> All of the AI projects we have observed as a team are failing. Every single one – we have seen 0% success in a year and a half, not only amongst projects we have been asked to participate in, but even within projects that we have observed in passing while doing totally unrelated work.
That's got to be hyperbole, which blows out their credibility. They chose to say 'AI' rather than, for example, LLM, or Transformer model, or Diffusion model. This means they are including a huge swathe of things dating back to Expert Systems in their claim.
And who hasn't seen productivity gains from more established AI technology - at least things like semantic search? Who hasn't seen diffusion models generating content in roles that might have done the work by hand before? Who hasn't seen some kind of regression algorithm (even using linear regression in a supervised context counts as AI - so you can absolutely do AI even in tools like Excel) improve operation productivity?
Even if they narrowed it to the Transformer model LLMs which re-ignited recent public interest in AI, less ambitious projects to give them to engineering staff to automate easy but boring tasks in the background generally have been a success. More ambitious ones that are beyond what you'd reasonably expect the models to be able to do - for sure, those tend to fail. For most of these, the failure is predictable in advance, while some are at the boundary of what's possible, and so it is harder to predict (these are rationally genuine R&D projects).
If you read the footnote, they follow up to say they've rejected 100% of the AI projects brought to them.
Go to their home page and one of their consulting selling points is recovering struggling projects.
One of their front-page selling points is that they use "ancient techniques" from books written prior to the year 2000, because presumably everything newer than that is bad?
> For non-executive management who might be struggling to deliver things that feel beyond their control, we have ancient techniques (see: books written between 1986 and 1999) to turn your team into the envy of the organisation, and we can drop in directly to get your team the resources it needs to save a struggling project.
This is entirely a selection bias issue that they've created for themselves: Advertise a consulting service for saving failing projects to companies that don't have internal expertise to handle it, then write blog posts that 100% of the projects you see are failing. Also refuse to help them, to guarantee they can't be converted to successful projects to keep the success number at 0%.
it's not surprising as LLM's will generally fail you when you yourself don't know what you want. it's surprising how many people and organizations just blunder around scribbling code without clear goals in mind - for such people LLM's are a net negative as they will just end up with even more scribbles and no hope to make sense of it all.
There’s some selection bias involved especially based on their marking, but you’re misreading what they are saying.
> even within projects that we have observed in passing while doing totally unrelated work.
The kind of companies with failing projects seem to be very bad at using AI. That’s different from the normal mix of success and failures at most large companies.
Companies with projects that are doing well don't try to hire consultants who specialized in saving failing projects.
No company with a good AI strategy is going to go to the Hermit Tech website where text written in Ye Olde Font explains that they'll use ancient techniques from the 80s and 90s to make your software work and thinks, "These are the right people for our AI job!"
These people are trying to carve out a niche for themselves as being anti-modern, anti-AI, and being contrarian consultants that you can bring in when you want some external consultants to agree with you in a very specific way.
I think you are right on target. I’ve seen this pattern:
Company does x. X sucks, and is largely derided as being a bad idea, but it’s important-employee-bobs baby. No one wants to be the guy that tells bob that x is killing the company in some small or large way, because bob can get them fired or is protected by someone who can, so you hire consultants xyz who specialize in “transforming productivity through x management” and they come in and “transform” x into something less lovecraftian, or just explain why it’s bad in a way that makes bob look like a genius.
> Companies with projects that are doing well don't try to hire consultants who specialized in saving failing projects.
Large companies where every project is going well are rare.
I agree companies with good AI strategies are unlikely to use these people, but excluding failures is just as much a bias as excluding winners. AI lets you shoot your sold in the foot even faster shouldn’t be surprising, but it’s still something to consider.
Tale as old as time. Revolution has been merchandised. The action and the reaction, they all have suppliers, which often turn out to be the same entities supplying to both sides.
Yeah, and how and what someone jokes about will often tell you something about what and how they think. I don't think they'll literally just ignore everything after that point, but I do expect that they are going to tend towards an old-school approach.
The prominent "Business Owners" link on this company's website is broken. "Not Found".
Perhaps this will get downvoted but I personally take with a grain of salt anything written/stated by a company that can't even get the most basic functions (like running a simple website) correct.
And the even bigger irony here is that the author has a ranty blog post in which he claims he saved his employer $500,000 by clicking a button. "[It] is fucking wild that an inefficiency that took me five minutes to solve in a GUI configuration panel was allowed to persist," he wrote.
I don't think this is surprising at all. I'm pretty sure I've saved my employers similar amounts in the past, because what happens is something gets configured months or years ago, the dev leaves, everyone forgets and assumes things are how they are.
I've had to delete some really silly code that would slow things down or just force waits as a result of either dealing with legacy APIs or some other arcane reason. Not without testing or making sure it wasn't there for a reason mind you, but these inefficiencies can sometimes be hidden big problems.
The point isn't that this stuff happens. It's that a person who is trying to sell consulting services with ranty "this industry is a sham" and "it's a disgrace that it was even possible" blog posts has a website with just 5 major navigation links...and one of them is 404.
> And who hasn't seen productivity gains from more established AI technology
You have to be very careful about claiming productivity gains. There may have been some instances of gains in a specific part of a workflow but does it slow down others or result in overall gains is yet to be empirically measured and validated. We’re seeing metrics like more lines of code, better unit tests, documentation, faster PRs etc. but the actual gains of businesses are still a question mark. Do more PRs lead to faster features being shipped or does it lead to slower reviews or bug ridden code that breaks user experience? I’ve see a lot of companies tout their metrics around more code being shipped but the same companies aren’t talking about how that translates to an actual dollar amount.
These things have been "good" for a while now. And yet, companies like Amazon and Microsoft aren't showing significant improvement in their most visible products.
If it's not obviously showing for the all-in, AI-selling companies, I simply don't expect serious improvement for everyone else.
They're undeniably neat tools, but so far there's no observable evidence that they're transformative.
I think there's a big difference between individual employees using AI tools to boost their productivity - with things like Claude Code and Codex - and "AI projects" where companies build custom software on top of LLMs.
The former is easy to get right. Any software engineer (at least provided they aren't actively resisting the technology )can get useful results out of Claude Code these days.
The latter is really hard. LLMs are a strange beast to build software on, and most of the obvious projects - like the internal chatbots described in this article - are easy to have over-promise and under-deliver.
I get what you are saying, and while I'll say it is definitely not 0%, I have seen very little in the way of useful software that is primarily generated. The vast majority just does not go the distance for whatever reason. I could explain many reasons, but I am getting really tired of explaining myself. If the tools were as great as everyone says, we'd be going through a software Renaissance, but we're not. I would argue a software dark ages since it feels like things are getting worse and I find bugs in what was historically very long running and stable applications. But, whatever. I think the author is clearly talking about modern AI, I don't think they need to be explicit about models.
In general I find their submissions tend towards extreme grandiosity. I find I really appreciate people who have some nuance about the world, can see some duality, and the many many many submissions here are (I admit) often quite fun and enjoyable, but spoken much more from a bully pulpit perspective, with a zeal and self certainty that I find rarely coincides with truth-seeking.
This part toward the end of the article resonated with me:
> If you’re being asked to review huge volumes of terrible AI code, just assume that the organisation is going to burn you out and fire you. You will not convince the person drowning you in 2000 line PRs to stop. Start looking for a new job as if you have already been fired. I have seen this happen many times now
I suspect we will see this phenomenon more and more as organizations more widely adopt agentic development.
Their company does data projects. That plus context makes me think they’re talking about internal work process automation type of work, although it also seems like they’re talking about conversational interfaces (chatbots).
I completely buy the “emperor’s new clothes” argument for work process automation. I’m surprised they don’t address AI-assisted engineering, which seems to be going positively for a lot of folks (although I have doubts about its sustainability). I disagree about the success of chatbots, if the problem is narrowly-defined and chosen properly. My previous company built a conversational interface to a vector database and saw good results. (Although, arguably, the vector database was the real magic, and a traditional UI would have been faster and more accurate.)
In general, I think OP is more right than wrong, though, particularly about the AI mania and unrealistic expectations sweeping the C-suite.
Do yo have links handy for AI-assisted engineering going positively? The case I have on my mind of it going negatively is this recent Ford case [1] It's not that I believe it couldn't go positively, of course.
[1] Ford rehires human engineers after AI fails to match quality checks
That Ford story was really misleading. It wasn't about modern LLMs, and the way it was reported implied that Ford had fired and then hired people but if you read closer that wasn't necessarily the case at all - it sounded more like they were re-hiring people who had retired because they needed expertise that had left the company.
> Over the last three years, Ford says it has hired 350 veteran engineers, many of them former employees and others from suppliers, to help address seemingly intractable quality woes that have cost the automaker billions. [...]
> “We had been relying more and more on automated quality systems” and not getting the desired results, Galhotra said. “We brought back technical specialists” and “they hunt for failure points before a part ever reaches the plant floor.”
I think it's not clear from the article why they left (e.g. could be anything from retirement to going to work at another firm/contracting to being fired to switching careers), and likely it's going to be a mix, plus not all were previous Ford employees. Similarly the "AI" isn't clearly defined (but like you I would be surprised if it were LLMs). I suspect though why the article exists (and a possible source of your downvotes) is signalling against "AI", which if Ford wants more expert employees (given their issues), is something Ford wants to present.
As @simonw said, the Ford example isn’t a good one.
As for AI-assisted engineering going well, I think the jury is still out. Here on HN and with the engineers I know, you see people claiming multiples of productivity on coding tasks. But you also see people complaining about drowning in slop PRs.
I think there’s a lot of confounding factors to these reports. The type of work matters a lot: bug fixing good, prototyping good, big legacy codebases not so much, but maybe good for increased understanding. The type of automation matters: aggressive autocomplete good, vibe coding bad, dark factory (vibe coding with fancy harnesses and auto-“correcting” eval loops) questionable.
And then finally, the perennial mistake our industry makes, which is to value speed of creation over maintenance costs. Personally, I think this is where AI-assisted engineering is going to fall down really hard, but the jury’s still out on that one.
Anyway, there’s a really big spread in experiences with AI, that I think chalk up more to all this context rather than religion and belief. OP didn’t address it at all, which I think is a big gap in their essay, but I do think think they describe the executive-level mania pretty well.
> As for AI-assisted engineering going well, I think the jury is still out.
Anecdotally, AI-assisted engineering has helped me flesh out ideas or to learn extremely complicated APIs faster than trying to understand the docs (which usually are labyrinthine). MS COM ones, for example. I can go read the docs but it's easier to get a quick idea of what I need to do if I ask Claude to provide me an example of doing something specific with it, because MS's code samples (particularly their full ones in, say, the windows Desktop SDK repo) have always been annoying for me to wade through because I have to filter out a bunch of noise. I can't (and won't) try to guestimate "productivity" improvements though, but as an assistant AI has (somewhat) helped. I still do all the engineering work though. Along with it giving me tips on using more modern language features for languages like C++.
> I disagree about the success of chatbots, if the problem is narrowly-defined and chosen properly.
If you can narrow the problem down, then you could design a much better interface for it than a text box and free form text (unless that's the better solution).
As for as AI assisted engineering goes, the thing is that after some time with a project, you already have much of the workflows and routines nailed down as scripts and other various combinations of tooling. And unless it's spaghetti code, you will have various snippets you can copy from for new code. The one thing I've observed about AI projects is that there's often little technical design coherence about them. It's always a kitchen sink of technologies and practices.
> If you can narrow the problem down, then you could design a much better interface for it than a text box and free form text (unless that's the better solution).
Yes, I agree, in that the chatbot we built probably would have worked just as well with a traditional UI, and would have been done a lot faster. But it would have been a lot less sexy (actually important for the bottom line!) and there are future directions that could take advantage of the conversational interface that’s potentially better than a traditional UI.
On the down side, good chatbots are really frikkin difficult to write. These things (LLMs) are not reliable at scale. The basic functionality came together in weeks. Getting it to behave consistently and obey guardrails took months, and even then we had to accept a low level of failed conversations.
They run a consulting company called "Hermit Tech". Their websites has an olde style font.
They boast about "ancient techniques" from books written prior to the year 2000
> For non-executive management who might be struggling to deliver things that feel beyond their control, we have ancient techniques (see: books written between 1986 and 1999) to turn your team into the envy of the organisation, and we can drop in directly to get your team the resources it needs to save a struggling project.
So yeah, of course these people hate AI and everything about it.
No serious company is reaching out to these people for help with their AI project.
That has ever been the case. As soon as it works reliably, it's not "AI" any more. Take spellcheckers or collaborative filtering as examples, but there are lots more. Hofstadter in G.E.B. said it well:
> There is a related “Theorem” about progress in AI: once some mental function is programmed, people soon cease to consider it as an essential ingredient of “real thinking”. The ineluctable core of intelligence is always in that next thing which hasn’t yet been programmed. This “Theorem” was first proposed to me by Larry Tesler, so I call it Tesler’s Theorem: “AI is whatever hasn’t been done yet.”
Sure, or maybe the actual applications of “AI” are small and unobtrusive, like the dictation of doctor’s notes example, and it’s not actually the massive revolution it’s claimed to be.
In this moment, the opposite is happening. Everything is getting called "AI", whether it uses LLMs to prompt LLMs about how to prompt LLMs, uses "conventional" machine learning, or just looks mysterious enough that they can expect the market to not ask questions.
I am reminded of "game AI", which for the most part has historically been just giant decision trees, encoded one way or another, because if you hook up any sort of real AI to a game entity or collection of game entities that does any sort of learning or training, even simple 1980s-era reinforcement learning, it turns out the game entities will roflstomp the human players, and the human players aren't interested in paying for that experience. We've been calling those collections of if statements and for loops "AI" for a long time, though, because who wants to hear about how deliberately stupid their opponents are?
This post feels like its correct but also doesn't align with my own personal usage of claude for writing advanced sql and python code. I haven't with my own eyes seen an AI chatbot actually deliver a user experience that lets them query data with natural language BUT I have personally experienced writing extremely advanced queries using natural language and it is absolutely able to get close to (by my estimate) 80-90% of the way there.
There surely are companies out there using AI in such a way that is actually advancing them above and beyond their competition. They are probably quietly doing it rather than announcing it loudly.
The whole point of the article is about the big corps with hordes of management and people in them. My argument is that they have always been that way. Before AI it was "data science and analytics" or (as the author says) "blockchain".
Maybe maybe not. In large organisations for each person creating value with LLMs there are probably more destroying value. Bob used to only waste one person time with LLMs he can waste the entire organisations time.
Anyone can use an LLM make a bad ideas sound like a good idea. I imagine this will lead to insane amounts of productivity loss as the entire organisation ends up pivoting to follow the bad idea of a mediocre VP etc.
> Checking out a parallel copy of our Go repository and telling the AI to rewrite the whole thing in Zig while I work on something else just so I can keep my job.
> Was it just sales fluff? The answer was a lot more interesting.... Executives at their customers were saying absurd things about achieving 100x productivity, and this meant that if any executive at the vendor said that these gains were not plausible, it would undermine the credibility of the customer’s executive, be perceived as an attack (or heresy), and possibly result in an enterprise contract cancellation.
You are basically calling out the fact that the Emperor has no clothes. Many said this before. While it is a true statement, it is not going to help. Because, as you rightly said, it is a mania - like the tulip mania of 17th century or the manias of many forms today. The mania continues to evolve and flourish through it's peak and then go down. For that matter, there is hardly anything that is not a mania. Think of agile processes, timesheets, LoC based productivity, ...
The corporate mindset keeps going through different mania at different times. It could be initiated by some consulting gurus (processes), or some security nerds (strap yourself down until you can't move), or peer pressure (fear of missing out), or presentation goals (show that you are a AI-powered and modern company).
We can't remove or stop manias. Infact that is not the goal. The music should go on and the dance should go on. Everyone is in this dance - customers, businesses, supply chains, governments, thinkers and philosophers. It's a world-wide dance. So it's OK. The music track won't last forever. It will change and dance will change.
This mania can also be called "herd behavior" or "crowd psychology". There was that book a while ago about the intelligence of crowds, but far more common is the stupidity and insanity of crowds. Business and politics are mainly driven by it, but the tech industry routinely falls to hype, trends, and decades-long mistakes that seem obvious in hindsight.
I think AI is only part of the problem. The Multi-crisis ahead, makes it even less of a proposition to be in charge right now- and AI offers a "responsibility" cope out for already stressed to the limit human systems who have no solutions for the problems, because there often arent any.
Take climate change- you have torrential rainfalls, sweping away whole city-parts in mountanous regions, in some enormous russian roulett. And it doesnt even factor into building evaluations because then it would basically reduce the prevalent pension scheme to cinders.
You have dry months in europe now, where some thrown cigarett butt could ignite a firestorm- and the obvious solution is to remove the dangerous greenery from the burbs. Nobody does it though.
And that is just the plain sight visible layer of this shit cake. If i was some missguided fool into heroics and leadership and signed up for a little more then i could take and fake- i would long for some magic box that lowers the burden too. Those up there are human after all.
As someone who uses LLMS for coding, ideas validation and research, I think the article is biased against AI forctge wrong reasons.
If you know what you are looking for and know what “a solution looks like”, AI is amazing at distilling ideas. If you have no clue, the AI will return “clueless” solutions.
It is just like before the AI: there are people who know how to search the web, read and understand documentation and so on. And then there are peole who are incapable.
AI is naking the latter category fail incredibly fast. Really, nothing new under the sun: garbage in, garbage out.
Everyone knows the ivory towers are full of people who shouldnt be there. Its marvelous to see them crumble even if you arent in the chain gang below.
Nothing will improve until things get bad enough. You need enough greedy yes men doing quailty control on airplanes to escalate.
It took me decades to understand the use of and need for escalation in big organisations.
I didnt understand seemingly unproductive strict job descriptions either. Hilarious situations with 10 people doing nothing at all their entire shift (really nothing) while i had work todo on my own that really required 5 people. A few days later someone showed up to tell me the qualty was below average. LOL
Now i know i should do only half a shift worth of work. When they come complaint about it i say: very good, write it down, make the official report.
Then i hear nothing and a year later its two people with work for 5 scheduled. I tell them to slow down but we still do 2.5 shifts because they dont understand how escalation works.
The nummers now show we are 5 times as productive which isnt good for the company. The beurocracy is slow to adapt and all it has is numbers.
For many years i tried to do all of the work but that means nothing is wrong. The numbers say all is fine most of the time. Someone grinning at how much work i did isnt going to get recorded or processed.
If it looks like an unattended LLM can do a better job it means you dont know what you are talking about. If you fire everyone who noticed you might buy time but reality will catch up.
It reminds me of when they first put computers in trains in NL, they ran on windows 3.11 and no one trusted it to do anything. The solution was to give it all the data so that it could display a nice overview but it didnt control anything. Lots of trains drove around with a blue screen of death or a boot error. If there was a problem it was slightly harder to diagnose but it wouldnt drive if [say] a door was open. If the gui said a door was open you could just ignore it. On its own it means someone has to replace a sensor. If it also didnt move anymore the message is a real issue.
I imagine LLMs are wonderful for that kind of thing.
The author doesn't distinguish whether these projects failed because of technology and execution, or failed because product market-fit. They simply blame AI being involved.
AI will only help if you use rapid iteration to cheaply/quickly produce ideas. All the normal project failure modes still exist.. Blaming AI because AI is dumb.
I only read up to point 3 because the hyperbole and frothing fervour was overwhelming.
The AI mania has been really fascinating to witness because on the surface it’s surprising that so many otherwise intelligent people have fallen into it. But I suppose intelligence as a concept is multifaceted and doesn’t include wisdom. I also wonder if it has to do with personality, where the people whose personality best suits leadership roles are more susceptible to this psychosis. I think there is also an archetype of “nerd” who believes they are smarter than they are and has all sorts of surface beliefs about AI from sci-fi that makes them susceptible.
It’s also a bit of a dilemma; if your boss has AI mania, and you buying into it or not is the difference between being promoted, keeping your job, passed up or even fired, the rational course of action for self-preservation is to also buy into the mania.
You can't get a fad like this with a true believer rate of 0. Too unstable. You have to have some true believers in the mix to see this situation arise, preferably widely distributed throughout the population.
Something I think about quite often with HN is there are probably people on this forum that have friends / family who work at AI companies or are highly monetarily invested in AI themselves.
I don't think everyone on this forum is as objective as we'd like to think. I know I'm not. A large part of why I dislike AI is because I view it as making my life, personally, worse. It has completely fucked up the career I've been building for going on two decades
And all I hear is "adapt or die" from assholes who are chugging the AI koolaid by the bucketfull
It's a very interesting technology, so it leaves some shockwaves while it's impacting the world. Over time this will be absorbed and the insanity will die out, but it's quite transformative, just like cars and the internet were (to name a few).
I read utter stupidity like this post and remember why am I wasting my time reading the human slop that infest this board? Just talk to Claude and learn something new and useful.
I love this author's article about saving half a million dollars with a click from a while back. Nikhil, if you're reading this, I have a decent war story about a similar situation (I was lucky enough that there were two such things, so I saved a full million a year and was still denied a $15k raise) and so I was really entertained at your post from a few years back. There are actually a lot of lessons to be learned about corporate politics there, about how you can save someone a million a year in perpetuity and promise to do that again next year (which I could have!) and see them still refuse to pay you a single extra cent.
Broadly I don't think this is quite so true, quite such a mortal threat.
What I see are that there are a lot of extremely fake humans, who want and need cover. Who have absurd ridiculous (and often dastardly or sinister) plans. Who want to do things, a-priori. But could never get away with their actions, in any just clear reasoned normal rules of society.
And AI is this new circuit breaker. It's innovative permission to move ridiculously fast and break everything, right now. Take the perhaps old IBM slide and flip it upside down,
> A computer can never be held accountable.
> Therefore a computer must never make a management decision
The people "using" AI today to "make decisions" are using it because AI cannot be held accountable therefore that is the cover for their decisions.
This is is all such a resounding PKD nightmare, a reality bring invaded by Fake Humans. It was that was already, just gobs of nonsense, the worst liars spreading the most ridiculous memetic caltrap everywhere: Bullshit Asymmetry Principle weaponized against reason to ever higher degrees, Fox News terrormongering advanced and advanced, Hastert Rule obstructionist politics by wicked pedophile protectors and system ruiners and monsters. AI is a rapid accelerant for burning down reality, for propagating the disreality that the fake humans require for existence. Un-people truly from some other dimension, who've worked and worked to get away with their twisted anti- reality over us all.
AI can and does help with a lot of decision making, in good ways. It's an incredibly tool. It can comb through incredible amounts of data. But it's primary use in "decision making" seems to be in deflecting responsibility, in making hideous choices no human system could reasonably make. In concocting fabulations. Both of management design, and endless fuel nightmare disreality slop video to dislodge any last bits of real reality still clinging on (hello ai faked campaign videos!).
The "frothing excitement" here is the frothing excitement to destroy society, to be and bring out the most wicked brutal careless world that can be brought upon us, to raise up the Theil-istic/(Octavia) Butler-ian nightmare neofuedalim. It is to escape accountability, to give cover for sin savagery and sabotage.
(Regarding the article, I do think it's worth tempering ones read of this article by reading the authors previous work on AI. Which to me exposes their baises and in my view makes them so vastly unreliable & overdramatic a narrator as to be near worthless. Their other submissions are less greviously clearly full of it, but also tend towards ridiculous over-grandiosity. https://news.ycombinator.com/item?id=48002795 )
honestly just seems like selection bias right? 0% success rate with ai projects, really? based on some of their other about us material it seems like they probably only attract companies that aren't motivated to adopt ai successfully.
Honestly, this is close to my experience within my org. It is not 0%, but it is very close. It seems like you're just uncomfortable with the author's experience conflicting with yours.
Perhaps it is the wording. For my org, I think people are getting value out of AI use supplementally, but what I would see as "AI projects" are definitely dead in the water.
I read the entire post, and I think this company has no expertise at all. So I'd rather they just used AI writing instead.At least Frontier model AI doesn't make such overblown claims.
They proudly claim that every AI project they've observed over the past year and a half had a 0% success rate, and that they've rejected all AI implementation work. While this is evidence that the market is crazy, at its core, it's a painful confession that they have no engineering expertise to implement and control modern AI architectures like RAG, Agentic Workflow, and context window optimization to meet business requirements. I find it fascinating how they're packaging that. It's basically saying, 'We're behind the times.'
There are already products that have achieved results by using AI as part of their development process, yet lumping all different types of AI usage into a single failure category is not only inaccurate but also misleading.
Same goes for the Snowflake Cortex anecdote. Even a freelancer like me can explain technical limitations and distinguish between what's possible and what's not, especially when clients are eager.
There's no engineering analysis in this entire post about why AI fails. No mention of technical bottlenecks like vector DB retrieval quality degradation or prompt injection failures.
I've also worked on RAG for a specific company. For internal knowledge chatbots, it often fails depending on document collection rates and chunking. But none of that is mentioned.
So I understand that AI projects and related things are bad. But there's no analysis of why.
For example, regarding Snowflake, I'm not sure, but did they discuss accuracy in terms of what query set or what ground truth they were using? You're consultants, aren't you?
Honestly, I don't understand why people are excited about this. I'd rather they just used AI. TIt's not about whether human writing is good or bad. It's that this kind of writing feels like a deception of the reader.
When making overgeneralizations, there's a basic minimum standard required.
Saying that making token usage a KPI makes it hard for employees to report is just an 'obvious' fact that's already appeared in far too many essays. Wake up. You're 'consultants.' Consultants are supposed to provide metrics and directions, but all you're doing is shouting into an echo chamber and asking for agreement.
If a significant portion of corporate AI investments are shoddy, you could at least propose specific metrics like document collection rates or user evaluation scores using the very skills you claim to have. I really don't get it.
Just use AI. I wish the OP had used AI. Let me be realistic.
Clicking the footnote for the weird "All of the AI projects we have observed as a team are failing" is equal parts enlightening and confusing:
> We have rejected all AI implementation work. It is absolutely a gigantic bubble and we have minimized our exposure to it – every single one of our current contracts would be totally unaffected by OpenAI collapsing, save for perhaps some second-order effects such a recession causing a client to become unable to pay us. And there’s nothing we can do to insulate ourselves from that anyway.
Following the link to their company page goes to Hermit Tech, where the primary advertisements for their services are about helping failing projects and troubled teams.
So this is just one huge selection bias example? Start a consulting company for recovering struggling projects, then make claims like "100% of the projects we've seen are struggling"?
There's so much more in this blog post that feels like they're working hard to ignore anything that disagrees with their bubble. Building an AI data pipeline with evals such that you can swap between AI APIs is standard. It's actually part of doing a decent job because you need to select which model hits the right cost/performance tradeoffs and be in a position to pivot when that math changes. Harboring ideas that OpenAI is going to collapse and bring your projects down with it is the kind of talk you hear out of people who don't understand how AI projects work or that there's an ecosystem to it beyond a single company.
The latest projects I'm working on even include open weight models that can be run on reasonable local hardware as cost and performance benchmarks. Even if all of the AI providers collapsed at the same time and nobody offered any services (not going to happen) these projects can still continue on.
It's a very weird time in technology. You can have one foot in a world where people are adopting technologies and using them intelligently, then you can run into articles like this from people who have built their own little self-selecting bubble that confirms all of their ideas who can't even imagine that successful projects exist right now.
What do you expect from a person other than the generalizations they see from the experiences they have had...? What 'evidence' could possibly be given other than extensive anecdotes?
The article presents hypothesis as fact with insufficient science. I have no problem discussing speculation, but if the author wants to promote their claims to any more than that I would want to see more journalism.
> All of the AI projects we have observed as a team are failing. Every single one – we have seen 0% success in a year and a half, not only amongst projects we have been asked to participate in, but even within projects that we have observed in passing while doing totally unrelated work.
That's got to be hyperbole, which blows out their credibility. They chose to say 'AI' rather than, for example, LLM, or Transformer model, or Diffusion model. This means they are including a huge swathe of things dating back to Expert Systems in their claim.
And who hasn't seen productivity gains from more established AI technology - at least things like semantic search? Who hasn't seen diffusion models generating content in roles that might have done the work by hand before? Who hasn't seen some kind of regression algorithm (even using linear regression in a supervised context counts as AI - so you can absolutely do AI even in tools like Excel) improve operation productivity?
Even if they narrowed it to the Transformer model LLMs which re-ignited recent public interest in AI, less ambitious projects to give them to engineering staff to automate easy but boring tasks in the background generally have been a success. More ambitious ones that are beyond what you'd reasonably expect the models to be able to do - for sure, those tend to fail. For most of these, the failure is predictable in advance, while some are at the boundary of what's possible, and so it is harder to predict (these are rationally genuine R&D projects).
If you read the footnote, they follow up to say they've rejected 100% of the AI projects brought to them.
Go to their home page and one of their consulting selling points is recovering struggling projects.
One of their front-page selling points is that they use "ancient techniques" from books written prior to the year 2000, because presumably everything newer than that is bad?
> For non-executive management who might be struggling to deliver things that feel beyond their control, we have ancient techniques (see: books written between 1986 and 1999) to turn your team into the envy of the organisation, and we can drop in directly to get your team the resources it needs to save a struggling project.
This is entirely a selection bias issue that they've created for themselves: Advertise a consulting service for saving failing projects to companies that don't have internal expertise to handle it, then write blog posts that 100% of the projects you see are failing. Also refuse to help them, to guarantee they can't be converted to successful projects to keep the success number at 0%.
Supposedly 70% of software projects failed even before AI. So are AI projects better or worse than that?
Also like people have asked about the 70% failure rate, how do you define failure/success.
https://medium.com/@trienpont/why-do-over-70-of-software-pro...
it's not surprising as LLM's will generally fail you when you yourself don't know what you want. it's surprising how many people and organizations just blunder around scribbling code without clear goals in mind - for such people LLM's are a net negative as they will just end up with even more scribbles and no hope to make sense of it all.
Cheaper than consultants!
There’s some selection bias involved especially based on their marking, but you’re misreading what they are saying.
> even within projects that we have observed in passing while doing totally unrelated work.
The kind of companies with failing projects seem to be very bad at using AI. That’s different from the normal mix of success and failures at most large companies.
Companies with projects that are doing well don't try to hire consultants who specialized in saving failing projects.
No company with a good AI strategy is going to go to the Hermit Tech website where text written in Ye Olde Font explains that they'll use ancient techniques from the 80s and 90s to make your software work and thinks, "These are the right people for our AI job!"
These people are trying to carve out a niche for themselves as being anti-modern, anti-AI, and being contrarian consultants that you can bring in when you want some external consultants to agree with you in a very specific way.
I think you are right on target. I’ve seen this pattern:
Company does x. X sucks, and is largely derided as being a bad idea, but it’s important-employee-bobs baby. No one wants to be the guy that tells bob that x is killing the company in some small or large way, because bob can get them fired or is protected by someone who can, so you hire consultants xyz who specialize in “transforming productivity through x management” and they come in and “transform” x into something less lovecraftian, or just explain why it’s bad in a way that makes bob look like a genius.
> Companies with projects that are doing well don't try to hire consultants who specialized in saving failing projects.
Large companies where every project is going well are rare.
I agree companies with good AI strategies are unlikely to use these people, but excluding failures is just as much a bias as excluding winners. AI lets you shoot your sold in the foot even faster shouldn’t be surprising, but it’s still something to consider.
Christ they are profiting off of reaction, they're like Trump but even worse since at least Trump is not stupid enough to be anti-technology.
Tale as old as time. Revolution has been merchandised. The action and the reaction, they all have suppliers, which often turn out to be the same entities supplying to both sides.
Yeah, the "ancient techniques" thing is a joke.
Yeah, and how and what someone jokes about will often tell you something about what and how they think. I don't think they'll literally just ignore everything after that point, but I do expect that they are going to tend towards an old-school approach.
The prominent "Business Owners" link on this company's website is broken. "Not Found".
Perhaps this will get downvoted but I personally take with a grain of salt anything written/stated by a company that can't even get the most basic functions (like running a simple website) correct.
And the even bigger irony here is that the author has a ranty blog post in which he claims he saved his employer $500,000 by clicking a button. "[It] is fucking wild that an inefficiency that took me five minutes to solve in a GUI configuration panel was allowed to persist," he wrote.
https://ludic.mataroa.blog/blog/i-accidentally-saved-half-a-...
This is an interesting strategy for trying to impress potential customers.
Most consultants try to impress you by talking about the great companies they've worked for.
This blog post screams, "Look how dumb my past coworkers were!" from top to bottom, then expects us to be impressed with their experience?
I don't think this is surprising at all. I'm pretty sure I've saved my employers similar amounts in the past, because what happens is something gets configured months or years ago, the dev leaves, everyone forgets and assumes things are how they are.
I've had to delete some really silly code that would slow things down or just force waits as a result of either dealing with legacy APIs or some other arcane reason. Not without testing or making sure it wasn't there for a reason mind you, but these inefficiencies can sometimes be hidden big problems.
The point isn't that this stuff happens. It's that a person who is trying to sell consulting services with ranty "this industry is a sham" and "it's a disgrace that it was even possible" blog posts has a website with just 5 major navigation links...and one of them is 404.
> And who hasn't seen productivity gains from more established AI technology
You have to be very careful about claiming productivity gains. There may have been some instances of gains in a specific part of a workflow but does it slow down others or result in overall gains is yet to be empirically measured and validated. We’re seeing metrics like more lines of code, better unit tests, documentation, faster PRs etc. but the actual gains of businesses are still a question mark. Do more PRs lead to faster features being shipped or does it lead to slower reviews or bug ridden code that breaks user experience? I’ve see a lot of companies tout their metrics around more code being shipped but the same companies aren’t talking about how that translates to an actual dollar amount.
These things have been "good" for a while now. And yet, companies like Amazon and Microsoft aren't showing significant improvement in their most visible products.
If it's not obviously showing for the all-in, AI-selling companies, I simply don't expect serious improvement for everyone else.
They're undeniably neat tools, but so far there's no observable evidence that they're transformative.
I think there's a big difference between individual employees using AI tools to boost their productivity - with things like Claude Code and Codex - and "AI projects" where companies build custom software on top of LLMs.
The former is easy to get right. Any software engineer (at least provided they aren't actively resisting the technology )can get useful results out of Claude Code these days.
The latter is really hard. LLMs are a strange beast to build software on, and most of the obvious projects - like the internal chatbots described in this article - are easy to have over-promise and under-deliver.
>semantic search
I'm doing fundraising for my tf-idf startup. It's named after a very big number!
I get what you are saying, and while I'll say it is definitely not 0%, I have seen very little in the way of useful software that is primarily generated. The vast majority just does not go the distance for whatever reason. I could explain many reasons, but I am getting really tired of explaining myself. If the tools were as great as everyone says, we'd be going through a software Renaissance, but we're not. I would argue a software dark ages since it feels like things are getting worse and I find bugs in what was historically very long running and stable applications. But, whatever. I think the author is clearly talking about modern AI, I don't think they need to be explicit about models.
Their previous article on AI shows a pretty strident ham-handedness. https://news.ycombinator.com/item?id=48002795
In general I find their submissions tend towards extreme grandiosity. I find I really appreciate people who have some nuance about the world, can see some duality, and the many many many submissions here are (I admit) often quite fun and enjoyable, but spoken much more from a bully pulpit perspective, with a zeal and self certainty that I find rarely coincides with truth-seeking.
I would love to learn about and from the details of their projects.
zero percent of statistics online are made up!
[citation needed]
https://hermit-tech.com/business-owners
This part toward the end of the article resonated with me:
> If you’re being asked to review huge volumes of terrible AI code, just assume that the organisation is going to burn you out and fire you. You will not convince the person drowning you in 2000 line PRs to stop. Start looking for a new job as if you have already been fired. I have seen this happen many times now
I suspect we will see this phenomenon more and more as organizations more widely adopt agentic development.
I'm in this quote and I don't like it.
> All of the AI projects we have observed as a team are failing. Every single one – we have seen 0% success in a year and a half,
What is an "AI project"? The post doesn't define it.
Is it writing some software from scratch? Using an LLM chatbot by non-coders, either internally or externally? Or something else entirely?
Some examples would really help.
Their company does data projects. That plus context makes me think they’re talking about internal work process automation type of work, although it also seems like they’re talking about conversational interfaces (chatbots).
I completely buy the “emperor’s new clothes” argument for work process automation. I’m surprised they don’t address AI-assisted engineering, which seems to be going positively for a lot of folks (although I have doubts about its sustainability). I disagree about the success of chatbots, if the problem is narrowly-defined and chosen properly. My previous company built a conversational interface to a vector database and saw good results. (Although, arguably, the vector database was the real magic, and a traditional UI would have been faster and more accurate.)
In general, I think OP is more right than wrong, though, particularly about the AI mania and unrealistic expectations sweeping the C-suite.
Do yo have links handy for AI-assisted engineering going positively? The case I have on my mind of it going negatively is this recent Ford case [1] It's not that I believe it couldn't go positively, of course.
[1] Ford rehires human engineers after AI fails to match quality checks
https://www.bbc.com/news/articles/cgrkd41n2v9o
That Ford story was really misleading. It wasn't about modern LLMs, and the way it was reported implied that Ford had fired and then hired people but if you read closer that wasn't necessarily the case at all - it sounded more like they were re-hiring people who had retired because they needed expertise that had left the company.
You need to get through the Bloomberg paywall: https://www.bloomberg.com/news/articles/2026-06-25/ford-has-...
> Over the last three years, Ford says it has hired 350 veteran engineers, many of them former employees and others from suppliers, to help address seemingly intractable quality woes that have cost the automaker billions. [...]
> “We had been relying more and more on automated quality systems” and not getting the desired results, Galhotra said. “We brought back technical specialists” and “they hunt for failure points before a part ever reaches the plant floor.”
(I made these points on the HN thread about it 3 weeks ago and got voted down and I'm still salty about it https://news.ycombinator.com/item?id=48674446#48675045 )
I think it's not clear from the article why they left (e.g. could be anything from retirement to going to work at another firm/contracting to being fired to switching careers), and likely it's going to be a mix, plus not all were previous Ford employees. Similarly the "AI" isn't clearly defined (but like you I would be surprised if it were LLMs). I suspect though why the article exists (and a possible source of your downvotes) is signalling against "AI", which if Ford wants more expert employees (given their issues), is something Ford wants to present.
As @simonw said, the Ford example isn’t a good one.
As for AI-assisted engineering going well, I think the jury is still out. Here on HN and with the engineers I know, you see people claiming multiples of productivity on coding tasks. But you also see people complaining about drowning in slop PRs.
I think there’s a lot of confounding factors to these reports. The type of work matters a lot: bug fixing good, prototyping good, big legacy codebases not so much, but maybe good for increased understanding. The type of automation matters: aggressive autocomplete good, vibe coding bad, dark factory (vibe coding with fancy harnesses and auto-“correcting” eval loops) questionable.
And then finally, the perennial mistake our industry makes, which is to value speed of creation over maintenance costs. Personally, I think this is where AI-assisted engineering is going to fall down really hard, but the jury’s still out on that one.
Anyway, there’s a really big spread in experiences with AI, that I think chalk up more to all this context rather than religion and belief. OP didn’t address it at all, which I think is a big gap in their essay, but I do think think they describe the executive-level mania pretty well.
> As for AI-assisted engineering going well, I think the jury is still out.
Anecdotally, AI-assisted engineering has helped me flesh out ideas or to learn extremely complicated APIs faster than trying to understand the docs (which usually are labyrinthine). MS COM ones, for example. I can go read the docs but it's easier to get a quick idea of what I need to do if I ask Claude to provide me an example of doing something specific with it, because MS's code samples (particularly their full ones in, say, the windows Desktop SDK repo) have always been annoying for me to wade through because I have to filter out a bunch of noise. I can't (and won't) try to guestimate "productivity" improvements though, but as an assistant AI has (somewhat) helped. I still do all the engineering work though. Along with it giving me tips on using more modern language features for languages like C++.
The Ford case is not about AI coding. It's about computer vision processes that went wrong. This was less about AI and more about Ford being dumb.
> I disagree about the success of chatbots, if the problem is narrowly-defined and chosen properly.
If you can narrow the problem down, then you could design a much better interface for it than a text box and free form text (unless that's the better solution).
As for as AI assisted engineering goes, the thing is that after some time with a project, you already have much of the workflows and routines nailed down as scripts and other various combinations of tooling. And unless it's spaghetti code, you will have various snippets you can copy from for new code. The one thing I've observed about AI projects is that there's often little technical design coherence about them. It's always a kitchen sink of technologies and practices.
> If you can narrow the problem down, then you could design a much better interface for it than a text box and free form text (unless that's the better solution).
Yes, I agree, in that the chatbot we built probably would have worked just as well with a traditional UI, and would have been done a lot faster. But it would have been a lot less sexy (actually important for the bottom line!) and there are future directions that could take advantage of the conversational interface that’s potentially better than a traditional UI.
On the down side, good chatbots are really frikkin difficult to write. These things (LLMs) are not reliable at scale. The basic functionality came together in weeks. Getting it to behave consistently and obey guardrails took months, and even then we had to accept a low level of failed conversations.
They run a consulting company called "Hermit Tech". Their websites has an olde style font.
They boast about "ancient techniques" from books written prior to the year 2000
> For non-executive management who might be struggling to deliver things that feel beyond their control, we have ancient techniques (see: books written between 1986 and 1999) to turn your team into the envy of the organisation, and we can drop in directly to get your team the resources it needs to save a struggling project.
So yeah, of course these people hate AI and everything about it.
No serious company is reaching out to these people for help with their AI project.
i feel like the useful applications of AI get silently integrated into workflows.
the ill-conceived moonshots by and for a non-technical audience get labelled as "AI projects/initiatives" and they fail.
That has ever been the case. As soon as it works reliably, it's not "AI" any more. Take spellcheckers or collaborative filtering as examples, but there are lots more. Hofstadter in G.E.B. said it well:
> There is a related “Theorem” about progress in AI: once some mental function is programmed, people soon cease to consider it as an essential ingredient of “real thinking”. The ineluctable core of intelligence is always in that next thing which hasn’t yet been programmed. This “Theorem” was first proposed to me by Larry Tesler, so I call it Tesler’s Theorem: “AI is whatever hasn’t been done yet.”
Sure, or maybe the actual applications of “AI” are small and unobtrusive, like the dictation of doctor’s notes example, and it’s not actually the massive revolution it’s claimed to be.
In this moment, the opposite is happening. Everything is getting called "AI", whether it uses LLMs to prompt LLMs about how to prompt LLMs, uses "conventional" machine learning, or just looks mysterious enough that they can expect the market to not ask questions.
I am reminded of "game AI", which for the most part has historically been just giant decision trees, encoded one way or another, because if you hook up any sort of real AI to a game entity or collection of game entities that does any sort of learning or training, even simple 1980s-era reinforcement learning, it turns out the game entities will roflstomp the human players, and the human players aren't interested in paying for that experience. We've been calling those collections of if statements and for loops "AI" for a long time, though, because who wants to hear about how deliberately stupid their opponents are?
This post feels like its correct but also doesn't align with my own personal usage of claude for writing advanced sql and python code. I haven't with my own eyes seen an AI chatbot actually deliver a user experience that lets them query data with natural language BUT I have personally experienced writing extremely advanced queries using natural language and it is absolutely able to get close to (by my estimate) 80-90% of the way there.
There surely are companies out there using AI in such a way that is actually advancing them above and beyond their competition. They are probably quietly doing it rather than announcing it loudly.
The whole point of the article is about the big corps with hordes of management and people in them. My argument is that they have always been that way. Before AI it was "data science and analytics" or (as the author says) "blockchain".
Maybe maybe not. In large organisations for each person creating value with LLMs there are probably more destroying value. Bob used to only waste one person time with LLMs he can waste the entire organisations time.
Anyone can use an LLM make a bad ideas sound like a good idea. I imagine this will lead to insane amounts of productivity loss as the entire organisation ends up pivoting to follow the bad idea of a mediocre VP etc.
> Checking out a parallel copy of our Go repository and telling the AI to rewrite the whole thing in Zig while I work on something else just so I can keep my job.
> Was it just sales fluff? The answer was a lot more interesting.... Executives at their customers were saying absurd things about achieving 100x productivity, and this meant that if any executive at the vendor said that these gains were not plausible, it would undermine the credibility of the customer’s executive, be perceived as an attack (or heresy), and possibly result in an enterprise contract cancellation.
A lot of excellent anecdotes here.
Dupe:
https://news.ycombinator.com/item?id=48955929
https://news.ycombinator.com/item?id=48956153
Incredible skill issue selecting for profit instead of fun. Why does everyone need to make money?? What about like... shareware? Remember shareware?
Small discussion yesterday (43 points, 7 comments) https://news.ycombinator.com/item?id=48956153
You are basically calling out the fact that the Emperor has no clothes. Many said this before. While it is a true statement, it is not going to help. Because, as you rightly said, it is a mania - like the tulip mania of 17th century or the manias of many forms today. The mania continues to evolve and flourish through it's peak and then go down. For that matter, there is hardly anything that is not a mania. Think of agile processes, timesheets, LoC based productivity, ...
The corporate mindset keeps going through different mania at different times. It could be initiated by some consulting gurus (processes), or some security nerds (strap yourself down until you can't move), or peer pressure (fear of missing out), or presentation goals (show that you are a AI-powered and modern company).
We can't remove or stop manias. Infact that is not the goal. The music should go on and the dance should go on. Everyone is in this dance - customers, businesses, supply chains, governments, thinkers and philosophers. It's a world-wide dance. So it's OK. The music track won't last forever. It will change and dance will change.
This mania can also be called "herd behavior" or "crowd psychology". There was that book a while ago about the intelligence of crowds, but far more common is the stupidity and insanity of crowds. Business and politics are mainly driven by it, but the tech industry routinely falls to hype, trends, and decades-long mistakes that seem obvious in hindsight.
I think AI is only part of the problem. The Multi-crisis ahead, makes it even less of a proposition to be in charge right now- and AI offers a "responsibility" cope out for already stressed to the limit human systems who have no solutions for the problems, because there often arent any.
Take climate change- you have torrential rainfalls, sweping away whole city-parts in mountanous regions, in some enormous russian roulett. And it doesnt even factor into building evaluations because then it would basically reduce the prevalent pension scheme to cinders.
You have dry months in europe now, where some thrown cigarett butt could ignite a firestorm- and the obvious solution is to remove the dangerous greenery from the burbs. Nobody does it though.
And that is just the plain sight visible layer of this shit cake. If i was some missguided fool into heroics and leadership and signed up for a little more then i could take and fake- i would long for some magic box that lowers the burden too. Those up there are human after all.
Dupe https://news.ycombinator.com/item?id=48956153
As someone who uses LLMS for coding, ideas validation and research, I think the article is biased against AI forctge wrong reasons.
If you know what you are looking for and know what “a solution looks like”, AI is amazing at distilling ideas. If you have no clue, the AI will return “clueless” solutions.
It is just like before the AI: there are people who know how to search the web, read and understand documentation and so on. And then there are peole who are incapable.
AI is naking the latter category fail incredibly fast. Really, nothing new under the sun: garbage in, garbage out.
Everyone knows the ivory towers are full of people who shouldnt be there. Its marvelous to see them crumble even if you arent in the chain gang below.
Nothing will improve until things get bad enough. You need enough greedy yes men doing quailty control on airplanes to escalate.
It took me decades to understand the use of and need for escalation in big organisations.
I didnt understand seemingly unproductive strict job descriptions either. Hilarious situations with 10 people doing nothing at all their entire shift (really nothing) while i had work todo on my own that really required 5 people. A few days later someone showed up to tell me the qualty was below average. LOL
Now i know i should do only half a shift worth of work. When they come complaint about it i say: very good, write it down, make the official report.
Then i hear nothing and a year later its two people with work for 5 scheduled. I tell them to slow down but we still do 2.5 shifts because they dont understand how escalation works.
The nummers now show we are 5 times as productive which isnt good for the company. The beurocracy is slow to adapt and all it has is numbers.
For many years i tried to do all of the work but that means nothing is wrong. The numbers say all is fine most of the time. Someone grinning at how much work i did isnt going to get recorded or processed.
If it looks like an unattended LLM can do a better job it means you dont know what you are talking about. If you fire everyone who noticed you might buy time but reality will catch up.
It reminds me of when they first put computers in trains in NL, they ran on windows 3.11 and no one trusted it to do anything. The solution was to give it all the data so that it could display a nice overview but it didnt control anything. Lots of trains drove around with a blue screen of death or a boot error. If there was a problem it was slightly harder to diagnose but it wouldnt drive if [say] a door was open. If the gui said a door was open you could just ignore it. On its own it means someone has to replace a sensor. If it also didnt move anymore the message is a real issue.
I imagine LLMs are wonderful for that kind of thing.
Seems like this consultant needs a consultant to help them adapt to modern technology.
The author doesn't distinguish whether these projects failed because of technology and execution, or failed because product market-fit. They simply blame AI being involved.
AI will only help if you use rapid iteration to cheaply/quickly produce ideas. All the normal project failure modes still exist.. Blaming AI because AI is dumb.
I only read up to point 3 because the hyperbole and frothing fervour was overwhelming.
The AI mania has been really fascinating to witness because on the surface it’s surprising that so many otherwise intelligent people have fallen into it. But I suppose intelligence as a concept is multifaceted and doesn’t include wisdom. I also wonder if it has to do with personality, where the people whose personality best suits leadership roles are more susceptible to this psychosis. I think there is also an archetype of “nerd” who believes they are smarter than they are and has all sorts of surface beliefs about AI from sci-fi that makes them susceptible.
It’s also a bit of a dilemma; if your boss has AI mania, and you buying into it or not is the difference between being promoted, keeping your job, passed up or even fired, the rational course of action for self-preservation is to also buy into the mania.
Yeah but that doesn't explain the people on anonymous forums talking like they have a brainworm.
You can't get a fad like this with a true believer rate of 0. Too unstable. You have to have some true believers in the mix to see this situation arise, preferably widely distributed throughout the population.
Something I think about quite often with HN is there are probably people on this forum that have friends / family who work at AI companies or are highly monetarily invested in AI themselves.
I don't think everyone on this forum is as objective as we'd like to think. I know I'm not. A large part of why I dislike AI is because I view it as making my life, personally, worse. It has completely fucked up the career I've been building for going on two decades
And all I hear is "adapt or die" from assholes who are chugging the AI koolaid by the bucketfull
It's a very interesting technology, so it leaves some shockwaves while it's impacting the world. Over time this will be absorbed and the insanity will die out, but it's quite transformative, just like cars and the internet were (to name a few).
I read utter stupidity like this post and remember why am I wasting my time reading the human slop that infest this board? Just talk to Claude and learn something new and useful.
I love this author's article about saving half a million dollars with a click from a while back. Nikhil, if you're reading this, I have a decent war story about a similar situation (I was lucky enough that there were two such things, so I saved a full million a year and was still denied a $15k raise) and so I was really entertained at your post from a few years back. There are actually a lot of lessons to be learned about corporate politics there, about how you can save someone a million a year in perpetuity and promise to do that again next year (which I could have!) and see them still refuse to pay you a single extra cent.
Ai Mania is making evangelists and non-believers alike sound like they are taking sides in a cult.
Broadly I don't think this is quite so true, quite such a mortal threat.
What I see are that there are a lot of extremely fake humans, who want and need cover. Who have absurd ridiculous (and often dastardly or sinister) plans. Who want to do things, a-priori. But could never get away with their actions, in any just clear reasoned normal rules of society.
And AI is this new circuit breaker. It's innovative permission to move ridiculously fast and break everything, right now. Take the perhaps old IBM slide and flip it upside down,
> A computer can never be held accountable.
> Therefore a computer must never make a management decision
https://simonwillison.net/2025/Feb/3/a-computer-can-never-be...
The people "using" AI today to "make decisions" are using it because AI cannot be held accountable therefore that is the cover for their decisions.
This is is all such a resounding PKD nightmare, a reality bring invaded by Fake Humans. It was that was already, just gobs of nonsense, the worst liars spreading the most ridiculous memetic caltrap everywhere: Bullshit Asymmetry Principle weaponized against reason to ever higher degrees, Fox News terrormongering advanced and advanced, Hastert Rule obstructionist politics by wicked pedophile protectors and system ruiners and monsters. AI is a rapid accelerant for burning down reality, for propagating the disreality that the fake humans require for existence. Un-people truly from some other dimension, who've worked and worked to get away with their twisted anti- reality over us all.
AI can and does help with a lot of decision making, in good ways. It's an incredibly tool. It can comb through incredible amounts of data. But it's primary use in "decision making" seems to be in deflecting responsibility, in making hideous choices no human system could reasonably make. In concocting fabulations. Both of management design, and endless fuel nightmare disreality slop video to dislodge any last bits of real reality still clinging on (hello ai faked campaign videos!).
The "frothing excitement" here is the frothing excitement to destroy society, to be and bring out the most wicked brutal careless world that can be brought upon us, to raise up the Theil-istic/(Octavia) Butler-ian nightmare neofuedalim. It is to escape accountability, to give cover for sin savagery and sabotage.
(Regarding the article, I do think it's worth tempering ones read of this article by reading the authors previous work on AI. Which to me exposes their baises and in my view makes them so vastly unreliable & overdramatic a narrator as to be near worthless. Their other submissions are less greviously clearly full of it, but also tend towards ridiculous over-grandiosity. https://news.ycombinator.com/item?id=48002795 )
honestly just seems like selection bias right? 0% success rate with ai projects, really? based on some of their other about us material it seems like they probably only attract companies that aren't motivated to adopt ai successfully.
If you read the article, the go over a case where they introduce AI into their pitch and were met with results they were not comfortable with.
Fair point as far as there being low success rate in some ways and over-enthusiasm. But 0% success is a dishonest exaggeration.
He's ramped his own AI spite up to a manic level.
Honestly, this is close to my experience within my org. It is not 0%, but it is very close. It seems like you're just uncomfortable with the author's experience conflicting with yours.
Perhaps it is the wording. For my org, I think people are getting value out of AI use supplementally, but what I would see as "AI projects" are definitely dead in the water.
I read the entire post, and I think this company has no expertise at all. So I'd rather they just used AI writing instead.At least Frontier model AI doesn't make such overblown claims.
They proudly claim that every AI project they've observed over the past year and a half had a 0% success rate, and that they've rejected all AI implementation work. While this is evidence that the market is crazy, at its core, it's a painful confession that they have no engineering expertise to implement and control modern AI architectures like RAG, Agentic Workflow, and context window optimization to meet business requirements. I find it fascinating how they're packaging that. It's basically saying, 'We're behind the times.'
There are already products that have achieved results by using AI as part of their development process, yet lumping all different types of AI usage into a single failure category is not only inaccurate but also misleading.
Same goes for the Snowflake Cortex anecdote. Even a freelancer like me can explain technical limitations and distinguish between what's possible and what's not, especially when clients are eager.
There's no engineering analysis in this entire post about why AI fails. No mention of technical bottlenecks like vector DB retrieval quality degradation or prompt injection failures.
I've also worked on RAG for a specific company. For internal knowledge chatbots, it often fails depending on document collection rates and chunking. But none of that is mentioned.
So I understand that AI projects and related things are bad. But there's no analysis of why.
For example, regarding Snowflake, I'm not sure, but did they discuss accuracy in terms of what query set or what ground truth they were using? You're consultants, aren't you?
Honestly, I don't understand why people are excited about this. I'd rather they just used AI. TIt's not about whether human writing is good or bad. It's that this kind of writing feels like a deception of the reader.
When making overgeneralizations, there's a basic minimum standard required.
Saying that making token usage a KPI makes it hard for employees to report is just an 'obvious' fact that's already appeared in far too many essays. Wake up. You're 'consultants.' Consultants are supposed to provide metrics and directions, but all you're doing is shouting into an echo chamber and asking for agreement.
If a significant portion of corporate AI investments are shoddy, you could at least propose specific metrics like document collection rates or user evaluation scores using the very skills you claim to have. I really don't get it.
Just use AI. I wish the OP had used AI. Let me be realistic.
Clicking the footnote for the weird "All of the AI projects we have observed as a team are failing" is equal parts enlightening and confusing:
> We have rejected all AI implementation work. It is absolutely a gigantic bubble and we have minimized our exposure to it – every single one of our current contracts would be totally unaffected by OpenAI collapsing, save for perhaps some second-order effects such a recession causing a client to become unable to pay us. And there’s nothing we can do to insulate ourselves from that anyway.
Following the link to their company page goes to Hermit Tech, where the primary advertisements for their services are about helping failing projects and troubled teams.
So this is just one huge selection bias example? Start a consulting company for recovering struggling projects, then make claims like "100% of the projects we've seen are struggling"?
There's so much more in this blog post that feels like they're working hard to ignore anything that disagrees with their bubble. Building an AI data pipeline with evals such that you can swap between AI APIs is standard. It's actually part of doing a decent job because you need to select which model hits the right cost/performance tradeoffs and be in a position to pivot when that math changes. Harboring ideas that OpenAI is going to collapse and bring your projects down with it is the kind of talk you hear out of people who don't understand how AI projects work or that there's an ecosystem to it beyond a single company.
The latest projects I'm working on even include open weight models that can be run on reasonable local hardware as cost and performance benchmarks. Even if all of the AI providers collapsed at the same time and nobody offered any services (not going to happen) these projects can still continue on.
It's a very weird time in technology. You can have one foot in a world where people are adopting technologies and using them intelligently, then you can run into articles like this from people who have built their own little self-selecting bubble that confirms all of their ideas who can't even imagine that successful projects exist right now.
Interesting how op describes his own experience and then assume that every other company around the globe experience exactly the same.
He generalizes CEO's behavior but provides no evidence. Cool.
What do you expect from a person other than the generalizations they see from the experiences they have had...? What 'evidence' could possibly be given other than extensive anecdotes?
The article presents hypothesis as fact with insufficient science. I have no problem discussing speculation, but if the author wants to promote their claims to any more than that I would want to see more journalism.