The financial market things are over my head and I don't have a dog in the game, but I think "Nobody is replacing salesforce with their internally vibe coded software" is just false? Both taken literally [0] [1] and as denying the general trend. Just in my company we already replaced WMS software subscription with own solution, and I wouldn't be able to write it fast enough and maintain it by myself without the use of Claude Code.
I'd say "Not perfectly or with every edge case handled, but well enough that the CIO reviewing a $500k annual renewal started asking the question “what if we just built this ourselves” " is an accurate description.
I didn't think something could be worse than everyone using Salesforce, but everyone using a different, constantly broken, incompatible SF clone that no one understands may be that.
Agreed. If anything, it puts downward pressure on pricing. Even if the CIO still buys Salesforce or whatever other tool, they won't be willing to pay as much.
I thought there would be one or two results, perhaps a result of poor reoccurring phrasing. Nope.
>i'm going to change your diaper and burp you
>Carlito is a very good boy Go piss in your diaper you big baby
>He doesn't care. He is a big baby who filled up his diaper with pee pee and poo poo
>you are a big baby and i am going to change your diaper and burp you <
>To be clear I call executives of multi trillion dollar companies scumbags and if you can't deal with that I'm not sure what to do. Burp you? Change your diaper?
>I am going to change your diaper and burp you
>Yeah man it's real authoritarian to say your second name is doodoo. Go change your diaper you big baby
>Yeah because you're a big baby with a big full diaper
>Hello sir this is your uber outside. I have your order from the diaper store
Ed's main thesis is that cost is unsustainable for AI companies but this is clearly wrong.
The unit cost is going down and has gone down by more than 20-30x over the years. Sure, the fixed cost of training is going up but that's because of the implied returns. Once the returns to training don't happen, it would simply reduce modulo cutoff date updates. The companies have a choice to just stop training and focus on inference cost reduction.
What am I missing here? Unless the consumers decide that they are no longer willing to pay the same amount as before and their expectations are rising with prices falling, what else?
I do enjoy a good Ed Zitron sneer. The fact that the original article moved markets says a lot about the critical thinking skills of stock market traders.
You should look into how he destroyed the small indie MMO Darkfall and gave the game 2/10 without ever playing it, in a Eurogamer review a few years ago. The developers had receipts and could prove that he hadn't played it.
It doesn't have any material effect on this article, but it says something about his ethics.
I'm somewhat halfway through the original memo, and I hate the fact that kernels of truth lie here and there. For example, I used to work as a full-stack developer for about 2 years, and now I got forced into what the memo calls "gig economy" just to pay the rent, because companies slowed down hiring junior developers thanks to... Honestly, I don't really care at this point what I have to thank for.
All I know is, whenever I read testimonies from people whose companies suddenly decided to force LLM usage for productivity to be "AI first", having colleagues opening PR's who are only machine reviewed with implementations they cannot justify themselves outside of "Claude wrote it", makes me burnout just reading them. And it's only going to get worse until it becomes better, but not for the developers.
Honestly, the one thing that I could see justify all the investment companies make for LLM-assisted coding is the full automation of software production. I can only see the current state of things as the "end game" for them, only if they suddenly decide to jack up pricing to tap directly on the corporate budget and not the individual developer's budget.
I don't think Ed doesn't comment about the actual tech. Here are some things he has said before and please tell me if these still hold in the spirit?
> You cannot "fix" hallucinations (the times when a model authoritatively tells you something that isn't true, or creates a picture of something that isn't right), because these models are predicting things based off of tags in a dataset, which it might be able to do well but can never do so flawlessly or reliably.
ChatGPT is fairly reliable.
>Deep Research has the same problem as every other generative AI product. These models don't know anything, and thus everything they do — even "reading" and "browsing" the web — is limited by their training data and probabilistic models that can say "this is an article about a subject" and posit their relevance, but not truly understand their contents. Deep Research repeatedly citing SEO-bait as a primary source proves that these models, even when grinding their gears as hard as humanely possible, are exceedingly mediocre, deeply untrustworthy, and ultimately useless.
This is untrue in spirit.
> You can fight with me on semantics, on claiming valuations are high and how many users ChatGPT has, but look at the products and tell me any of this is really the future.
Imagine if they’d done something else.
Imagine if they’d done anything else.
Imagine if they’d have decided to unite around something other than the idea that they needed to continue growing.
Imagine, because right now that’s the closest you’re going to fucking get.
This is what he said in 2024. He really thought ChatGPT is not in the future.
There are so many examples and its clear that he's not good faith and has consistently gotten the spirit wrong.
Well from my point of view. When they talk about gigawatt datacenters, then yes it is economically nonviable. You just need to know the scale of a gigawatt to realize that we need to start building power plants and fortifying the power grid to ship a gigawatt of power to a single location. Until the build out which takes years mind you, it is competing with other consumers of power. Lets take another huge consumer of power like a large steel mills use 100 megawatt. So if that power becomes more expensive because of datacenters, then the price of steel will go up. And if the price of steel goes up it affects a lot of things in the economy.
We are facing a situation that the short term effects are on memory and storage prices going up and lack of jet engines. Long term we wont be able to build actual buildings and ships without financing it with even more debt than today and everyone in the economy is going to service that debt through the price.
Cost, debt, difficulty forming a moat, gap between what the product promises and what it can do, and the difficulty actually raising capital required.
His style is acerbic and (imo) excessive sometimes. But he's also one of a minority of journos actually looking at the numbers and adding them up. Which seems to be a rarity
Disagree. He's cherry picking an extremely limited subset of numbers, based on a weak understanding of the industry and a lack of access to a lot of private data, and taking advantage of vulnerable people.
> "What if our AI bullishness continues to be right...and what if that’s
actually bearish" - what if pee pee was poo poo
Despite the vulgarity, it is exceptionally illuminating to how much some of these slop pieces are just a mere pretension of rhetoric. I see this pretty consistently with a lot of the material I come across on the job that's gone through the LLM meat-grinder.
Also, the comment made me giggle like a little kid.
What's pretend-rhetoric about it? They're positing agents will prove to be very capable, but that this would ultimately be a bad thing by automating away too much of the economy. You can argue whether that's plausible or not, but it isn't an incoherent or vapid argument.
I suggest you read the annotation if that question isn't just rhetorical. I'm not familiar with Ed, but he has a pretty good take down in here if you can get past his somewhat juvenile writing style.
It is a problem when your doomsday timeline for obsolescence is behind the minute you publish.
The memo itself was fantasy doomer porn on day 1.
> "Here is an annotated version of the Citrini Memo with my own intro. It is analyslop - scare-fiction written to ingratiate AI boosters and analysts/traders with tales of ultra-automation and socialist data center policies. Shameful that the markets reacted at all."
I've started to feel like Ed Zitron is actively hurting people I care about.
I'm lucky to have worked in the field for a long time, and be able to spend a lot of tokens. In the last month it's become clear to me that the tech works. The science is done, and what's left is engineering.
There are a lot of risks and mitigations and theory to build, but it's all solvable. The tech isn't mature, but neither was the Internet 30 years ago. And we built transatlantic cables and ran new wires to everyone's house.
People I care about, engineers with 20 years of experience, are having mental health breakdowns, caused by Zitron's work. They insist the tech will never work, and avoid learning about it, becoming progressively more paranoid and isolated. I'm trying to be supportive and help them start to recover, but it's slow going.
If someone is having a crisis about this, I hope they start talking to a therapist. I don't need them to agree with me, but I do need them to not harm themselves.
There's nothing to recover from, what are you even talking about? I'm not a token user (and I can't make predictions about the future and whether it will force me to use token but still). That the industry is collectively having a delusion about what constitutes good software (in all senses of the word - functionality and consequences for society) is clear to see, something I too fear we might never recover from, but I stand quite clearly on the side of people not of corporations hoping to extract more more more.
Ed Zitron, from what little I have heard of him, seems incredibly irrational. I don't think I've ever seen anybody stick their head deeper in the sand more than I've seen him do.
It's one thing to dislike or even detest something, but to constantly claim it is worthless and without use when people are already benefitting from it everyday is nothing short of delusion.
The financial market things are over my head and I don't have a dog in the game, but I think "Nobody is replacing salesforce with their internally vibe coded software" is just false? Both taken literally [0] [1] and as denying the general trend. Just in my company we already replaced WMS software subscription with own solution, and I wouldn't be able to write it fast enough and maintain it by myself without the use of Claude Code. I'd say "Not perfectly or with every edge case handled, but well enough that the CIO reviewing a $500k annual renewal started asking the question “what if we just built this ourselves” " is an accurate description.
[0] https://lovable.dev/blog/how-a-startup-replaced-a-salesforce...
[1] https://seekingalpha.com/news/4144652-klarna-shuts-down-sale...
I didn't think something could be worse than everyone using Salesforce, but everyone using a different, constantly broken, incompatible SF clone that no one understands may be that.
Lol give it 12 months.
Agreed. If anything, it puts downward pressure on pricing. Even if the CIO still buys Salesforce or whatever other tool, they won't be willing to pay as much.
> Commented [1]: what if pee pee was poo poo
Rarely do I read something that starts off with such promise!
Don’t encourage his diaper fetish! [0]
[0]: https://bsky.app/search?q=from%3Aedzitron.com+diaper
I thought there would be one or two results, perhaps a result of poor reoccurring phrasing. Nope.
>i'm going to change your diaper and burp you
>Carlito is a very good boy Go piss in your diaper you big baby
>He doesn't care. He is a big baby who filled up his diaper with pee pee and poo poo
>you are a big baby and i am going to change your diaper and burp you <
>To be clear I call executives of multi trillion dollar companies scumbags and if you can't deal with that I'm not sure what to do. Burp you? Change your diaper?
>I am going to change your diaper and burp you
>Yeah man it's real authoritarian to say your second name is doodoo. Go change your diaper you big baby
>Yeah because you're a big baby with a big full diaper
>Hello sir this is your uber outside. I have your order from the diaper store
"Search is currently unavailable when logged out"
Ed's main thesis is that cost is unsustainable for AI companies but this is clearly wrong.
The unit cost is going down and has gone down by more than 20-30x over the years. Sure, the fixed cost of training is going up but that's because of the implied returns. Once the returns to training don't happen, it would simply reduce modulo cutoff date updates. The companies have a choice to just stop training and focus on inference cost reduction.
What am I missing here? Unless the consumers decide that they are no longer willing to pay the same amount as before and their expectations are rising with prices falling, what else?
I do enjoy a good Ed Zitron sneer. The fact that the original article moved markets says a lot about the critical thinking skills of stock market traders.
You should look into how he destroyed the small indie MMO Darkfall and gave the game 2/10 without ever playing it, in a Eurogamer review a few years ago. The developers had receipts and could prove that he hadn't played it.
It doesn't have any material effect on this article, but it says something about his ethics.
Its word against word in this situation. The logs prove nothing as they are easily modifiable and the devs had a good reason to do so.
The original memo: https://www.citriniresearch.com/p/2028gic
HN discussion: https://news.ycombinator.com/item?id=47114579
I'm somewhat halfway through the original memo, and I hate the fact that kernels of truth lie here and there. For example, I used to work as a full-stack developer for about 2 years, and now I got forced into what the memo calls "gig economy" just to pay the rent, because companies slowed down hiring junior developers thanks to... Honestly, I don't really care at this point what I have to thank for.
All I know is, whenever I read testimonies from people whose companies suddenly decided to force LLM usage for productivity to be "AI first", having colleagues opening PR's who are only machine reviewed with implementations they cannot justify themselves outside of "Claude wrote it", makes me burnout just reading them. And it's only going to get worse until it becomes better, but not for the developers.
Honestly, the one thing that I could see justify all the investment companies make for LLM-assisted coding is the full automation of software production. I can only see the current state of things as the "end game" for them, only if they suddenly decide to jack up pricing to tap directly on the corporate budget and not the individual developer's budget.
Offtopic - The success of coding agents must be Ed Zitron's nightmare.
He has been a perpetual bear
I don't think so.
His argument is not "this tech doesn't work", but rather "these businesses aren't economically viable"
And that the smoke and mirrors accounting and perpetual thirst for more billions indicates just how unviable it is
Whilst he does dunk on LLM capabilities, the framing is the business angle - can Anysphere etc. actually form a moat and make a profit?
I don't think Ed doesn't comment about the actual tech. Here are some things he has said before and please tell me if these still hold in the spirit?
> You cannot "fix" hallucinations (the times when a model authoritatively tells you something that isn't true, or creates a picture of something that isn't right), because these models are predicting things based off of tags in a dataset, which it might be able to do well but can never do so flawlessly or reliably.
ChatGPT is fairly reliable.
>Deep Research has the same problem as every other generative AI product. These models don't know anything, and thus everything they do — even "reading" and "browsing" the web — is limited by their training data and probabilistic models that can say "this is an article about a subject" and posit their relevance, but not truly understand their contents. Deep Research repeatedly citing SEO-bait as a primary source proves that these models, even when grinding their gears as hard as humanely possible, are exceedingly mediocre, deeply untrustworthy, and ultimately useless.
This is untrue in spirit.
> You can fight with me on semantics, on claiming valuations are high and how many users ChatGPT has, but look at the products and tell me any of this is really the future.
Imagine if they’d done something else.
Imagine if they’d done anything else.
Imagine if they’d have decided to unite around something other than the idea that they needed to continue growing.
Imagine, because right now that’s the closest you’re going to fucking get.
This is what he said in 2024. He really thought ChatGPT is not in the future.
There are so many examples and its clear that he's not good faith and has consistently gotten the spirit wrong.
>His argument has never been "this tech doesn't work", but rather "these businesses aren't economically viable"
Why? because of cost?
Well from my point of view. When they talk about gigawatt datacenters, then yes it is economically nonviable. You just need to know the scale of a gigawatt to realize that we need to start building power plants and fortifying the power grid to ship a gigawatt of power to a single location. Until the build out which takes years mind you, it is competing with other consumers of power. Lets take another huge consumer of power like a large steel mills use 100 megawatt. So if that power becomes more expensive because of datacenters, then the price of steel will go up. And if the price of steel goes up it affects a lot of things in the economy.
We are facing a situation that the short term effects are on memory and storage prices going up and lack of jet engines. Long term we wont be able to build actual buildings and ships without financing it with even more debt than today and everyone in the economy is going to service that debt through the price.
Cost, debt, difficulty forming a moat, gap between what the product promises and what it can do, and the difficulty actually raising capital required.
His style is acerbic and (imo) excessive sometimes. But he's also one of a minority of journos actually looking at the numbers and adding them up. Which seems to be a rarity
cost is going down 20x, 30x over the years so he's wrong about this.
Those are very large reductions - can I ask you for a source?
And why is the error bar so large?
https://epoch.ai/data-insights/llm-inference-price-trends
> The rate of decline varies dramatically depending on the performance milestone, ranging from 9x to 900x per year
Disagree. He's cherry picking an extremely limited subset of numbers, based on a weak understanding of the industry and a lack of access to a lot of private data, and taking advantage of vulnerable people.
>taking advantage of vulnerable people
What on earth do you mean by this? Who is getting taken advantage of?
I'm not sure how anyone can respond to that, without asking you to divulge that private data
Which success? I still see those things churning out laughably wrong code at every turn.
It's not a point and click code machine, but it's laughably wrong to say they just churn out laughably wrong code.
Half of developers are below average. Half of developers say that the code Ai produces is amazing. Would you like a venn diagram?
You’re saying that AI is already good enough to replace 50% of developers. Sounds like you agree it will be very important.
He's saying the good half of developers have to deal with the increased slop output of the bad half. Probably will be overwhelmed by it, in the end.
You're telling on yourself here
I guess everyone stopped using Claude Code already then, if it doesn't work.
It doesn't work precisely because people are using claude.
If it worked, there'd be no people using it.
Nobody goes there, it's too crowded.
sorry but this is just cope at this point
The copium is strong in these threads. There was merit to that a year or two ago, but at this point it's just sticking the head in the sand.
This is not off-topic at all!
Some one should compile concrete predictions that he made vs how they turned out.
Have they been successful?
He is funny and entertaining but does he provide constructive investment advice? I am not so sure.
> "What if our AI bullishness continues to be right...and what if that’s actually bearish" - what if pee pee was poo poo
Despite the vulgarity, it is exceptionally illuminating to how much some of these slop pieces are just a mere pretension of rhetoric. I see this pretty consistently with a lot of the material I come across on the job that's gone through the LLM meat-grinder.
Also, the comment made me giggle like a little kid.
What's pretend-rhetoric about it? They're positing agents will prove to be very capable, but that this would ultimately be a bad thing by automating away too much of the economy. You can argue whether that's plausible or not, but it isn't an incoherent or vapid argument.
I suggest you read the annotation if that question isn't just rhetorical. I'm not familiar with Ed, but he has a pretty good take down in here if you can get past his somewhat juvenile writing style.
It is a problem when your doomsday timeline for obsolescence is behind the minute you publish. The memo itself was fantasy doomer porn on day 1.
Where does the title come from?
What is this document?
What is the context?
Good questions. Here is the author's BlueSky post about it:
https://bsky.app/profile/edzitron.com/post/3mfkc63h6222l
> "Here is an annotated version of the Citrini Memo with my own intro. It is analyslop - scare-fiction written to ingratiate AI boosters and analysts/traders with tales of ultra-automation and socialist data center policies. Shameful that the markets reacted at all."
I've also heard Cory Doctorow recently offer a similarly dismissive view, describing AI as "just statistics".
I've started to feel like Ed Zitron is actively hurting people I care about.
I'm lucky to have worked in the field for a long time, and be able to spend a lot of tokens. In the last month it's become clear to me that the tech works. The science is done, and what's left is engineering.
There are a lot of risks and mitigations and theory to build, but it's all solvable. The tech isn't mature, but neither was the Internet 30 years ago. And we built transatlantic cables and ran new wires to everyone's house.
People I care about, engineers with 20 years of experience, are having mental health breakdowns, caused by Zitron's work. They insist the tech will never work, and avoid learning about it, becoming progressively more paranoid and isolated. I'm trying to be supportive and help them start to recover, but it's slow going.
If someone is having a crisis about this, I hope they start talking to a therapist. I don't need them to agree with me, but I do need them to not harm themselves.
hopefully they rope soon! :)
There's nothing to recover from, what are you even talking about? I'm not a token user (and I can't make predictions about the future and whether it will force me to use token but still). That the industry is collectively having a delusion about what constitutes good software (in all senses of the word - functionality and consequences for society) is clear to see, something I too fear we might never recover from, but I stand quite clearly on the side of people not of corporations hoping to extract more more more.
Ed Zitron, from what little I have heard of him, seems incredibly irrational. I don't think I've ever seen anybody stick their head deeper in the sand more than I've seen him do.
It's one thing to dislike or even detest something, but to constantly claim it is worthless and without use when people are already benefitting from it everyday is nothing short of delusion.