purely information-theory perspective, the simplest solution is to say that yes, any content derived from existing information carries no information itself.
From a realistic perspective in the context of people copy-pasting LLM output, my thoughts are that asking an LLM to research for you is more defensible, but it's still better to read the LLM's research results and write the important parts in your own words (partly because the LLM probably used way more words than necessary for the context).
> Writing is fundamentally the transfer of information from your brain to my brain. If you have 1000 bits of semantic information you want to transfer, you can't give 300 bits of semantic information to an LLM and have it fill in the remaining 700, because it doesn't know what those 700 bits are.
Sure you can. LLM doesn't know what those 700 bits are, but you do. You may not realize it, and may not even know it at the time of prompting, but you do by the time you're sending.
Typical case is like this: you have 500 bits of semantic information to transfer. You give 300 of them to LLM, and get back the 500 bits you knew you have, and extra 500 you can quickly confirm are correct and relevant. Some of them are just dereferences of your input - where you recalled a pointer, but not what it pointed to. Some of it is information you never had before, but are able to easily validate.
You send that to me. I likely immediately realize the message was AI-assisted, but I trust you to be a decent human being, and not an asshole that lobs unverified LLM vomit over the fence for others to deal with. End result: you communicate 1000 bits of information to me, instead of planned 500, and you yourself learn extra 500 bits.
This is the optimistic scenario, but it does happen when LLM operator is not an asshole.
(Excuse the strong language, but I spent a lot of effort every day on both dealing with inconsiderate people lobbing LLM output at me, and making sure never to act like one myself, so it's a topic close to my heart.)
I think you missunderstand. If the llm can "add back" 200 bits of information, these 200 bits are superfluous by definition. They can quiet literally be inferred from the starting 300 bits. And the llm has proved it.
If the receiver wanted these 200 added bits, she could infer them either herself or even use an llm to do it.
In principle, all of mathematics can be inferred from the axioms. The field of mathematics is not superfluous and something a receiver could infer themselves if they wanted it.
My brain does not contain all the information that can be added by an LLM; a human brain could contain it, even the biggest LLMs are about 1% of the (if you approximate synaptic count ~= parameters) parameter count of a human brain, but none actually will.
What my brain may actually contain is the information necessary to verify the (in this example) more than 200 bits the LLM claims to have added and trim out the parts which are false, retaining the (in this example) 200 "new" bits of new information added by the LLM.*
Concrete example: I am a software developer by training, though not a web developer. If someone who does not have any developer experience asks me to make a web app, I am forced to use an LLM as I do not know enough JS etc syntax to get it done myself. But as we all know, LLMs are only "ok" but not "good" at making software, so there are a lot of rough edges and outright mistakes. My experience as a software developer extends to detecting such failures and I can usually correct them.
The original person, someone who has no developer experience, can also prompt the LLM. Right now, this would result in something that retains all the errors, because they didn't have someone like me intermediating between them and the LLM.
I add bits by removing noise, the LLM adds bits but they contain noise.
I do not know for how long this will remain true, but today it is true.
* Feels like P versus NP to me. The answers AI generate are at their best when they're easy to verify. Then again, when they're easy to verify, they can be RLed to get good at this quickly and the need to verify goes down, leaving them still pretty bad at things that are hard to verify.
> If the llm can "add back" 200 bits of information, these 200 bits are superfluous by definition.
No, they're not. Getting those bits takes energy.
SOTA LLMs know way more than any individual on approximately anything there is to know (and what they don't, they can look up faster than people can). It's very easy for them to make the "missing" 200 bits explicit, rather than implicit, which in practical terms is the same as adding 200 bits that weren't there before.
Theoretically, an idealized omnipotent mind / AGI could derive the unifying theory from reading your HN comment on a phone screen. There is enough information there, if you were able to extract every bit of evidence available from it. But you are not. Neither am I. It would take us practically infinite work to try, solving this most cruel mathematical riddle.
How often do you modify the LLM output before sending it? If it’s less than 50% of the time, it means that by not modifying it, you have added at most one bit of information to what you originally wrote. (If you don’t understand why, think of it this way: Instead of sending the LLM response, you could send the prompt and one extra bit indicating whether the LLM response to the prompt should be modified, followed by the modifications.)
> How often do you modify the LLM output before sending it?
Me specifically, I never send anyone LLM output I haven't give at least a quick read (not skim, read) to make sure it's reasonable and there is no obvious bullshit there. And then I still mention it's LLM-sourced.
> If it’s less than 50% of the time, it means that by not modifying it, you have added at most one bit of information to what you originally wrote. (...) Instead of sending the LLM response, you could send the prompt and one extra bit indicating whether the LLM response to the prompt should be modified, followed by the modifications.
It's not the case, though. Prompts are not interchangeable with output. There is no guarantee that if you send a prompt, and recipient passes it to their LLM, they'll receive anything similar to what you did. It may have mistakes - different mistakes - or just spend focus differently.
The extra bits I claim LLMs can add to the message hinge strictly on you vouching for the response. Of course, you can just prompt an LLM, learn from the response, and then write your message clean, containing both the bits you originally had, and the bits you gained. But at that point, the LLM already gave you text containing all those bits - if you can vouch for it, you may as well copy it over and save yourself the trouble.
> It's not the case, though. Prompts are not interchangeable with output. There is no guarantee that if you send a prompt, and recipient passes it to their LLM, they'll receive anything similar to what you did. It may have mistakes - different mistakes - or just spend focus differently.
I’m not saying that the response is interchangeable, but that due to the data processing inequality, it cannot convey strictly more information than the prompt.
> The extra bits I claim LLMs can add to the message hinge strictly on you vouching for the response.
My argument is that if you vouch at least 50% of the time, the vouching only adds one bit of useful information – either you vouch or not.
> due to the data processing inequality, it cannot convey strictly more information than the prompt.
Only in the case where the LLM message is not reviewed before sending, and only if we assume reliable LLM (so that the receiver could recreate the same output if given the original prompt). This is not a realistic scenario.
> My argument is that if you vouch at least 50% of the time, the vouching only adds one bit of useful information – either you vouch or not.
The alternative to vouching isn't "not vouching", but "correcting and cutting out wrong bits and vouching for the rest", which means the single "vouched for it" adds all the bits that are in final message but weren't there in the prompt.
> Some of it is information you never had before, but are able to easily validate.
This is exactly the use case an LLM might (huge emphasis on might, depends on workflow, agentic vs. relying on contextual which can hallucinate) be good at and yet humans are notoriously bad at, because we are swayed by emotional responses and it is easy to have an emotional response to text that is programmed to look good for you and you alone.
> you communicate 1000 bits of information to me, instead of planned 500, and you yourself learn extra 500 bits.
Extremely optimistic. If this were the ideal scenario, you would USE the LLM to garner information ABOUT those 500 bits and then reframe them in a way that you yourself would put it. If there is insight, your "word" in your mental register now expands from the original 1000 bits to 1500 or 2000, and then are "processed" by your human brain that includes subconscious choices that are meaningful to the end result. There are tons of hidden semiotic data in your diction and wording (think resource forks in classic MacOS/HFS, only visible to the filesys) that is lost when you rely on another source to put together words for you; it's as if it is a game of Telephone. These are subtleties which you may intend for your recipient to receive and which are crucially important to your recipient and are irretrievable, it is intrinsically lossy. You have an alphabet soup of words, they cannot be put together by an LLM in exactly the way your brain did. We must rely on the fact that we ourselves put this together, the "aha" moment when an LLM does it for you is illusory and does not itself provide meaningfully important confirmation that you indeed say what you mean to say. Of course, humans say things and put things in way we do not intend to all the time. I still fundamentally believe this is more honest than relying on a third party that is not capable of understanding human emotional nuance to put together language for you, when language is and always has been a manner in which to dictate human emotional nuance.
> (Excuse the strong language, but I spent a lot of effort every day on both dealing with inconsiderate people lobbing LLM output at me, and making sure never to act like one myself, so it's a topic close to my heart.)
Does not negate the fact that LLM output itself, at least when used to convey human emotions or thoughts, is lossy. A very highly compressed JPEG with added interpolation from an upscale algorithm might come up with cool details that were never present in the original, and may look cool to both you and the recipient but are not honest to the source material. You receiving 240p JPEGs on a day-to-day basis is irrelevant to this. For the purposes of communication, it is a massive error which has the potential to compound, regardless of whether or not you or the recipient believe this to be the case.
> If this were the ideal scenario, you would USE the LLM to garner information ABOUT those 500 bits and then reframe them in a way that you yourself would put it.
Yes, but at that point in practice we're getting into over-optimizing territory. In this optimistic case I presented, you could learn those 500 bits yourself and formulate a clean message yourself, with all 1000 bits in it, but since LLM already gave you the text, and you feel you vouch for, you may as well send it over and save yourself the effort.
In reality the numbers are probably lower, and writing the message yourself is IMO also a good way to be truly sure you vouch for the "extra" 500 bits, as it forces you to actually pay attention. There's a chance you'll find inconsistency in output, or in your own understanding. I don't begrudge people for eventually cutting the process off here, for practical reasons - it's the fuzzy line between accuracy and perfectionism.
> A very highly compressed JPEG with added interpolation from an upscale algorithm might come up with cool details that were never present in the original, and may look cool to both you and the recipient but are not honest to the source material.
Again, I think it's a wrong take. LLMs aren't pulling the information out of their asses, and you are also not able to express every information directly. LLM can "upscale" information and you can take a look and recognize, "yes, this is exactly as it was", even without being able to write out that "upscaled" version by yourself. Verification is often easier than direct recall.
> Typical case is like this (…)
> This is the optimistic scenario
So is it typical or optimistic?
> I spent a lot of effort every day on both dealing with inconsiderate people lobbing LLM output at me, and making sure never to act like one myself
So why are you so eager to defend your fantastical scenario? It doesn’t matter how considerate you are, truth is the overwhelming majority of people aren’t and won’t be. We’re discussing reality here, not “what could be if we lived in a utopia which will never come to pass”.
> So is it typical or optimistic?
The optimistic case is "having 500 bits, giving LLM 300, getting back 1000, and learning extra 500 in the process". Real numbers are lower. People don't vouch thoroughly and don't catch all mistakes.
But reasonable people don't send every output from LLMs to others without giving it a cursory glance (obvious hallucinations or nonsense would paint the sender as incompetent or inconsiderate), and that alone eliminates the worst levels of noise. A cursory read and cutting out obvious bullshit before sending is enough to make the message carry more bits of information than the propmpt.
> the overwhelming majority of people aren’t and won’t be.
In my experience, the "overwhelming majority" are giving something between a cursory glance and cursory edit; whether the resulting message has more or less information than prompt then depends on how much noise LLM added on top. The inconsiderate people I deal with, they often send "net more bits than in prompt" outputs, but those outputs are also verbose and not fully filtered for bullshit, thus it's effortful to tease out the signal from noise.
While I generally agree, the difference is I can also hand an LLM my pile of code and documents which is... a lot more information dense than basically anything I could write. Sure, a one-off prompt in a chat window isn't very helpful.
Summarization—especially of private context—is definitely one of the main exceptions to LLM writing being useless. However, it's still preferable to turn your private context into shared context and then give pointers to that, rather than having an LLM attempt to summarize it.
Oh fully agree. I hate reading LLM generated text, just wanted to provide a counterpoint to GP.
>If it's able to guess those 700 bits correctly, then they aren't true semantic information, and you really only have 300 bits you want to transfer.
A math teacher only has "this class is about math" to transfer, the rest is known. ;-)
Jokes aside, I think we're in a weird transition now where AI is used to generate text that looks good but is bad. In a few years people will know that and be more critical.
I think we went through a similar phase when DTP had it's breakthrough. Suddenly school papers were laser printed 300ppi times new roman and got more attention than better papers written by hand. But eventually that became the baseline.
I think Ai will make it so that well-written texts with clarity, good layout, correct illustrations, callouts etc become the norm, and will no longer impress anyone unless the information itself is actually good.
And if the information is good, it won't matter if it's AI generated or not.
Maybe things will change in the future, but I know where things stand right now. Would you rather learn about math by interacting with the teacher, by using the internet (including public LLMs), or by having the teacher tell an LLM "teach a math class" and then copy-paste its response to you? Option 1 is the best, but option 2 is at least better than option 3.
Oh it very much depends, your opinion there is certainly not always true and not shared by everyone; you only know where you stand right now with any certainty.
It depends on the teacher. I’ve had good and bad math classes, and LLMs today are quite a bit better than the bad ones. The worst human teacher in my memory didn’t offer interactions. He walked in, turned his back to the class, wrote equations on the board for 45 minutes, and then left. The best math teachers, the ones better than LLMs, are the ones who share the joy and sense of discovery and history of math, and not just the mechanics. But there aren’t that many teachers of that sort.
Learning by using the internet without LLMs is rarely very good, but more often than not in my experience sucks much worse than using LLMs. If you include using public LLMs in internet usage, then it’s not very different from just using LLMs that search the internet. I have heard that a huge swath of today’s high school and college kids are reaching for chatGPT before Google (which is incidentally OpenAI’s goal), and that many of them would rather talk to chatGPT than talk to a teacher. I’m going to refrain from making any claims, but I believe there are a lot of people who disagree with your ranking of the options.
Foreign language learning is one case where I love using LLMs, because it’s not typically an option otherwise. You can practice non-stop and have conversations with someone fluent in a language who will be infinitely patient with your mistakes. This is true of math and other subjects too; using LLMs to practice, so that the human teacher isn’t the bottleneck, to supplement and reinforce the human interactions, is usually better than using the internet without LLMs. The other reason many people prefer talking to LLMs is the lack of judgement. If you aren’t getting it and ask the teacher one too many basic questions, they treat you differently. Sometimes it’s necessary and helpful, and sometimes it’s harmful and takes a long time to change. LLMs don’t do that, they just explain and explain. That lack of judgement is a big reason many people prefer LLM interaction to human interaction.
Okay, sure. The difference between option 1 and option 2 is completely irrelevant to this discussion and I gave what I thought would be an uncontroversial take on it because I thought it didn't matter. The point is at least one of option 1 or 2 is almost always better than option 3. Option 3 is what I and others are arguing against.
Irrelevant? How so? And why did you present two different options if you thought it was irrelevant? I thought you were trying to make a point about human interaction. From my POV, your options 2 and 3 are much closer together than 1 & 2, so if you’re not suggesting that human interaction is preferable to LLM interaction, then what’s your justification for claiming any option is better than any other? I still completely disagree with your clarified point, and suspect many many other people do too, and are demonstrating that preference by using chatGPT for their classes instead of using either the teacher or Google. It’s a simple fact that interacting with an LLM can be preferable to interacting with some teachers some of the time, and LLMs currently often make interacting with the internet more pleasant and more efficient. Your ‘almost always’ still sounds to me like one person’s opinion that is not shared by all.
>Maybe things will change in the future, but I know where things stand right now
It's not an either-or though, LLM generated text already provides lots of value in many situations right now. It's also used for fluff, yes, but much of what humans write is fluff too, reporters often get paid by the word.
The point is that the value in a text has nothing to do with whether it was generated by an LLM or not. What matters is if it's useful or not.
> you can't give 300 bits of semantic information to an LLM and have it fill in the remaining 700, because it doesn't know what those 700 bits are. If it's able to guess those 700 bits correctly, then they aren't true semantic information, and you really only have 300 bits you want to transfer.
It doesn't actually follow, because maybe the LLM is smarter than the original writer (at least in the domain the writing is about) and hence really is able to complete the ideas in a way the writer can't. As an existing example, consider formulating a conjecture and having an LLM prove it. But I agree; if I wanted to read an LLM's output I'd simply ask it myself rather than read someone's supposedly-human writing.
Lets do a thought exercise:
You are stranded in a desert island. You start writing a message "Help, I am..." and pass at that point.
Somebody finds the message. They can no doubt come up with plausible continuations like "Help, I am Robinson Crusoe" or "Help, I am hungry" but they cannot create information. No matter how smart and how long you stare at the message, that is not going to tell you what the original person would have written.
Isn't it from Claude Shannon that information lowers uncertainty? Infinite regurgitation or massaging of data does not create new information. You will get the information form the LLM, not from that original person.
Once you've determined what the information-content of a message is, then you can apply information theory to it. But different receivers can derive a different amount of information from the same message.
Consider, for example, that if somebody doesn't know English at all, then before receiving the message, their best guess at what it is is some probability distribution over all English characters (or sounds, depending on what we assume them to know), and after knowing the first part is "Help, I am" that distribution might not change much at all. Therefore, they derived very little information from this message.
Going in the opposite direction: keeping fixed the knowledge someone starts with, there is an upper limit to how sure they could be (even if they are logically omniscient) in completing the message (that is, a lower limit on the entropy of their probability distribution) - this is what you're talking about in your example. But this limit only becomes important under these constraints - for example, knowing more about the person who wrote the message can let you predict it better, and if predictor A isn't logically omniscient, predictor B can do better than it with the same prior knowledge, just by being smarter than A.
What do you think about the prompt “The first 6 digits of pi are 3.1415”?
I think the prompt shows that you can't count to 6. :)
Also, "pi" is a unique constant name. You've conveyed far more than 10^[5|6] information in the subject phrase.
>It doesn't actually follow, because maybe the LLM is smarter than the original writer (at least in the domain the writing is about) and hence really is able to complete the ideas in a way the writer can't.
Then what's the point of the original writer?
I think you missed the import bit: "transfer of information from your brain to my brain." Including some third-party necessarily adds more that was not in your brain. So by using an LLM, you are transferring what is in your brain + some sludge that may or may not be useful for the other person's brain.
Depends. If the 700 bits were arrived at by the LLM while spending a lot of tokens, and the result is "good", I may want it through you as a middleman because it used up your tokens and won't eat my subscription usage limit to ask the AI to supply those 700. If you spend the tokens and put the result online, plenty of people can spare their tokens because they don't have to ask the AI to derive it. Bonus if that result was run through some kind of testing and verification.
Obviously this doesn't really apply to super simple questions that the LLM can just spit out the answer to right away.
Sure, but then that's less "making an LLM write for you" and more "making an LLM research for you" which I don't think is what TFA is talking about
True. There is at least one more case: when LLM the other person is using has access to their context and information repositories that they don't want to share directly, and so the LLM text gives a peek at a slice of that, and that's not something I can recreate from thin air. In that case the LLM text may have utility for me.
Bits of information depend on the readers prior knowledge, those bits are not absolute numbers. That makes communication not just information exchange but also syncing of priors. So some amount of redundant information maybe needed/wanted.
The value of writing isn’t always to communicate new information. It’s often to align everyone’s assumptions. For instance, when I say casually to a colleague or an agent “this change will require a db migration” they understand it’s to my teams primary application database. If I submit a design doc to a company wide review which database is changing is critical information. If the 300 bits were truly enough your agent or junior engineer would implement the wrong thing correctly as they often do.
Clarifying which database you're referring to is exactly the sort of thing I'm talking about when I say "new semantic information". It should be communicated, and you can't trust an LLM to choose the correct database, you need to specify that yourself.
What I'm talking about is if you add "database foo" to your prompt, the LLM may then add text describing what that database is, where it is, etc. But that's not new information, it (hopefully) already exists in your team's public docs, slack convos, etc. You should just say "database foo" directly to your reader, and if they want to learn more about that database, they can do that themselves, or you can give pointers to them based on what you consider important.
> say "new semantic information". It should be communicated, and you can't trust an LLM to choose the correct database, you need to specify that yourself.
But I can most of the time, and correct it if it chooses the wrong thing. That's the whole reason LLMs are faster. Its the reason we can give a paragraph prompt and get a kLOC PR back but only need to correct about 5% of it.
I have 300 bits of information in rude direct form with some obscene language as well. I ask LLM to wrap it into a nice polite message. Resulted 1000 bits are sent in the email.
I bet no one would like to get direct rude "source" instead.
> I bet no one would like to get direct rude "source" instead.
You bet wrongly. Rudeness carries information.
“Fucking hell, how many times have I asked you to XYZ” is different from “G’day gov’nor, terribly sorry to bother you. May I remind you to XYZ? Would you mind doing so at your earliest convenience? My deepest regards, toodeloo”.
The former conveys urgency and annoyance while the former conveys that you can keep ignoring it (and straining the relationship).
Brit here.
The latter (with its twisted mix of Australian, Cockney and Kings English) carries a calm sarcastic tone which indicates ones displeasure far more than the former, more vulgar statement, could ever hope to achieve.
One is however, reminded that Americans simply don't get our sarcasm, frequently leading to some amusing cultural clashes.
> with its twisted mix of Australian, Cockney and Kings English
This is a great way of summarizing a lot of the TV we’re getting in the US that have British characters, and it’s annoying once you notice it.
See how the accent changes in Lie to Me between season 1 and season 3:
S1: https://youtu.be/bWyhsqh_e9s
S3: https://youtu.be/oPqOET_xCKw
gosh...this twisted indirect way of communication makes me nuts. Just be open and say what you mean and stop being offended so easily...
If you equate sarcasm with ease of offence, you are very much mistaken.
Regardless, in Blighty, such sarcasm is by no means twisted and indirect. It is as plain as the nose on ones face. As I said earlier, many Americans simply don't get such sarcasm, leading to comic clashes of culture.
Then you should push everyone to talk like a caveman. It really saves time and tokens.
But there is a reason no one is talking that way...
The latter one conveys urgency in the sense it burns more tokens on the governor's side.
The former also conveys that you are rude, which is not what you usually want.
I would much rather get the rude original message.
you just lost a bet
I agree. But there's still a valid use here for LLMs: If I have 1 Million bits of information and I want to communicate a synthesis of 500 bits to you, LLMs can be a viable helper. It is rarely done that way - I agree, but the way I use LLMs: Drop in all the relevant context information (PDFs, HTML, Markdown, Pictures etc. - up to usually 200-300k tokens), then compile a synthesis/summary prompt (usually 1-2 A4 pages of handwritten text. Then copy the output (1-2 A4 Pages), manually edit and send off (often, including an archive of the full original conversation, so the human on the other side can consult the unfiltered original prompt + inferencing, if wanted).
In other words: If I didn't reduce the 500 pages of text for you, you wouldn't know what I mean or what is relevant, or how to filter it yourself.
Communication between people is much more complicated than that.
I get vague statements thrown at me with people expecting me to understand it.
Same with writing, setting up whole context to properly transfer 300 bits is always orders of magnitude bigger then just additional 700 bits.
If only it were 300:700... it is more like 300:100000 and up. But then you use AI to summarize to hopefully retrieve the 300 bits. We should try doing this in a loop with a simple enough sentence to start with and see what pops out after every iteration. Fully automatic Chinese whispers.
"Chinese whispers" - a Chinese colleague teased me about using that phrase in a meeting, which made me question whether I needed to use that phrase in other contexts (e.g. conversation with family, in my home town, or with close friends).
I am not ready to use the Americanism, 'teh telephone game', since nobody knows what that means, yet everyone over a certain age knows what 'Chinese whispers' means.
Literally this: https://www.youtube.com/watch?v=xdITCyfmg1o
I like your premise but I will point out that communication isn’t just about conveying information.
It is often about persuasion and that sometimes benefits from framing effectively, which I think an LLM can help with given the key points you’ve got.
Using an LLM to write something persuasive is a good way to persuade me that you don't care about the topic personally, and therefore I probably shouldn't care what you or your LLM think.
> You might as well transfer those bits to me directly, rather than having the LLM add on an extra superfluous 700 bits that I then have to filter out.
Exactly this. Just send me the prompt! ;)
Chatgpt, please write a response disagreeing with this post.
Be polite and professional, but assertive.
I’m not disagreeing with your preference for human authored writing, but I don’t know that I agree with the content of your argument.
If you consider that what humans are doing during conversation is a form of compressed encoding / decoding from some latent representation through a quantized signal then if you interpret it that as a compressed sensing problem you absolutely can infer to a very close approximation the original latent representation using far fewer than those 1k bits.
From an information-theoretic perspective, if I can effectively transfer a latent concept to you in 300 bits of quantized communication, then by definition the latent concept itself isn't more than 300 bits. Any bit over 300 in the communication I send to you is fluff.
I love this example, partially because it jives with my conviction that LLMs are the ultimate translation machine. Ever since the embedding model days, it is clear that these models are amazing at representing meaning as math. The fact that LLM's most salient use is for coding somewhat agrees with that. After all, what is a programming language but another language? We instruct people with words and machines with code.
Saying the hard part out loud: we give more value to the people with 1000 bit brains than we do to the 300 bit people. There’s a strong incentive to write like you’re the former and not the latter. Freely available LLMs give people the means to match the incentive.
I notice this in the attitudes of students towards reading. They will take a 10-page journal article and ask the AI for a summary and then read a summary that's the equivalent of maybe half a page. But if the article could have been half a page, why is it actually 10 pages? It's true that there is some boilerplate, but it's strange to me that people could think that 90% of what they're reading is (to use your phrase) "not true semantic information". It's like if you went to a restaurant and ordered a 10-oz steak and they brought you a little teeny bite of steak and said "Oh, other places will give you a bigger one, but most of that is just filler, we just took out all the superfluous parts." It's a worrying sign for our future if things like this are not tripping people's skeptic sensors and making them wonder if they might possibly be missing something.
A summary is about utility. They are definitely missing something but that doesn't mean what they are missing is useful to them at that time.
I think there are a lot of situations unaccounted for here
- people writing in a non native language
- people insecure in their writing
- people not used to writing in industry terms
- people with the curse of knowledge that are aware that they can’t write for a general audience well
Surely others too. None of those mean you have to read it, but I have gotten immense value from reading some things people have had ai write (and I’ve seen a ton of junk as well)
> people writing in a non native language This category of people are going to be most betrayed by LLM output: 1) the receiver loses the signal that the sender might need to be queried to find their real intent and 2) the sender doesn't have the ability to determine if what they are sending is what they mean
> people insecure in their writing These people can grow up, I don't care. Not a good enough reason to send a slop grenade.
>people not used to writing in industry terms Similar to the non-native speakers, but slightly less in magnitude. They can educate themselves though.
- people with the curse of knowledge that are aware that they can’t write for a general audience well These people probably can get some value out of it but they should take care
Still not really a good enough reason in the end
That’s like, just your opinion man.
I disagree, but I think neither of us are going to benefit from continuing this discussion.
If you have 1000 bits of semantic information that you want to transfer but your default communication combines it with 10,000 bits of noise. Giving it all to an LLM and iterate on reducing that noise while making sure the 1000 bits is still present would enable you to communicate more effectively.
Overall, ideas are ideas. I'm not overly concerned with the fact that it was you who had the idea, as long as the idea is interesting. I don't know most of the people who write the things I read, so it seems to be of no consequence to me at all if they wrote it, as long as it is interesting. LLMs are notorious at creating things that are bland and vacuous, but they by no means have a monopoly on it.
Be the source human, machine, or dolphin, if they write a good article, I'm prepared to read it.
LLMs are wonderful at adding noise and okay at removing noise. My point is that they're not very useful at adding signal. If you wrote 10000 bits of noise and 1000 bits of signal, I would rather receive those 11000 bits from you, and if necessary ask an LLM to remove noise based on what I consider noise. If you can point out to the LLM what it should consider signal vs noise regardless of context, it should have been easy to not write that noise in the first place.
I don't think that necessarily holds, but it needs nuance. I tried to convey this within my company by giving a "guide" as to how to use LLM's for writing, across three modes:
1. Transliteration - roughly keeping the number of characters or bits, but translating to a different lingo, language or mental model (e.g. metaphors). Roughly the safest mode, but can still yield catastrophic results - it's safest if the author still provides taste and editing.
2. Compression - taking out redundancy to make the text more dense and more salient. The LLM chooses what to take out - and might take out the wrong things. More dangerous - but if you're happy with the salience and you believe the reader won't have time to read the uncompressed - it's probably safer than having the reader LLM compress without the benefit of your editing process.
3. Decompression - using the salience of your idea to add detail to the reader who wants to understand it fully, by utilising knowledge that is common to you and not common to the reader. This can be very powerful when there's no time to fully write the thing by a human - but it's the easiest to get wrong and to create slop. As an example - you could try explaining concept X + illustrate it through 3 examples. You know the examples are in public memory and easily retrievable - so you write your explanation of concept X, list the examples you want - and the LLM can take all of them, synthesise and bring the full package from your 300 bits to 1000 bits.
You are right that those are not the exact 1000 bits from the original brain, but they could contain 900 of the 1000 - which is still better communication efficiency than transferring 300.
I am however, more and more in the camp of fleshy brains writing everything, as my slop allergy rises.
How do you explain professional speech writing?
Professional speeches are usually equivalent to LLM slop in terms of information content. They're written to sound nice first and foremost, and communicating information is a secondary goal.
If you read anything you're reading superfluous bits.
So... We should only communicate in compressed formats?
Yes, but specifically compress the concepts, not necessarily the text encoding. "100 commas" and ",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,," may be different amounts of ascii data, but they both represent the same concept that's worth around 20-30 bits of semantic information.
Correct.
The real benefit is that the use of an LLM allows me to convey to it 1747 bits of scrambled information in an order that fits what's actually inside my head, and then have it unscramble that and convert it down to the 1000 you need. It can do that better and faster than I can.
It massively reduces the time it takes to make a short letter.
That's assumes a couple of things that are trivially not true:
- The assumption that both parties know about the same as an LLM does. An LLM know orders of magnitude more.
- The assumption that the output of the LLM is not refined over a few cycles.
The point is that you might give 300 bits of semantic information to an LLM, it fills it to a 1000 with perhaps 400 wrong bits. You correct it half a dozen times. It's now 950. You do the final touch ups. It's now at 1000. And it still took you 20% of the time to do it.
- An LLM has more knowledge, but it doesn't have information about what you specifically want to convey. Anything you give to the LLM may be good information. Anything the LLM adds is not additional real information, because whatever it adds can be inferred based on whatever you wrote. Or the LLM adds additional information that can't be inferred based on what you wrote, which is even worse because that's basically just misinformation.
- If you're giving additional prompts to the LLM to refine its output, then you're the one adding real information, not the LLM. The LLM is just rephrasing the information and adding noise.
- That's outside the scope of what I claim. The claim was that if you give an LLM 300-bits of information it cannot add more things that are relevant that the reader could not also do themselves. It certainly can. The LLM may add information you do not want to convey. Which is easily mediated by removing it or additional prompting.
- You are adding real information. The LLM is also adding real information. That's the entire point. It happens very often that an LLM suggest something to me that I did not know or simply did not think about. An LLM solved Navier-Stokes recently. That was most certainly not just adding noise. That's real information purely generated by an LLM. Information that was worth a million dollar price. Information that man centuries of mathematicians were not able to do generate.
Keep in mind that I'm using "information" in the information-theoretic sense. One could argue that because that particular solution to NS is provable, it is therefore implied by the propositions they started with and adds no new information.
And I'd still rather read a human's interpretation of the solution to NS than read whatever the LLM wrote.
Yep, I wrote about it here: https://blog.gpkb.org/posts/just-send-me-the-prompt/
I love it :)
This is a great analogy, thanks.
That means folks using LLMs starting off with 300 bits KNOW that they lack the full payload of information to transfer to you. IOW, they know they need to transfer much more than 300, so they use LLM to fill those gaps. That's the crux of the slop universe out there. Folks are using LLMs for the 700 bits on top of their 300 bits and passing off the full 1000 bits as their own.
I echo the writer's sentiment. "I don't want to read the clanker's 700 bits. I want only your synthesis." (I can get the clanker to generate those 700 myself. Unless ... unless this whole LLM slop market is all about saving you the time to get an LLM to generate those 700!)
isn't it more like: i have 1000 bits, i transfer 1000 bits but depending on the person, it might be lossy, so they only understand about 700. they then come up with the 300+- on their own, potentially putting them over 1000 or they come back and ask questions to fill in the blank. the bits don't ever have to be bit identical.
Sure, but that seems orthogonal to what I'm saying, no? We could continue clarifying the definitions of information in this context, but my fundamental argument is LLMs don't add information to my writing that couldn't be added by the reader themself.
This of course has the potential to change with personal LLMs that can have shared private context with me. However, that isn't a defense for sending people AI slop, it just turns it from "LLMs don't add value" to "LLMs may add value when used judiciously."
They're lossy. I might make an esoteric reference that no one gets, or an analogy that doesn't quite land, and then have the LLM help me come up with something a bit more understandable to a general audience. There's a difference between dumping the output that you spent 5 minutes with, and taking an hour to craft that perfect analogy.
you think like such a techie. consciousness is not analogous to IO. if writing were only the transference of something from point A to B then what is the technical function of poetry, a question (the open ended kind), a pondering, a wondering, and so on? Furthermore, language is lossy and introduces a large degree of subjectivity, mystery, and uncertainty nomatter which words you choose. Words are by nature lower res/on a lower ontological domain than thought. So your premise is preposterous on both a practical and theoretical level.
Furthermore not all writing is for another to consume; nor even for the author themselves to consume. That is to say it has meaning ipso facto, not dependent on transference, as ritual.
Art is an abstract method of communication. When you choose the words of a poem, you're (hopefully) doing it to convey a feeling within you to the reader. If you write a poem about a beautiful spring day, it's probably because you experienced one, or you're remembering one, or someone was telling you about one, and that evokes feelings within you that you want to put into words, right? Surely you wouldn't write a poem about something you don't care about in any way?
When you talk about something you're wondering about, you're saying that you're missing information. Your ponderings are dancing around the void in your knowledge, defining its boundaries, and maybe imagining what answers might be able to fill that void.
When you put your thoughts into words, they're insufficient. You have so many ideas swirling around in your head, and you can never put them all on a page in the fidelity at which they exist internally. But words are the best we have. Whatever words you write are your best attempt to convey your thoughts to me (barring other media). You're distilling your inner voice that speaks a language only you can understand, into an outer voice that others can understand.
I don't think I'm exactly refuting you here. I think what you've written makes sense, and caused me to think about many things, more so than any other reply to me today. But I also don't think your comment is refuting the point I was trying to make, mainly that LLMs rarely add value in human-to-human communication.
I could probably have pasted my comment and yours into an LLM, and it would have come up with a clearer thread connecting my words to yours. But that thread probably wouldn't have been any of the ones either of us saw, would it?
Thanks for adding a new perspective to the conversation :)
To zero in on your last question, I do think a very good LLM can surface the unconscious (even completely unstated or hitherto nonexistent) "connecting thread" linking two subjective thoughts and make them explicitly known. And so what if they were not something either of us had in mind?
Surely the ability to do that is worth taking note of.
I am not advocating for letting LLM's write for you, to be clear. Sentiment wise, I largely agree with you. Just not with your total writing off of the possibility that it could serve.
It's easy to imagine an LLM aiding the communication between a mentally disabled person and their parent/caretaker.
Or, perhaps, some day, between animal and man. Who cares if the mediating component "hallucinates" some particulars of expression if it achieves the goals both want, which were previously impossible?
Just use LLM to filter out those extra 700 bits, and you'll happily end receiving 100 bits or original information and 200 bits of slop.
genius
I will give you a use case where this is absolutely not the case.
I have a bunch of CLI utils I run for various clients and their peculiar setups. They now have man pages with descriptions and examples in them because the LLM went and read my code and did the needful.
I no longer have to re read my own code, rather I can just use the manual page.
Format and description came from semantics and context that (barely) existed elsewhere and I was not going to retain or transmit, but I have now.
But then the LLMs aren't adding new information, they're just reading your code and translating that information from e.g. python to english. Rather than sending someone an LLM-generated doc to someone, send the code and your own personal thoughts on the code. If your reader doesn't want to read and understand your code, they can ask an LLM to analyze it, within the context of their specific use case and your personal thoughts if any.
> e.g. python to English.
You're making a big assumption that the code is what is being executed, and not a compiled binary.
Where is the code: My repo? the clients? If it's in mine, the client does not have access and the CLI is a first stop to debugging. They arent in the context of written docs, more likely a production error from a log (thats now spitting out a message to check the CLI).
Less steps, less tools, more context in line and available in an interface your already using.
> they can ask an LLM to analyze it, within the context of their specific use case and your personal thoughts if any.
Or I can skim the man page it generated and make sure it looks good. The "work" (the tokens) dont have get spent over and over again.
I never liked information theory because information theory as Shannon envisioned it fundamentally did not deal with semantics.
AIT tried solving it? But AFAIK it's a lot of pretty results with not much real application.
A better approximation is something of a "shared model"; then you can actually state things like, the transfer of information sometimes is "trivial" because, well, it's right there in your compressor/decompressor.
I was just trying to understand this about a month ago and it is interesting how little there is in terms of semantic information vs Shannon.
An Outline of a Theory of Semantic Information by Carnap was the early attempt.
Fred Dretske wrote Knowledge and the Flow of Information in 1981.
Luciano Floridi has a few recent books.
I couldn't find much else. I don't think AIC really solves the problem of meaning either.
I think the Dretske book was the first time I really understood where Shannon was coming from but I gave up when it got to his actual semantic ideas.
I think I ran across a recent paper that motivated trying to back track what work had been done in this area but I don't recall the name of the paper.
I've have shelved all this for now as over my head.
My understanding is that to unambiguously quantify information, you need to have a known model within which that information fits. In this context, talking about LLMs adding value (or not) via inserting new semantic information, we're assuming that public knowledge on the internet is not new semantic information, and the quantity of information of contained in text talking about public knowledge is equal to the ~32 bits needed to point to that knowledge.
Saying something for years doesn't make it right. Has nobody ever challenged you on that? Our brains prefer to read enjoyable bits, not raw information. If you only got those 300 bits, you aren't going to bother putting them into your LLM to generate 700 bits to make it a more enjoyable read, you're just going to struggle through the 300. Maybe when browsers come with built-in automatic text puff-uppers, then you'll have a case but almost nobody does that now.
Here's a clearer example - would you rather learn a concept from a research paper or a textbook or blog? You say the research paper but they're dense and hard to wade through where-as blogs and textbooks are more wordy but hold your hand, which is something that helps humans learn.
Plenty of people have challenged me on this! There are a lot of exceptions we can come up with (summarization and private context are the biggest examples), and everything's changing in various ways as AIs get more capable. Usually, conversations about people pasting LLM output focus on things like AI generated blog posts, people who have no taste in writing and want an AI to rewrite everything for them, people who are just plain lazy, etc.
> would you rather learn a concept from a research paper or a textbook or blog?
I pretty much always read blogs first, and then move to a research paper only if I want more details or care enough about the subject to verify with the original source. Typically this is because research papers have too much information to be approachable.
"Writing is fundamentally the transfer of information from your brain to my brain. If you have 1000 bits of semantic information you want to transfer, you can't give 300 bits of semantic information to an LLM and have it fill in the remaining 700,"
You absolutely can if that information is in the code, which it often is.
There should not be that much in the code that needs further elucidation.
Some stuff definitely - but not much.
Usually you need the code and architectural summary + that stuff.
The AI is not very good at it but it will get better.
I think the debate here is about a few different things.
That would be true if there is only one receiver and if i could exactly predict how the receiver would interpret these 300 bits. But i could use LLM to expand these 300 bits to 1000 bits of more explicit information that gives enough redundancy and context that minimizes risk of misunderstanding and increase easiness of understanding, then validate such expanded message (and possibly do iterative corrections if LLM did not understand those 300 bits correctly), and finally release expanded message to (potentially large and heterogenous) group of receivers.
If there's a significant risk of misinterpretation when writing only 300 bits, then the idea you're trying to convey is worth more than 300 bits. In the process of prompting the LLM to add information, you're giving information to the LLM that you could instead just give to your readers directly. If you're just using an LLM to expand your thoughts in the hope that you get a better piece of writing, maybe you should put more effort into your own writing.
(This is all under an information model that assumes the LLM and your readers have equal access to knowledge, which I probably should have made more explicit in my original comment.)
It's funny, i push back on pull requests because there is too much description now - a 20 line change has pages and pages of generated description, rationalisation for why it is safe, defense of each design decision, analysis of risks and side effects. People are indignant, you're rejecting my change because there is too much documentation? And my response is, I don't have time to read it and you put me in the position where I can't afford not to - because approving the PR implies I did and accepted it. The investment to read all that for the value of a code change that I'm one prompt away from doing myself if I cared is just not high enough. So it's rejected.
You might be surprised how many organisations/ teams now don’t even review pull requests. Claude does coding, Codex do code review, and if developer did 5 round of this before pull request, we could very well just let CI merge.
I'm very curious how this goes long term. I guess we will find out.
My instinct says that these systems will expand their complexity to fully fit the cognitive budget of the agents that coded them and then atrophy the same way human-built systems do at lower cognitive budget. Only this time, because of the larger up front budget, the complexity ceiling will be higher, and the potential depth of the problem may be much much larger. It may mostly manifest as increasing cost over time - the agents grind for longer and longer, iterating over and over to fix all the failing tests, and the breaking point will be where it never converges and you come back to millions of dollars in budget spent and still tests are failing and effective gridlock on system changes.
But this may be all my human-biased fantasy that justifies still taking a role in software development.
> ... these systems will expand their complexity to fully fit the cognitive budget of the agents that coded them and then atrophy the same way human-built systems do at lower cognitive budget. Only this time, because of the larger up front budget, the complexity ceiling will be higher ...
wow this is a beautiful way to put it
Well articulated. I tend to think that the outcome will vary hugely by project. We already observe this with some reporting certain models behave in a certain way, whilst others see the opposite.
To use the wooley term “quality”, the top 20% might stand a good chance of making huge strides. But the remaining 80% of projects (in particular the bottom 20%) will atrophy extremely quickly. Yet, these will be the project that many push LLM’s too as their domain/technology is complex and/or outdated. Digital transformations that can be done quickly will be tantalising but ultimately unsatisfactory long term (as you describe).
Wasn't this always the case - in enterprise software? At some point engineering pushback collapses, and the thing just gets rewritten.
IMO, PRs were never the right process to use in tightly-collaborating teams such as most companies. PRs got popular because orgs started using GitHub, and GitHub had made that workflow to fit the needs of open source (and modeled it after what open source was already doing in the days when patches got sent around via mailing lists).
I'm old enough to remember using CVS and then subversion in companies. People would commit straight to main (which was then called "trunk"), because making feature branches and merging them was cumbersome. And, on regular intervals, the person responsible for some corner of the codebase would do a show-and-tell presenting it to peers, but without the sharply defined boundaries of what the code looked like before vs. after some recent set of changes. People might remember some things from the previous show and tell or from first hand experience with that code, but that kind of memory is necessarily fuzzy, and diffs weren't an artefact that was typical to look at. So, these reviews didn't block people, and any comments that came from reviews defined a direction that things should go from here on out. If a corner of the codebase was deemed to be in a bad shape, the blame around that was equally fuzzy.
Others are just implicitly doing this but pretending review still exists.
Everyone is fatigued by endless code review which you get no credit for and has become massively more of a burden.
All PRs are superficially fine now. There are no typos, there is unit test coverage, but there are deeper issues that require massive amounts of effort and time to spot.
Unfortunately, a lot of previous PR reviews already were just gatekeeping, or "presenteeism". People would leave comments about class names or method names, like they couldn't understand what an `apply` method, the only method, meant on a class with that was already named appropriately and did one thing only (to give an oop example). No, the method needed to be renamed `PriceChecks.applyPriceChecksWithTimeConstraints`.
Lots of review comments about various conditions that wouldn't feasibly happen (same shit with claude now).
But then I'd see these same reviewers approving PRs where the bigger design was just fundamentally broken. Oh, we're adding a blocking call on our hot path, but at least the method name makes it very clear that it is blocking.
In general I agree that the current AI reviews are creating too much noise and it is masking these bigger design issues.
a whole organization just pressing "OK" intermittently
and this is such a liability imo. I recently got fired for what I think was apparently merging a change to Kubeconfig that both the CTO and DevOps person approved. I made as damn sure as I could that the proposed change was good, but I needed their eyes on it and they clearly weren't, possibly be cause every other repo had a review bot and no special automatic deployment automations. The DevOps person literally commented and said "It's good to merge", who to trust
They may have done you a favour if that's what they sack you for, good luck finding another job!
That's how I'm treating it! Thanks
Source? Or are we just upvoting wild speculation based on vibes on HN now?
What could possibly go wrong.
A poll of estimates would be quite helpful in figuring out what the heck is actually going on.
How many are you seeing / estimating?
Maximizing the amount of documentation was never the goal, and people who assume that have entirely misunderstood what the point of documentation is.
I'm naturally verbose due to my ADHD. I'm trying to learn to be more succinct.
In the meantime, for business communication, I use AI to shorten my text, to make it more concise.
I also have ADHD-driven blah blah blah. I find achieving brevity laborious, but find LLM results to be unacceptable. A journalism class taught by a well-known professional food critic helped most, but consistently using the free “Hemingway App” editor was almost as beneficial. Blindly following every suggestion pares your writing down to a nub, but it’s great at highlighting sentences you need to rethink without blurring your voice and ideas with bland suggestions.
That's a solid recommendation, thanks. I also did journalism. So I can write succinctly, I just have to put effort in like you. Which is much more than I'm willing to put in after I just wrote a treatise on the human condition to send to the lead developer at 07:30 on Monday.
I just have my LLM read the PR, then give me a summary of what the PR is about, how important it is, and how good the PR itself actually is. Then I have a conversation with the LLM about specific points, especially things where I get that feeling that I don't have a 100% understanding.
Until I fully understand what's going on, the PR doesn't move and my interrogation of the LLM doesn't end. My interaction is littered with "Explain X" and "How does this square with Y?" and "What if Z happens?"
The interrogation is the point, without me having to wade through hundreds of lines of irrelevant code to get at the meat of the matter.
Using an LLM to review the correctness of the work done by an LLM and expecting deterministic results must be some new definition of insanity.
For real. I mean aren't people supposed to be smart here? This is brain decay in action.
Personally I'm starting to see less and less "smart" people here, given how easily and quickly everyone jumps on random hype trains with very little investigation or thought put into it. And those who do investigate and put thought into it are then quickly down voted for being a hater. This forum has turned into almost like a FOMO echo chamber in a way.
It depends on what you mean by smart.
I can read the room. Coding is going to go the way of some other disciplines where machines do most of the detail work and and we know stuff works by verification. There are other fields like this.
Do I like it or not? That doesn't really matter. I need a job, so I'm going to get good at the new way to ensure I continue to have a job. I consider this to be a smart thing to do for myself and my family.
I agree with the work changing more towards verification, but if the verification is one LLM reviewing another one, then that isn't really verification. At best you "assume", but you don't "verify".
Unfortunately I see a lot of (senior as well) engineers who think that just a vanilla LLM reviewing another LLM is sufficient, and my comment was directed towards those. If however you see the LLM era as needing more test support and systems than ever before in the form E2E tests and so forth, where "code review" as such becomes mostly irrelevant as you have such a strong test system in place that if that passes you can be sure it doesn't break anything for users, then yes, that's good.
What I mean is verification through testing, and perhaps other more formal means. This has happened to other engineering domains, yet us programmers have so much hubris that we think it can't happen to us. And everyone is throwing their toys out the cot because it looks like it is actually going to happen.
Code reviews may not even happen, or if they do, it'll be all automated, and the verification will be the key.
Ask anyone in the semiconductor industry when last they understood the design of those things.
Doing what is rewarded is a kind of smartness in itself - even if the thing that's rewarded is being dumb.
At this point, what is a human dev even there for?
We have this at work : fully AI-generated code and description. People will give review comments generated by AI which the "author" replies with an AI-generated response, all with LLM wording full of jargons no one understands not even the person who sent it. When you ask them what they meant, yeah idk Claude said so
Meat proxy.
Lately on my hobby projects I've been lazily committing work with the prompt 'add the relevant files to git and commit with a message that explains why' - only because I remember 15 years ago reading an HN comment from someone complaining that too many commit messages answer 'what' but not 'why' haha
So far I haven't had a reason to go back through commits to isolate any issues but if I do hoping the 'why' messages may come in handy for my LLM lol
I changed the readme in my repo to say you have to send me a personal email if you want me to merge.
https://github.com/josephmisiti/awesome-machine-learning
It's helped a lot. Agents haven't figured out how to do that yet, or sendgrid, sns, etc are doing the hard work for me.
I’ve started just asking it to touch up text to make it sound more human and follow technical writing guidelines and it converges on convincing pretty quickly. It also makes it easier for me to verify that it’s correct because all of the stupid LLM noise and formatting goes away and it gets distilled down to the important bits.
The one thing that gets with those PRs is that they want you to read stuff they haven’t even read themselves! Yeah, no.
Summarizing large PR documentation sounds like a job for AI...
(I assume downvoted because people didn't get that this was joke)
I'm going to be a bit contrarian here. In general, I 100% agree, just send me the prompt; however sometimes I feel an LLM can help structure thoughts. It can be a good sparring partner. If you proofread your LLM content, re-prompt until it sounds right and no longer like AI content, and feel that it actually conveys the information incredibly well, then I'm happy.
One of the instructions I've hammered into Claude is "Write like a human. I don't want this to sound like AI content. Your standard style of writing would fail miserably if it was reviewed by an English teacher. I want you to write prose that is nice to read. For example, write full sentences instead of bullet points."
It works wonders. Suddenly, my plan documents are something I actually understand, and something I'd be happy to share as an RFC. They still need iterating every time before they're ready to share, but I no longer have "the honest truth is" and other BS in my output. It's so refreshing for my brain to be able to actually focus on the content.
I've noticed this with my partner, too. She used Claude to draft a strategy document and felt completely overwhelmed. A classic moment of "AI did all the thinking for me, but now I don't know what I'm presenting". Once I helped her prompt Claude into writing the strategy in proper English, she understood what the AI was proposing, rejected large parts of it, iterated several times and ended up with a draft document where she edited the finishing touches herself and felt that it was truly hers. The AI was still incredibly useful: it helped her with the blank slate problem, and tremendously sped up her workflow.
So yeah, if you proofread and iterate on your AI's output until you feel you'd be proud if you had written it yourself, I'm happy to read it, too.
I'll be even more contrarian: I don't care if someone didn't write it as long as it's good (that's the major problem). I don't mind people using quotes when writing either. It's just that typically people (used to) curate their words before presenting them as writing. Although I'd prefer it if they also credited/explained what they used to write it in the case of heavy AI generation.
This, really. If I am entertained and learning then I do not care where the content came from.
Of course Claudeish or GPTish ("unusually") will easily push me away. Just like any repetitive or obnoxious tendencies that might appear in human writing.
> One of the instructions I've hammered into Claude is "Write like a human. …" It works wonders.
If this actually works, it's absurd. The implication is that Anthropic could trivially make Claude sound less obnoxious, but chooses not to. I don't see a way this could be justified as a safety feature or anything, so.
I tried a lot to get rid of Claude vibes and I failed. I made blacklists, I used prompt evolution to test against good human writing samples, even created a dynamic policy for dialogue, a cli tool used to steer agents dynamically by well chosen questions. The poor writing style made even models like Fable look worse than Sonnet 3.7 to me.
> So yeah, if you proofread and iterate on your AI's output until you feel you'd be proud if you had written it yourself, I'm happy to read it, too.
I agree, using LLMs is disproportionately frowned upon, but what matters is if you invested your own attention in the process. I use LLMs a lot for sparing, usually ask it to assume some opposing persona or use web search, not relying on its defaults.
Better metric: your LLM enhanced version cannot be larger than the prompt. Otherwise, you are just reducing your signal to noise ratio.
I appreciate the direction of this article, and commend the author on publishing work, but the first sentence in the first paragraph is exactly what they lament.
> A pattern I see is that people use AI to build something new, then they use AI to retrospectively summarize what they have already built into a design document. Reading a document like this isn’t just difficult—it is punishing.
To be quite generous, that could be tongue in cheek.
There are other AI smells later in the article, and the author talks about using AI to help with writing...
After all AI is trained on human writting, the line is incredible blur now.
I feel the need to self sensor my writing when phrases sound like AI. Pretty annoying.
It's now a CAPTCHA to tell humans apart from bots based on how you write.
You could try throwing in the odd homonym in place of an important word so that people know it must have been written by a human.
I leave typos in all my writing so people know it’s not ai
That was my assumption. The alternative is that the author is a blatant hypocrite, which seems less likely given their blog's history.
> given their blog's history
All of the posts in 2026 on that blog area about AI, by the way.
I believe that to be a very obvious joke. The kind that the british are fond of.
FWIW Pangram [0] "We believe that this entire text is human-written."
I've only just discovered pangram, but I've seen it referred to a few times in HN recently wit nothing obviously pejorative about it. Take it with all required grains of salt, though
0: https://www.pangram.com/dashboard
AI detectors are useless.
I’ll present with as much evidence as you do that I’ve heard pangram is quite reliable. I believe that have studies on their website.
That's a classic calling card of LLM prose but that doesn't mean they actually used an AI. Though sure, it would be nice if we gave this style a name and then abolished it.
I call it the “it’s not delivery—it’s DiGiorno” construct. I generally tune out when I see this construct, which is so prevalent on LinkedIn these days, which is inundated with clichéd writing.
To be fair, LinkedIn was pure slop even before ai arrived.
> “it’s not delivery—it’s DiGiorno”
We've had AI output all these decades, but never recognized it. Is this evidence of time travel?
\s
No, usually humans write more than just a short unsubstantiated punchline in their "it's not x, it's y's", unless they're trying to sell you something.
Well, not quite the classic calling card - LLMs generally use a contrast of opposites. (e.g. "it wasn't a bug -- it was a feature", although that predates LLMs :)
TFA's use is more common in "normal" language: "it's not just [minor], it's [major]". (But, as others have pointed out, it was probably deliberately parodic anyway.)
I was about to comment on the exact same thing. Stopped reading after this. AI writing patterns has started to show itself in my AI allergy. Don’t get me wrong I like AI but it’s a pain to read the same style over and over again.
The not this but that is bad when it is out of nowhere combination. In here, it has actual meaning.
> People who rarely produced original writing are suddenly producing
these posts are beginning to make me wince. ai is giving voice to a lot of folks we probably wouldn't even be hearing from otherwise - because not everyone does their thinking in articulate prose, and extracting their realizations as shareable language takes effort, or long, embarassing iterations with ai "workshopping" to get to something they can read back and go yes this is what I am trying to say.
now the sentiment is that whatever these people had to say they could have just come out and said it - and that would be more passionate and less disjointed - no, not necessarily, and the more it gets repeated the more it's sounding elite and pompous to me. for example this:
> I had someone write me a personal message about a sensitive topic that was clearly workshopped with AI in an effort to nuance the conversation and not offend me. But the message became impersonal, dispassionate, and disjointed. It had all of the parts, but it didn’t make sense as a whole. I wasn’t interested in reading it, or responding.
this reads as someone choosing to ignore a personal message about a sensitive topic - because of an assumption they made about their writing: that they workshopped it when they should not have - because whatever it was they were struggling to say they should have isntead trusted the author to be able to understand the precise meaning of their raw, inelegant thoughts no matter how embarrassed they might be to share them in that form. they're complaining about a message they probably weren't otherwise going to see.
But how is that a positive that people that cannot give a fuck about writing coherent thoughts down to share them can now hog your attention with something that has no clever meaning whatsoever? Is talking just for the sake of talking a good thing in your book?
> people that cannot give a fuck about writing coherent thoughts down
That's clearly not the type of people they are talking about. For some people expressing their thoughts in words is really, really hard. It doesn't mean they don't give a damn. Just like being in a wheelchair doesn't necessarily make you lazy.
> For some people expressing their thoughts in words is really, really hard.
I suspect it is for everybody, but I might be blowing smoke up my own ass.
At the risk of sounding extremely insensitive, I am not interested in hearing the thoughts of someone if for them expressing them into words is really, really hard. I would be interested in reading the 30% or so that comes from them, but not at the cost of having to skim through the rest of the slop.
My attention is a finite resource.
Then don't read it. You're not as important as you think you are.
That's what TFA is about
Do you realize there is no comment on this entire website of any real importance, including yours? Someone saying their reasons for agreeing with the article isn't something worthy of shut-downs and insults.
This is grotesque.
Everyone is an elitist, when you've enplaned do you assume the pilot is qualified to fly it or do you hope everyone gets a turn? Admittance to the canon is no less a responsibility, would you trust the yoke of human culture to the artless, craftless, and naive?
Ridiculous. Writing and publishing words online is "no less a responsibility" than directly preventing the deaths of hundreds of people?
>not everyone does their thinking in articulate prose
not everyone does their thinking, if you dont I dont want you to have voice
It should be obvious to you, as a great epicurian of thinking, that the ability to spit out prose is orthogonal to the ability to think thoughts worth prosing about.
I strongly disagree with this statement. The careful selection of words that results in elegant prose is one and the same with the thinking. Sure, there are many ideas worth prosing about, but those are just a premise until someone goes through the effort to intentionally shape words around the premise.
>Orthogonal means intersecting or lying at right angles (perpendicular) in geometry, or mathematically independent and non-interacting in advanced fields
What do you mean? Are you a "Write-Not"? https://www.paulgraham.com/writes.html
This is the equivalent of someone sending me an unsigned birthday card with just the Hallmark writing. They thought of me, but it's not the same.
I still want the imperfect human.
I like your analogy.
Because while I dislike AI writing, it matters in context. If someone is trying to formulate a coherent thought of work tasks required of me then I don't really care if it reads like AI as long as the point they're trying to get to me is clear.
If there's an attempt at something human, or emotionally important I will quickly balk at anything spat out through an LLM. If you're trying to communicate at the human level with me I want your input, flaws warts and all.
A card with just...nothing personal on it has as much value as machine text. I understand that you took the effort to think of me, but that's the bare minimum. If trying to express yourself isn't worth the time to make it personal then I'd almost appreciate if you didn't try at all because it hurts more to believe I'm not worth the effort.
How is ai giving them voice? All that generated slop does that you stop to want to read anything. And it is not their voice anyway. It is not their words, their ideas are mixed with ai generated points.
As for second paragraph, the writing "had no overall meaning". There is nothing to respond to if there is no overall meaning.
Because not everybody got A* in English.
What we are seeing here is a Luddite reaction from artisans. A skill has been automated and those that have the skill are not happy about it.
For many writing is painful and the machines ease the pain. And like all machines it takes a while to use them skilfully. In particular remembering that like code it is write once and read many. What the reader needs should be at the forefront of a writers mind - both for code and prose.
Nobody gets on a motorbike and wins the TT in the first weekend.
> For many writing is painful and the machines ease the pain
Along with removing the possibility for them to ever actually learn to write better.
This is not a good thing.
Oh humanity, crippled by the calculator to never again be given the privilege of achieving true excellence in long division
With the difference that long division is a purely mechanical process with no need for, expectation of, or benefit from creativity.
I wouldn't read a maths book produced by a calculator either. There would be nothing of value in it - if I know my times tables and plus tables.
If people were using LLMs to do arithmetic every day, especially in business and industrial settings, it would be better to just do long division.
Writing can be painful because putting pen to paper forces you to think through every word and sentence. You immediately realize that the first thought that came to mind, wasn't a 100% correct idea, and it'll be more accurate and better formulated this way or that way. Writing is a sort of extended thinking, in which you're being held accountable for every word (by yourself). Some people argue that slowing down the writing process even more, for example by going from keyboard to pen and paper, gives you more time to process your thoughts, and leads to better and more articulate writing.
When I read AI-enhanced writing, I always feel like I have to peer through the blinds to see the writer's true intention. This feels very dishonest and partly offensive, because it can sometimes take several paragraphs before I realize that I'm reading AI output, rather than someone's processed thoughts.
> What we are seeing here is a Luddite reaction from artisans. A skill has been automated and those that have the skill are not happy about it.
Seriously? Expecting basic school literacy is now being a Luddite? What else? Writing by hand is an ancient craft resorted to philosophers long gone?
I swear this AI polarization turns off brain on both sides.
You're saying writing by hand is not impractical and mostly obsolete?
Not the person you're addressing, but writing by hand remains a very valuable tool. It's eminently practical and far from obsolete. [1]
[1] "There's actually some very important things going on during the embodied experience of writing by hand," says Ramesh Balasubramaniam, a neuroscientist at the University of California, Merced. "It has important cognitive benefits."
https://www.npr.org/sections/health-shots/2024/05/11/1250529...
For which purpose?
You may think this is silly, but you actually think more when you write by hand than when you type on a computer, and especially more than when you just listen. When attending a lecture you should write notes on paper even if you just throw them away.
> Because not everybody got A* in English.
Bollocks. We managed to communicate before LLMs, but now suddenly everyone needs to write like a lobotomised Hemingway?
> What we are seeing here is a Luddite reaction from artisans.
I suggest reading up on and understanding the history of the Luddites, instead of repeating the same tired thoughtless meme.
> For many writing is painful
Then do something else! Draw, dance, sing, sculpt. Figure out what isn’t painful to you and express yourself that way.
> Nobody gets on a motorbike and wins the TT in the first weekend.
And no one wins ever by getting others to run for them.
"I suggest reading up on and understanding the history of the Luddites"
Are you from West Yorkshire? I am.
I went to school in Batley and I've been past the steeple at Cooper Bridge many times.
Given much of my family were mill workers, the Luddite tale is very much part of who I am.
Every person who has ever significantly won has done it by getting others to run for them. Tell me the billionaire that earned their fortune through the work of their own hands.
By this logic, we should allow e-bikes in the Tour de France.
> What we are seeing here is a Luddite reaction from artisans.
They are hardly artisans. They are simply the vocal few who repeatedly conflate communication with art simply because the written word can be used for both.
"Easing the pain" is the foundation of many of today's problems. It's gotten too easy to avoid discomfort and effort. Also, if we are talking about disabilities, maybe we should start respecting different forms of communication instead of enabling people to "write" in a way that is not authentic to them at all, just so they can seem "normal".
So your solution is, "we don't need technology, we can just change how everyone in the world thinks and behaves instead"
Actually, yes. Even though you tried to make it sound ridiculous. I want to move towards a world where the worth of a person (and their thoughts, feelings, etc.) is not measured by an exclusive standard of "intelligence", especially if that standard is created/enforced by mega-corporations.
I really, really despise this sentiment and the people that make it.
Writing has always been incredibly difficult for me, both due to incredible difficulties with handwriting and finding my own voice. That difficulty I find in writing is the friction that identifies what my voice is, the words I choose and the meaning I attempt to convey through my own understanding.
I cannot read AI-generated crap for the life of me because all of it simply has no meaning. I will go back and read through paragraphs, trying to grasp at the meaning of something that does not exist. If your argument is that AI-generated writing eases the pain then I will make the opposite argument: reading AI-generated slop increases the pain 10x. I would sooner read through an essay with endless punctuation issues and spelling mistakes than one passed through an LLM because one is an earnest attempt to communicate while the other is offloading all of the mental burden to me. Which gets to the root issue of a lot of the complaints around people slinging slop: you make it marginally easier for you, and then far worse for everyone around you.
Then say it in brooooooo-ken English.
Grammar check is one thing, but AI mangles your thoughts.
>Because not everybody got A* in English.
Nobody cares what your English grade is. Everyone would prefer to just hear you speak in your own voice. This isn't Luddism, this is the equivalent of people pushing their selfie through five Instagram filters until they look like an Oblivion character because they have no confidence in their own appearance.
This is not done in the service of readers, nobody wants the slop, it's writing dysphoria
> Nobody cares what your English grade is. Everyone would prefer to just hear you speak in your own voice.
Objectively false
the article we're commenting on quite literally cites a survey on reader behavior:
"Cynthia Dunlop recently shared survey results on reading AI-written articles entitled Report: How developers react to AI-scented blog posts. For the majority of readers, if they think an article is AI-assisted or AI-authored, they will stop reading (78%) and avoid the author in the future (71%)[...] the strongest result in the survey was people overwhelmingly preferred the author’s own writing (98%), with all its flaws and idiosyncrasies,[...]*
there's a say-do gap in play - similar to makeup/no-makeup studies: people claim to prefer natural/no-makeup while behavioral research reveals a preference for subtle/light/professionally-applied makeup
when it comes to ai people love to claim to hate it outright - this thread is full of descriptions of ai writing being impossible to read, totally incoherent, and superlatively bad/worse than average human writing. meanwhile evidence suggests that we might actually kind of like ai writing - with a growing portion of samples making hn's frontpage (https://www.salahadawi.com/hacker-news-ai-detector/state-of-...)
I've seen my share of ai slop and I get where the rage is coming from - I think folks could do a lot better than shouting down every use of ai as slop, especially while hypocritically using ai in their own writing. the pompous elitism on display in this thread alone is gross even for hn
the fact that the HN (or other) frontpages feature a non-trivial amount of AI content is to a large extent owed to the fact that many commenters don't even read the content they're commenting on, another problem in and of itself.
There is no elitism in demanding that people use their own words, or if you want to call it elitism then elitism is healthy and good. Imagine ten years ago you talk to someone online and they told you they outsourced their posting to a call center guy in the third world, that's what this is. Literacy and agency aren't gross, it's what we ought to expect of an educated adult. It's fundamentally a sign of honesty that if I talk to you or read what you write I want to hear words out of your mouth, not some meat puppet version of you hooked up to a chatbot.
"Reading a document like this isn’t just difficult—it is punishing."
I can't tell if this essay was written in earnest or as a subtle troll.
Whichever it is, I stopped reading regardless
On the note of imperfections... My preference is now strongly towards reading someone's stream of thought thrown down at speed over than that same stream of thought shat through the digestive track of an AI agent. At least that feels human.
To me, it also seems like they AI digestion is getting actively worse? As best as I can tell, all the agentic nature and reasoning for code is now making writing actively worse, as the agent pulls across your whole knowledge base and will take that one thought and eagerly join and context it thinks is relevant, with the reasoning spread throughout the page.
For better or for worse I’ve always preferred this, and IME a large portion of people (especially younger or who don’t write/communicate much for recreation or their work) will not read it or assume it was AI.
IMO it is a costly/goodhart-resistant way to “show your work” and help other people understand or challenge your mental model. (IE a justification for something you believe to be true). Overly polished writing is performative, it’s hard to take seriously once you’ve read The Economist/LW enough to see how poorly “well written” correlates with truth
To a certain extent are all wrong or ignorant about almost everything because our knowledge/time are very limited. But it is really important to understand what other people think in order to coordinate with them/align human goals and understanding.
It’s good that the average person taking the lowest-friction path to using an LLM in bad faith is easy to identify now. The more obvious and disliked it becomes, the more they’ll be hit with the stick to actually know things and not bother people. It’s so much worse to be “bad and stupid and not care” than “possibly cringe or wrong”
> At least that feels human.
I feel the same. I've actually started to appreciate things I used to dislike. Like typos, or grammatical errors. I used to see it as a lack of attention to detail. But more and more it now feels like "hey, something written by a fellow human!".
Same thing for video voice overs. Things like a bad quality microphone, or someone who doesn't pronounce things very clearly. Now I go: for sure a human!
There's a lot of misunderstanding here. The quality of LLM writing has not plateaued, it has dropped significantly. They've cut it to the point where people start noticing.
Do a search on "Claude Sonnet 4.5" on Reddit and you'll see lots of disappointed users [1]
If I could give out ratings,
Average human with a degree: 5/10
Sonnet/Opus 5: 2/10
GPT 6 Astra, 5.4 Sol: 3/10
Sonnet 4.6: 8/10
Claude Sonnet 4.5: 9/10
GPT 4o: 6/10
GPT 4.5: 10/10
GPT 3 Davinci (with a lot of coaching): 7/10
[1] https://www.reddit.com/r/claudexplorers/comments/1ta6f9c/i_s...
I'm curious if there have been any formal studies on this. I suspect that people have become better over time at identifying LLM writing styles, and are also combining that with selection/confirmation bias. Dunno what fraction of the perceived drop in quality comes from that.
Yeah, I think I'll start making my own creative writing benchmark. It's a little late to start data collecting, but better late than never.
Have you seen the episode "Darmok" of Star Trek The New Generation series? It is about a people with an impenetrable language aimed at packing as much information as possible by using references and metaphors, not descriptions. I guess this is where LLM language output is heading.
Have they got worse or have we simply gone past the novelty state and are more critical of their outputs?
You can verify this by switching to an older model.
There's a reason ASD-STE100 is mentioned everywhere now. They're at the point where people can't understand what they're saying.
(https://en.wikipedia.org/wiki/Simplified_Technical_English , for those confused.)
I feel like this is more of a curation issue than an issue with using AI itself. If someone is using AI to distill information to the point where important context gets lost, then yeah, I think they are definitely doing it wrong.
What I often notice in pull request descriptions isn't necessarily a problem with them being written by AI. It's the lack of nuance. They don't tell me exactly what the author expects the reviewer to pay attention to, what might be risky, or where they would particularly value another set of eyes. Being able to articulate those things is far more important to me than whether AI was involved in writing the text.
I would also push back a little on the author's point about imperfections. Being able to say that imperfections are fine somewhat ignores the elephant in the room: bias. Bias can have very real consequences. The author may personally be more tolerant of variations or imperfections in writing, but that's not something we can realistically expect from everyone. I'm also not sure it's fair to put the burden of dealing with those biases entirely on the person doing the writing.
That's also why I keep coming back to the expectation of having a distinctive "voice." Not everyone has the same attachment to their writing voice in the first place. For some people, getting their ideas across clearly and being understood matters much more. If voice isn't something they particularly value or identify with, then insisting that they preserve it feels somewhat moot.
“Language can become a screen which stands between the thinker and reality. This is the reason why true creativity often starts where language ends.” — Arthur Koestler
We do not experience the world as words. Words and language are already an hallucination on top of reality and LLMs are an hallucination on top of language.
There are many use cases for LLMs that will help humanity progress. Putting AI in between yourself and the people you want to build productive relationships with is not one of them.
My take here is, I think the real issue is AI pretending to be human by mocking a human style.
I'd always rather read an actual person's thoughts, clearly communicated and written down in their voice. But, there's clearly a subsection of people who don't want to (or aren't confident in / etc) writing their ideas fully.
If they have a bullet point list, and they want to use AI to neaten up the bullet point list and present it back, that's fine by me. I wouldn't have minded reading their thoughts in a messy list, but AI is serving to make it more readable.
What I think everyone hates, is AI being used to convert the list into pseudo-thought-out-prose. It just contains a bunch of bloat that wastes people's time.
So my advice if you use AI to write and don't want to change: keep going, just stop seeing the desired output as human prose. Just neaten up your bullet point list / promot / whatever, and send me that.
I wonder where this discourse is headed, we[0] keep making the same observations over and over again.
[0] myself included https://www.google.com/goto?url=CAESbQHrOzAViiJizn8otPF3AAeE...
A good code or commit comment should succinctly explain the subtle details or motivation not obvious from the content itself. AI comments tend to wax poetic about surface details.
The same goes for articles. I want the author to distill the material using their own voice and hard earned expertise. I don't want a generic text which sort of tells the tale.
> I would much rather someone be themselves and write with their own voice, or with some passion
I recently had a fairly successful article here on HN. The common thread among comments was appreciation for the passion and it being "hand-crafted". I am thrilled my passion shone through despite the dry subject. Still, it's sad that something simply being written by a human has become such a rare compliment.
It's tricky because I don't want to write what no one is going to read (but that for some reason I'm required to write).
Nobody is going to read anything regardless of whether you used a LLM to write it or not (but if you use a LLM they'll use that as their excuse for the virtue signalling aura farm), so honestly why bother writing.
If it's written down in a form that can't be repudiated, a succinct message to any stakeholders covering any decision of substance is good practice. They might not read it, but they cannot pretend I did not make a good faith attempt to inform them.
(And I try to be judicious about what counts as "of substance" so they can't accuse me of flooding them with messages)
I think AI can write good already. It’s all up to the prompt. Strong guidelines. Without them, it will sound like an AI because, just as the article said, the language reflects the writer.
this is intended as a silly observation and not an actual accusation, but for an article about not writing with AI, the following is one AI-ass sentence:
“Reading a document like this isn’t just difficult—it is punishing.”
There's nothing AI about that sentence.
You are oblivious to AI smells then. This is a very common AI writing pattern.
AI didn’t invent the em dash nor “It’s not just X, it’s Y” and sometimes those are appropriate for the situation. Prose should be judged accordingly.
Using both at the same time, tho? Eh.
LLM didn't invent it themselves, they are plagiarism machines. Obviously these kind of patterns had to be popular in the first place to be overused by LLMs
Also we have reached the point where LLMs are pubkishing more contents than humans so we have probably reached a point where humans will inconciously start copying and using LLMs style too.
c'mon, it's literally the top 2 AI writing tropes in a single sentence. i googled "AI tropes", and here are the first 2 tropes listed in the first search result:
1) negative parallelism: "It's not bold. It's backwards."
2) em-dash addiction: "The problem -- and this is the part nobody talks about -- is systemic."
again, i'm not saying the author wrote this sentence using AI. i was just tickled by how stereotypically AI it sounded. for better or for worse, those characteristics are just going to set off the AI detector in people's heads now, as evidenced by the fact that there are like 10 separate comments on this post that picked up on the exact same thing.
It's the stereotypical example of AI in my experience. I.e it's not this it's that.
It's a super common writing and speech pattern.
I kind of agree with this, but I think it’s model specific and a temporary problem.
Opus 5 and Fable 5.1 commit/PR messages are incomprehensible garbage.
However, Astra messages are nearly perfect for a copy/paste to less-technical stakeholders. I maybe fix a line or two.
Give it a year, and I suspect that I won’t even need to make those fixes.
The only problem right now is I can’t get my team 20x OpenAI accounts due to supply constraints. We’re all stuck waiting, hoping Anthropic either ups their game or OpenAI gets more capacity.
We would easily pay $1000 a month per developer/PM for a business tier ~30x account or similar that let everyone use Astra all week without running out of tokens. And that is entirely because of the writing and communication improvements.
These last few days, there has been a small stream of blogposts on HN expressing similar things, and I have enjoyed them all.
My question recently has been how to broach this subject with colleagues who really enjoy producing prose with AI. There is not yet a better cultural shorthand for this sort of thing than "slop" which is a harsh-sounding word and itself sort of a thought-terminating cliché. "I don't want to read what you didn't write" is maybe closer — but it needs a pithier and somewhat more encouraging encapsulation, like "I want to hear it from you".
Has anyone had good experiences setting up professional boundaries or team norms around AI-written docs?
It's something that my company has been struggling with. Developers are generating mountains of code, documentation, and Jira ticket comments. It's incomprehensible and overwhelming.
For the team I lead, my guideline is AI generated is fine but it needs to be human-edited and/or summarized. You want me to read what you're offering? Put some effort into it and meet me halfway. I don't want AI generated gibberish with made-up terms. You'd better also understand what you are presenting as your work. It's been fairly well-received though we're still working on it.
I have some co-workers on other teams who use AI to generate responses to literally everything. Ask a simple question? Get pages of AI generated nonsense in response. They are proving a tougher nut to crack.
There is one person who is so good at writing with AI assistance.
Everyone else, including me, not so much. I envy his talent and ability to so successfully use the new tools. It's just something to think and talk about regularly. Maybe one day we will all be able to use the tools as well as that guy.
Ask questions about how the audience is taking the blog posts. Say that many people seem to really not like LLM writing and ask how are they listening to their readers on this point.
You may find that brings a better and more human focused discussion. Who is the audience and what is best for them? Don't frame it about your preferences but about the real persons that your colleagues seek to talk to.
In my current position(DevOps) we are getting this kind of messages or issues a lot recently. While forwarding these issues to my team lead, i don't have to deal with them which feels great but once its assigned to me... Its another long process to understand what issue actually means and what i need to do.
Thats said, i don't know any professional way for this but currenlty trying to communicate with peers in order to work without a problem.
im sensing about a 20 month lag between hn sentiment and normy-coworker sentiment. managers and business-types pointing out the benefits and challenges of working with an llm i told them about early 2025. im just politely nodding along waiting for them to catch up. theres some emperors new clothes happening at my company so maybe im complicit
"Hey, the comments are too wordy. Can you clean it up so its only important information?"
As much as you want to say "actually do your job, jackass." Being goal oriented has had the best results for me.
I’ve been away. Does anyone have a collection of these posts?
Here are the two I was thinking of:
- https://news.ycombinator.com/item?id=49767937
- https://news.ycombinator.com/item?id=49784816
I’ve seen the problem crop up in several places.
At a broad level, the issue isn’t generated content, it is the ratio of verification capacity to generation capacity (V/G). Your pain is because generation capacity has increased significantly, while verification is laborious and capacity has not (and can not) catch up.
Unlike spam which is from external sources and can be ignored, messages from other employees have to be responded to. I guarantee this is creating bottlenecks all across the firm, outside of the individuals who are feeling productive.
For fixes, theres theoretical approaches that might work?
If you need leadership to help you, then this issue has to become something that is on their radar, which means that something needs to go wrong or costs need to be registered.
The shortest conversation for that is to make people aware that generation has improved individual productivity, while moving the costs of that production to the rest of the firm.
If leadership is not at the stage to listen, then you need to move the costs you are incurring to the people who are sending them to you. Maybe set time aside to sit down with whoever sent a PR and then read what they sent together, to understand it.
It also makes a difference if tokens are being subsidized or not. If the firm doesn’t care how many tokens are being used, then you are naturally going to have over production.
AI-generated writing does not bother me, unless it is intended as prose writing, where the flow communicates as much as the specific facts that are described. For technical writing, I usually just scan for important information anyway, and in that sense the only thing that bothers me about AI-generated writing is if it's too verbose and obscures the important information by leaning ineffectively on jargon and high-minded verbiage. I find that Claude Opus is very guilty of this, whereas GPT 5.6 variants are not.
>the only thing that bothers me about AI-generated writing is if it's too verbose and obscures the important information by leaning ineffectively on jargon and high-minded verbiage
But so many READMEs and design docs are degraded to this level by uncritical/unmonitored use of Claude. You make it sound like it’s an exception, not a rule.
In my experience, that brain damage inducing Claude-style technical writing is everywhere. It sucks not only because of its verbosity, but because it includes completely pointless tidbits extracted from random extended LLM sessions, with no attempt to prune redundant or gratuitous details.
It’s funny, this pattern is very similar to how I write code with AI these days.
I write almost every line but I have one or more agents always running against the changes alongside me evaluating correctness, suggesting approaches, auditing test coverage, etc.
I’ll also use it for grunt work refactors, scaffolding, resolving merge/rebase conflicts, etc.
It goes both ways.
If you don't get to the point and continue to go on and on, I will use a LLM to summarize what you wrote and then decide whether to spend time listening/reading what you wrote.
> Reading a document like this isn’t just difficult—it is punishing.
I hope this was an in-joke but my fingers still reflexively closed the tab as soon as I saw it and I had to go reopen it again…
We're at a point where I assume that whatever I read is something the "author" endorses, but may not have written. And that may be more important than the specific words an author may have chosen themselves. I doubt we can ever go back.
What should worry us is that AI may introduce a slant that the prompter assimilates during their conversation with the tool, which could introduce a subtle control lever over public opinion.
I don't disagree with the main points of the article. But I feel like soon with all the writing that's been hating on AI writing recently on HN, the LLMs are going to be really good at writing articles about how bad AI is for writing...
> Reading a document like this isn’t just difficult—it is punishing.
At this point I can't tell if it's a pun or the anti AI article is AI written as well. This sucks.
One aspect I am bothered by is how all writing style is becoming homogenized. Individual writing "voice" is becoming one collective AI voice.
The other comments here were written by LLMs that have a hard time detecting meta-irony.
Apparently most readers have read too much AI output that they don’t recognize a bit of irony as a joke.
> Apparently most readers have read too much AI output
You’re giving them too much credit.
The irony is that this article reads a lot like it was written by AI.
Yes, you can easily prompt this writing style.
>I had someone write me a personal message about a sensitive topic that was clearly workshopped with AI in an effort to nuance the conversation and not offend me. But the message became impersonal, dispassionate, and disjointed. It had all of the parts, but it didn’t make sense as a whole. I wasn’t interested in reading it, or responding. I would much rather someone be themselves and write with their own voice, or with some passion, even if they can’t quite find the words, or risk offending me.
Only a privileged person can say this. Not everybody is a native speaker or good communicator. As an IT guy somewhere on the spectrum, I have damaged my career on multiple occasions by clumsy writing.
I'm kind of okay with reading something you don't write(entirely). But reading something that you even neither write nor read? waste of time.
The most sensible solution that I've seen so far is to delineate human spaces from LLM spaces. Technical mumbo-jumbo can be LLM as an artifact, and human thought can just point to them.
To me it is just a bittersweet irony: people who don’t know what a book is for 30-40 years or in earlier days got the RTFM (gif smiley not included) treatment dump one liners to produce a giant load of text corpus that needs to be processed by an AI to get the gist out of it with sometimes quite some refinement by asking many questions.
So there it is: cruel people who don’t even read their own prompts while happily believing others read their text load and pat the genius on the shoulder for being such a great guy in writing, while in essence it turned the RTFM movement on its head: UUOAI - “Use your own AI.”
People cannot communicate better by using the digital equivalent of logorrhea.
It's funny to see this upvoted on a site where a huge portion of the front page content wasn't written by humans.
> I also see more and more pull request summaries clearly written by machines for machines. They are rich in detail: this was changed to that, these things were split, those things were merged, this was left untouched, tests were added for this, and so on. But I’m left asking: Why are we doing this? What is the value? How risky or urgent is this work? Where do you want my input? What should I pay attention to?
My own solution for this (for now at least) is to add a preamble section “AI;DR” to PR descriptions, filling what a TL;DR section would have done in the past. I’ve socialized to my peers that no part of such a section will be written by AI, and it should contain exactly this missing high-level content described by the author.
I've always wanted to write this essay, but now I no longer need to. My title would have been pretty similar: If You Want Me to Read It, Please Write It.
One minor addition: when a (conscious) human writes a proposal, a question or a request, they'd instinctively choose words with which they can express their uncertainty in some of the details. I as a reader can detect these uncertainties and I can choose to do more research assert or correct them, if it feels appropriate. AI-generated texts miss those cues, therefore they are much harder to read and connect to.
Writing for yourself is crucial. Writing is in a way thinking and understanding. If you can't write for yourself, then don't write at all if you expect someone else to read it.
I have a large and complicated project at $DAYJOB where I am the solo dev with a large team of domain experts. They AI dump me with suggestions (rather rarely, thank god) for features with complicated implementation details despite having no insight into the codebase. What is the point of that? I can't discuss it with them. If I jump on a call, they'd have to say "oh sorry — I don't know, I'll get back to you".
When I need to send them a message about something domain specific I have no knowledge of, I'll spend hours understanding it as far as I can. And then wording it in a way that makes sense to me. I use AI A LOT in the understanding and proof reading of messages, but the wording and writing is mine.
I'd rather understand something partially, write a bit poorly, and send an imperfect message than to AI dump a perfectly formed message which takes everything into account but not really understand everything that I'm sending. It closes the door to actual communication between two humans.
Just imagine the quality of text we will have if we re-feed this into an llm a thousand times...
The juxta-positioning of the ambi-dextrous personification of the meta-sematicism is going to be both rich and soul transpiring....
It's gonna be basically a fingerprint in your soul, from my soul...
My org, like every org, is struggling with this as well. We find ourselves writing skills like `/decomment` and often running several "minimize" passes on PR descriptions, etc.
One thing that has markedly helped is tweaking our AGENTS.md prompt and prose-heavy skills like `/pr` or `/ticket` to specify that all technical output needs to be written in ASD-STE100 Simplified Technical English. It has helped us so much. We still have to `/decomment` or minimize to remove unnecessary stuff that's been written in ASD-STE100, but the output is a lot easier to digest.
OTOH, we have a guy in marketing that has taken it upon himself to use Claude to draft these wildly complicated and dense implementation specs that he doesn't understand for various things that he just hands off to my team to implement. He has no context for how our product is built, the various layers and their interactions, etc. so they're often flat out wrong and overly complex, invent systems that do not exist, all of that. Then he gets mad that we're not moving as fast as he wants, because he gave us the implementation spec! I have been struggling to come up with a way to diplomatically tell him that his actions are slowing us down, and if he'd just describe what he needs instead of speccing it himself, we'd be twice as fast...
I feel the same way, and yet I'm also guilty of this. I'm trying to reach out to an expert, and because of the stakes, I want to write it exactly right. I'm using Claude to write it, but I don't like how Claude writes, so I rewrite it in my own words, and I still don't like it. I've iterated over this for several days. Finally I talk to my wife, she gives me some pointers, and I just write it down perfectly.
And yet, every person high up in business that I talk to, including my wife, advises me to use Claude for this sort of thing. I don't get it. Claude text sucks, and it hurts my eyes when I'm forced to read it.
As someone who is usually on the other side of this interaction (software engineer at a company receiving bug reports), it's much easier when people write in their own words (even if they sometimes use the wrong terms or say something which can't be true) than when they use an LLM to make things clear and correct. I want to know what you're trying to do, what you expected to happen, and what actually happened - that doesn't benefit from having an LLM rewrite it.
Sure, I use precise and technical language, but I'm still a human and I can understand what people are getting at when they use imprecise terms. But if the text passes through an LLM I can't tell whether the result actually matches the original problem or whether the LLM has made guesses or filled in blanks.
Bug reports aren't so high-stakes (although poorly written bug reports can often be useless). I've got something to sell, and I want this expert to confirm that what I've built is better than what's currently available. That's a much harder approach. You want people to send bug reports. You don't want everybody to try to sell something to you. But I really think that I've solved the problem that she identified, so I want her to take a look at it instead of dismissing it as most people would probably do when a random nobody claims to have solved something. It doesn't help that this entire issue is way out of my comfort zone.
If there's value in writing, then it's in the writing. It doesn't matter who wrote it, even if it was monkeys on typewriters.
I ask coworkers point blank to send me a non-llm version of their slop sagas. The initial exchange can be a bit awkward, particularly with new hires, but consistently asking for human prose has worked for me.
How do I type this character on my English keyboard? — also not reading what you didn't write.
On Windows, use wincompose. Space followed by a hyphen gives you an emdash with the cute half-width spaces around it. On Android, I can get an emdash by long pressing hyphen. There's also a keyboard shortcut on the mac for it. No idea about Linux/BSD or iOS, unfortunately.
On Windows I also use QuickAccent, now part of PowerToys. I could never remember the Alt-code for the em dash. Apparently it's Alt-0151. With QuickAccent I hold a letter and right or left arrow and up comes a menu of related symbols. It's pretty neat.
On Mac, it's option-shift-hyphen. Option-hyphen gives the en dash.
On Linux, you may be able to type Ctrl-Shift-U to get unicode entry and type 2014 and Enter. 2013 is the en dash. Again, I will never remember that.
In Linux if you enable the compose key it’s the compose key followed by 3 dashes.
cute little half-width spaces?
https://en.wikipedia.org/wiki/Whitespace_character#Hair_spac...
On Linux, it's whatever you say it is, after reading through `man xcompose`.
on a Mac, it's Option+Shift+-
— (em dash)
Option+- gives you an "en dash": – which is longer than a normal hyphen.
This sentiment gets echoed over and over. While I can certainly see the value in human-created creative works, if the subject is pure facts, there are only so many ways to state the facts without being caught in the weeds and missing your entire point.
Moreover, people are finding it hard to differentiate what is an is not AI-generated with newer models, often attributing original work with those of LLMs. It has just become an easy scapegoat for lazy comprehension and a desire to do less. You are jumping at AI boogeymen.
Just about the only thing here I can level with you on is, yes, AI is far from perfect and will continue to advance. Otherwise, so much of this reads as fruity prose to excuse apathy.
A lot of my problem with AI writing is that the facts aren’t checked. You get a big cheesy dramatic arc but the core story isn’t factually correct. There’s only so much time in a lifetime, too little to waste fact checking someone else’s work because they couldn’t be bothered to.
I have junior devs submitting huge PRs with multi paragraph descriptions about load baring root causes and "why this works". Then I ask them directly what are the steps to test this and they don't know, they have never run the code.
Rather than entertain the idea of reviewing the code I just send it back to them until they work out how to run it. And almost always they submit a new change because the last one didn't actually work, despite how confident and articulate claude was to them.
So, your company doesn't have CI/CD and gates? AI isn't your problem.
A change can pass CI without actually making the change required. The tests test that the code does what it does, not that it does what the user wanted.
AI is really good at making tests that pass even if the change itself is wrong.
Fwiw these harnesses run appropriate tests before pushing anyway, so CI CD pass is almost a guarantee. Correctness is another matter.
The title is the conditional that underpins the claim.
Do you read our AI output end-to-end and decide that it’s the accurate content communicated concisely? If yes, this is not an issue.
If you are sending people generated text you haven’t read, how do you know what you sent doesn’t fall in the category of writing the author describes?
It's curious that you're pinning the blame of "excusing apathy" on the reader and not the writer. Why doesn't it go both ways? Given that producing LLM text is so cheap, your proposal amounts to an asymmetric war which is ultimately a DDOS on my finite time and attention. Or in other words: if it wasn't worth your time to write it, it's not worth my time to read it.
Trying to understand your position.
Let us say someone had few good ideas, and seeded them into a prompt, and after few back and forth, web searches via agent, feedback from the author, a piece of work was produced and the author decided to share it as a blog.
This is not much different than how people are producing original work LLM in areas such as math.
Would you object to reading their work because it was a byproduct of collaboration between AI/Humans? What about songs? movies? math proves? and software produces as such?
Not Analemma_ but I'll give my two cents.
In my mind the issue isn't that it was crafted by AI, but that it wasn't crafted with the intention of getting the point across in a way that respects the reader's time or energy. The result is often needlessly wordy with information buried in a sea of flourish. This might be forgivable if the author had a certain entertaining style, but LLMs are aggressively mediocre by their nature, making wading through it painful.
Its certainly possible to get an LLM to write concise prose but thats not how most people are using them. We can talk all day on HN about the ideal LLM user that is conscientious, reviews the output, puts time and effort into stylistic concerns, etc. But none of that actually matters because thats not what people are doing; the entire reason they're using an LLM in the first place is so that they don't have to do that.
On that we agree then. I also don't want to go through mountains of tangled text written by an AI.
But what if the AI models improve their writing style considerably to match that of the best authors? which I think is plausible soon.
I feel in the near future, what really matters the most is the seed and the dialog/back and forth more than then who wrote what. And that the best content would be a by product of human/machine interaction atleast given the current generation of tech.
Until perhaps, in the the not so distant future, the math can encode our taste, judgments and develop imagination, then maybe the best human wouldn't be able to create anything better than the machine. But then again, we want to communicate the inner state of human being, because we want to build relationship, so content for consumption might be dominated by machines, but communication as way to build relationship might stay genuinely human...but even that might change, what if machines eventually become better at relationships than humans? that could be how AI truly "kill us"..I think within few decades we will discover that our intellect was a short period of cosmic evolution, and we will accept our fate as animals with analog organic intelligence that helped to bootstrap what comes next, this might be scary but humbling, and I think it's inevitable.
The lazy ones are the people spending 60s prompting then hitting copy/paste. They dump it on me as is and call it a job well done, which leads to me doing their job for them or calling them out to stop the behavior.
I don’t like doing either of these things. I don't want to do your job and I don't want to lecture about how it isn’t “your work” if you don’t touch the content after an LLM spits it out. It is rude and selfish to put me in that position.
Seconding the author, at my last job a new VP started pushing AI hard, in every use case, and suddenly I had to read AI-generated high-level design documents that were both overly-detailed, and... really difficult to read.
There was just something about them, like they were written by someone obsessed with detail but without a clear idea of the audience they were writing to.
The result always ended up being a document that was way too long, way too detailed, without context for those details, speaking over the head of the reader.
And I don't want to read the millionth iteration of some old man complaining about the kids these days, written by a human or not, yet here we are.
If you don't like AI slop. Don't read it. But wasting your time generating human slop to complain about AI slop is so obviously futile that it immediately identifies the writer as lacking the capacity for reason or emotional clarity, or merely seeking attention for their self-promotion with clickbait.
The problem is, some of us work with colleagues, or even bosses, who throw AI replies at us in nearly every conversation — so we don’t have a choice to NOT read it.
I am sorry, I didn't take into account people in literal slavery. I am very sorry to hear about your condition. Perhaps you should contact the UN or some human rights organization, they may be able to help.
Just throw it into your LLM and de-slopify it. I sometimes just ask "wtf is he on about, 2 paragraphs max". Tends to work. Sometimes you still have slop afterwards but at least it's 2 paragraphs of slop instead of a wall of it. (Not condoning the behaviour, just coping with it :P)
You don't have the choice to read or not read except when you read for leisure
I’m fine with reading something you didn’t write if it has the information I need. I’m not fine with reading something you didn’t read and will just waste my time.
"The design document is no longer a proposal to build consensus, bring people along with you, and refine ideas through slow, deliberate thinking"
I'm sorry but that's the last thing (almost last) I want to read.
I want an 'excellent summary' of the design including motivations and concerns.
Yes - the AI is bad at that but hopefully it will get better.
I think the future is people's AI agents evaluating the content generated by other people's AI agents and vice versa.
What the humans would see would be what is filtered up to them after all of this evaluation is done. And content generation would be the reverse. People would provide a seed of an idea and then their agents would turn that into potentially useful work.
I'm finding that executives at work are generating AI slop and sending it at each other and teams that are then meant to digest it, without any critical thought. Often these are sloppy and contain broad mistakes that aren't relevant to our company's setup.
For me the only thing that matters is the content the quality of the writing. If you're primarily interested in something else than content and quality, there is no point in reading at all. So I don't understand these anti-AI posts. If all they say is that most AI written texts are too long and badly written, I agree. However, that's also the case for many texts written by humans.
“Reading a document like this isn’t just difficult—it is punishing.”
I’d be curious whether the author composed this sentence himself or it was the output of AI. Personally I often find myself “it’s not X it’s Y” and then recoiling in disgust and rephrasing it simply because AI has made it so grating from overuse.
It was very clearly on purpose, if you look at the surrounding prose.
As an educator in university I have found the project proposals of students have become longer, to some degree more well written and quite certainly have LLMs involved.
This becomes problematic when it turns out people haven't even read what they sent me.
To clarify: this is about voluntary projects that do not win them extra points.
If it isn't worth the time for someone to write something themselves, it isn't worth my time to read it.
> I don’t want to live in a world where you use AI to summarize something important into unreadable text, and then I use AI in an attempt to decipher it. I want to hear you, imperfections and all.
It doesn't fundamentally change the equation if I use AI to prepare and then write it myself. If I'm using AI effectively, it's likely that you won't be able to tell.
This sort of post is increasingly coming off as high and mighty, where the user thinks they are being exceptionally creative and other people who are using AI are using it mindlessly.
Also, writing, indeed any making activity, forces you to think clearly. Often that is the deepest purpose of prose, even code - forcing the author to confront the problem and sharpen their own understanding of the world we share. That it then communicates something to another person (or to a machine), is a happy bonus.
"In this essay, I want to express my frustration, and describe how I’m finding value in AI for writing.'
"I don't like the way other people use AI, but the way I use it is okay."
I'm going to stop reading right there. I don't want to read that either.
> Reading a document like this isn’t just difficult—it is punishing
I stopped reading here. I’ve become so sensitive to LLMism, I don’t want to go on. Tip; omit the needless ”it’s not X, it’s Y” and just write ”it’s Y”.
It's clearly a joke. The rest is readable and well written, and free of LLMisms
Attention is expensive these days.
The article itself is written by AI lmao, notice the —
The problem is that whenever someone writes something they are accused of using AI. I know people who are accused of this because they know how to use em dashes, or even just write bullet-points. Reminds me of the bores who accused me of using Google Translate for something, and I had to show them that the language in question wasn't even on Google Translate.
If you are looking for top-notch corn-fed drivel, look no further than my comedy newsletter. Here’s the latest issue: https://yelluwcomedy.substack.com/p/sweater-weather?r=7c090l... It’s full of all the questionable wonders only a human mind can think of. Hope you enjoy it.
How do I type this character on your English keyboard? — also not reading what you didn't write.
On Macintosh keyboards it's ⌥⇧-. On the iOS soft keyboard, it's a long-press on the hyphen key. People often point out that various software will turn a double hyphen -- into one by default. We're overflowing with different ways to type an em dash, for those with any curiosity to learn.
I'd like to know how to type it with voice dictation.