Zitron is far more negative when it comes to the utility of LLMs than I am, but dear God would it be satisfying to watch him be proven right on the financials.
His as-of-late frequent appearances on Bloomberg, CNBC, and Scott Galloway's podcast, as well as his collaborations with Ed Elson from Galloway's podcast must feel like a great vindication for him already. Quite impressive for a person who is self-taught in that area, just over the past two years.
edit: And, of course, the post is flagged. We must bury our heads in the sand again.
It’s far from vindication. All these TV networks have slots dedicated for extreme doomers to come in and say “the sky is falling! There’s going to be a recession soon. Sell everything!!” Been seeing it every day for 15 years now.
Zitron is really really good. He taught me the strategy of "screenshot all the wild AI-bro claims" to keep track of them and call them out when it all collapses. The comments here villifying him are wrong and unhinged. Sometimes you need to be disliked a little to correctly predict the future.
>"screenshot all the wild AI-bro claims" to keep track of them and call them out when it all collapses.
But have you done the same for Zitron? Because the AI-Bros have consistently been more accurate and Zitron consistently pretty wrong.
Zitron proclaimed in (early) 2024 that LLMs were stuck because there was no more data to use, and synthetic data ruined models, so we were already at the brink of collapse.
Zitron predictions regarding LLMs, as technology, are meaningless. Literally nobody can predict the future, and he's not a computer scientist. What matters is his thesis for things that have actual numbers that can be verified. Look at the numbers and his claims, and decide for yourself if the thesis is valid
He sells outrage to ill informed people who don't know any better. There's real issues, but it's far more nuanced than Zitron seems willing to bring to his audience.
I welcome AI skepticism but Ed’s often does seem to be done in bad faith. I feel like his claims border on, “AI has no use case whatsoever.” Which clearly isn’t true in 2026.
I don’t think that’s Ed’s position. As I understand it, Ed’s position is that it has no real economic value.
AI is a tool, a useful tool, an innovative tool, but technology has introduced innovative tool after innovative tool for years. Apache, Twitter’s Bootstrap, Pull to Refresh, React, Kubernetes, all innovative, all delivered real business value. Nobody invested hundreds of billions of capital into infrastructure for React.
AI has endless use cases. Theres no evidence that these use cases deliver the value necessary to justify any capital expenditure. A technology can change the world and have no standalone business case.
Business is boring. Money in, money out. We are still waiting for Anthropic and OpenAI to show how LLMs have moved the needle on business.
Every single thing that an LLM can be used to deliver value for a business today can be achieved by a cheap to run open weight model that companies could host in house. Where does that leave the trillion+ worth of dollars being buried in the ground?
Another nuance: his position is that real economic value hasn't been demonstrated, and that due to how LLMs work, it cannot really be measured. Meaning the AI crowd actual believes are based on vibes and doesn't seem to match reality
(also, of course, that there is a massive, insane misallocation of resources to the AI sector compared to the level of demand)
Can you provide a link to where he has said that? I’m not an expert on Ed but I don’t recall ever reading him say that it is functionally useless, he doesn’t comment on its function, only the business, whether it has functional use is immaterial unless that generates value for the companies spending obscene money on hosting. The companies implementing AI as consumers can get a huge amount of value from it without a single cent going to any of the big labs, that’s part of why investment in AI infrastructure is so misplaced.
He often says in passing that he doesn't see value in LLMs outside of generating bad software and can help sometimes with search. But it's not really part of the AI bubble thesis itself IMHO, just him adding his personal commentary in interviews
That's just his own personal belief about the technology, the only relevant part of this is that he doesn't believe there will be enough demand to match the level of expanses. It doesn't matter if AI is useful or not for his arguments. Does the demand exist or is it manufactured by hyperscalers or other actors? Who cares about his take on the future of software
I have both read his blog and listened to his podcasts. Not all of it (the articles are basically books), but enough to understand his positions. The "ai is mostly useless" is very much his position, he will say it again and again and again.
This part:
> I don’t think that’s Ed’s position. As I understand it, Ed’s position is that it has no real economic value.
Is not true. I can agree with one of his positions and disagree with the other, but it really is fair to say he repeatedly and confidently states that ai is useless.
Stock market might dip 20%+, because OpenAI is a large buyer of GPUs, RAMs, servers and disks, every chip company will be impacted from removing huge customer from their revenue.
then flywheel effect might kick off and impact carriers and construction companies building data centers and delivering things there.
which impacts investor trust and they might start pulling their money from AI companies, which in turn, they either need to increase prices or downsize their efforts
I find it hard to justify any conclusion other than we're about to repeat the crash of 1929 in terms of stock prices. Back then, stocks dropped 90% over 3 years. We'll be lucky if 10% actual value remains after the bonfire that's about to happen.
That's up to you, but if there is a crash it's very unlikely to be before a major hyperscaler decides to reduce or stop their infra expenditures (which isn't happening, like, quite the opposite). That's what I would personally see as the top signal, but it's pretty dumb to exit the whole market because of a possible crash at some point in the future, without anything you see as a catalyst. Also, gold is completely overrated, it's not at all a safe place where to put your money and you have no idea how it would correlate with an AI bubble burst
Fiat money sure came in handy when 10% or more of the trees in my entire county were blow over toward the east.
Long term, you're getting eaten by inflation, but when the power grid, and most cell sites are down, cash is still king.
The real lesson I've learned is that a small generator consumes about 10 gallons of gasoline per day. There's no way to store enough of it to last through a total collapse, even if you could keep it from being stolen.
Chip makers may be fully booked now and have extra revenue because of AI, but that also creates deferred demand elsewhere. The prices will not fall immediately because of the crash of a single company. The scarcity has probably negative impact on the economy in general, so stock market should recover after that quickly or at least stay near a healthy valuation level.
when market sees huge storm of un-bought devices, prices will drop, companies waiting on the line will also drop their contracts because they don't want to buy chips with pre-crash prices and renegotiate their prices
Yes. Those on the demand side will benefit from it. And that‘s the whole point. It is much healthier situation economically than current, very distorted market.
Edit the config of the AI startups you have properly designed to be platform agnostic and make it use the other platforms models, wait the CI green and deploy prod
Not commenting on the rest of the piece but why is everyone assuming that OpenAI’s executive departures are all voluntary? Sam Altman has shown plenty of willingness to clean house among the top ranks in the past. He isn’t going to shy away from firing a revenue officer if targets aren’t being met.
I don't think they need to be voluntary to be pessimistic about the implication of the departures. It can be argued just as easily that so may high level execs being fired because of underperformance means OpenAI isn't achieving their goals. Revenue targets not being met is a large part of AI bubble promoters' thesis.
The article asserts that executives are leaving because they know OpenAI will fail and so their shares will be worthless soon. That’s a much more extreme read of the situation than the company simply being behind on some internal goals.
This. I also think there'll be more local AI models to fill gaps and people see how we've become needlessly dependent and we start using AI responsibly.
I would watch it happen with glee. I dont know exactly what all the other bad side effects it would had. I assume it will sux majorly for a lot of people and would not like that.
But, OpenAI investors and management loosing money and power would be the one part of it I would enjoy.
Just another article re-iterating the same bearish takes on these companies with sprinklings of quirky uses of profanity to let you know the author is a human. If you're not interested in reading the same thing you've been reading for the last few years, feel free to move on.
This seems to be an increasingly likely outcome. But In the aftermath, like the dot com crash, the debt, and assets are distributed in the liquidation. So it's not the death of AI, or even of ChatGPT. Doctorow wrote about this, but with less bombast than Zitron. https://doctorow.medium.com/https-pluralistic-net-2025-09-27...
A less bombastic quote from the cited Doctorow: " AI is the asbestos we are shoveling into the walls of our society and our descendants will be digging it out for generations". I am a big fan of Cory, but as a developer I tend to think that AI will continue to aid us in refining what we shovel as well as what was once already shoveled -- refining it from a carcinogen into a benign and load-bearing :D foundation. The metaphor is catchy, but may not be apt, unless somehow we unilaterally decide that use of AI itself is the cancer and we all act to eliminate it. I think a dismal fate for OpenAI and Anthropic is likely quite separate from the fate of AI itself, particularly for software development.
I think if you look at how AI is being used anywhere outside of computer science the quote makes a bit more sense.
AI is being used to dehumanize people's jobs, provide bad customer service, a bunch of people are being fired for dubious benefit, people are being turned into accountability sinks (the AI is in charge of decisions, but the human takes the blame).
The actual numbers he presents are well sourced and, sadly, correct.
Even if AI is the future of computing, we will have a 9.x economic earthquake headed our way.
Read his articles carefully. Don't be distracted by the salt, the swearing is just there to help the numbers go down easier.
Work it out with a pencil. For those of us with the usual innumeracy about billions, trillions, etc, it can be helpful to draw diagrams on graph paper.
Also, @dang, unflag this please, it is well within the HN ambit.
Zitron is far more negative when it comes to the utility of LLMs than I am, but dear God would it be satisfying to watch him be proven right on the financials.
His as-of-late frequent appearances on Bloomberg, CNBC, and Scott Galloway's podcast, as well as his collaborations with Ed Elson from Galloway's podcast must feel like a great vindication for him already. Quite impressive for a person who is self-taught in that area, just over the past two years.
edit: And, of course, the post is flagged. We must bury our heads in the sand again.
Doesn't look flagged to me. Though it triggered the flamewar checks I think
#142 now when it was on the front-page mere minutes ago: I guess, TFA is way too exaggerated & seeped in avoidable animosity to warrant it.
It’s far from vindication. All these TV networks have slots dedicated for extreme doomers to come in and say “the sky is falling! There’s going to be a recession soon. Sell everything!!” Been seeing it every day for 15 years now.
They generally don't show their hand, Zitron has shared his complete thesis openly
Time will tell.
I had to use Google to find the comments after I read the story. Y-combinator moderation is far from invisible.
Zitron is really really good. He taught me the strategy of "screenshot all the wild AI-bro claims" to keep track of them and call them out when it all collapses. The comments here villifying him are wrong and unhinged. Sometimes you need to be disliked a little to correctly predict the future.
>"screenshot all the wild AI-bro claims" to keep track of them and call them out when it all collapses.
But have you done the same for Zitron? Because the AI-Bros have consistently been more accurate and Zitron consistently pretty wrong.
Zitron proclaimed in (early) 2024 that LLMs were stuck because there was no more data to use, and synthetic data ruined models, so we were already at the brink of collapse.
Zitron predictions regarding LLMs, as technology, are meaningless. Literally nobody can predict the future, and he's not a computer scientist. What matters is his thesis for things that have actual numbers that can be verified. Look at the numbers and his claims, and decide for yourself if the thesis is valid
What happens if people realize Ed Zitron has been rage farming to harvest subscription dollars for a few years now?
He sells outrage to ill informed people who don't know any better. There's real issues, but it's far more nuanced than Zitron seems willing to bring to his audience.
look, he says right at the beginning he is NOT trying to be a buzzkill!!! this is real news!!! haha
It's not, it's an editorial, i.e an opinion piece...
[dead]
Thats basically all journalism
I welcome AI skepticism but Ed’s often does seem to be done in bad faith. I feel like his claims border on, “AI has no use case whatsoever.” Which clearly isn’t true in 2026.
I don’t think that’s Ed’s position. As I understand it, Ed’s position is that it has no real economic value.
AI is a tool, a useful tool, an innovative tool, but technology has introduced innovative tool after innovative tool for years. Apache, Twitter’s Bootstrap, Pull to Refresh, React, Kubernetes, all innovative, all delivered real business value. Nobody invested hundreds of billions of capital into infrastructure for React.
AI has endless use cases. Theres no evidence that these use cases deliver the value necessary to justify any capital expenditure. A technology can change the world and have no standalone business case.
Business is boring. Money in, money out. We are still waiting for Anthropic and OpenAI to show how LLMs have moved the needle on business.
Every single thing that an LLM can be used to deliver value for a business today can be achieved by a cheap to run open weight model that companies could host in house. Where does that leave the trillion+ worth of dollars being buried in the ground?
Another nuance: his position is that real economic value hasn't been demonstrated, and that due to how LLMs work, it cannot really be measured. Meaning the AI crowd actual believes are based on vibes and doesn't seem to match reality
(also, of course, that there is a massive, insane misallocation of resources to the AI sector compared to the level of demand)
His position is also that it is useless. It is pretty consistent and openly stated.
Can you provide a link to where he has said that? I’m not an expert on Ed but I don’t recall ever reading him say that it is functionally useless, he doesn’t comment on its function, only the business, whether it has functional use is immaterial unless that generates value for the companies spending obscene money on hosting. The companies implementing AI as consumers can get a huge amount of value from it without a single cent going to any of the big labs, that’s part of why investment in AI infrastructure is so misplaced.
He often says in passing that he doesn't see value in LLMs outside of generating bad software and can help sometimes with search. But it's not really part of the AI bubble thesis itself IMHO, just him adding his personal commentary in interviews
That's just his own personal belief about the technology, the only relevant part of this is that he doesn't believe there will be enough demand to match the level of expanses. It doesn't matter if AI is useful or not for his arguments. Does the demand exist or is it manufactured by hyperscalers or other actors? Who cares about his take on the future of software
I have both read his blog and listened to his podcasts. Not all of it (the articles are basically books), but enough to understand his positions. The "ai is mostly useless" is very much his position, he will say it again and again and again.
This part:
> I don’t think that’s Ed’s position. As I understand it, Ed’s position is that it has no real economic value.
Is not true. I can agree with one of his positions and disagree with the other, but it really is fair to say he repeatedly and confidently states that ai is useless.
I think you wanted to respond to another comment? What you’re quoting isnt my words
Probably a cambrian explosion of new startups. New stars are made of the refuse of old ones.
Stock market might dip 20%+, because OpenAI is a large buyer of GPUs, RAMs, servers and disks, every chip company will be impacted from removing huge customer from their revenue.
then flywheel effect might kick off and impact carriers and construction companies building data centers and delivering things there.
which impacts investor trust and they might start pulling their money from AI companies, which in turn, they either need to increase prices or downsize their efforts
I find it hard to justify any conclusion other than we're about to repeat the crash of 1929 in terms of stock prices. Back then, stocks dropped 90% over 3 years. We'll be lucky if 10% actual value remains after the bonfire that's about to happen.
So, what, should I withdraw my 401k and buy physical gold?
That's up to you, but if there is a crash it's very unlikely to be before a major hyperscaler decides to reduce or stop their infra expenditures (which isn't happening, like, quite the opposite). That's what I would personally see as the top signal, but it's pretty dumb to exit the whole market because of a possible crash at some point in the future, without anything you see as a catalyst. Also, gold is completely overrated, it's not at all a safe place where to put your money and you have no idea how it would correlate with an AI bubble burst
Uh oh, this is the comment chain where somebody is gonna say "fiat"
Fiat money sure came in handy when 10% or more of the trees in my entire county were blow over toward the east.
Long term, you're getting eaten by inflation, but when the power grid, and most cell sites are down, cash is still king.
The real lesson I've learned is that a small generator consumes about 10 gallons of gasoline per day. There's no way to store enough of it to last through a total collapse, even if you could keep it from being stolen.
Chip makers may be fully booked now and have extra revenue because of AI, but that also creates deferred demand elsewhere. The prices will not fall immediately because of the crash of a single company. The scarcity has probably negative impact on the economy in general, so stock market should recover after that quickly or at least stay near a healthy valuation level.
when market sees huge storm of un-bought devices, prices will drop, companies waiting on the line will also drop their contracts because they don't want to buy chips with pre-crash prices and renegotiate their prices
Yes. Those on the demand side will benefit from it. And that‘s the whole point. It is much healthier situation economically than current, very distorted market.
Edit the config of the AI startups you have properly designed to be platform agnostic and make it use the other platforms models, wait the CI green and deploy prod
Not commenting on the rest of the piece but why is everyone assuming that OpenAI’s executive departures are all voluntary? Sam Altman has shown plenty of willingness to clean house among the top ranks in the past. He isn’t going to shy away from firing a revenue officer if targets aren’t being met.
I don't think they need to be voluntary to be pessimistic about the implication of the departures. It can be argued just as easily that so may high level execs being fired because of underperformance means OpenAI isn't achieving their goals. Revenue targets not being met is a large part of AI bubble promoters' thesis.
The article asserts that executives are leaving because they know OpenAI will fail and so their shares will be worthless soon. That’s a much more extreme read of the situation than the company simply being behind on some internal goals.
Unfairly flagged, bumped way off front page: https://news.social-protocols.org/stats?id=49347207
It’s a long read with well-sourced numbers that make the point well.
I doubt they would go public if anthropic does and it’s a slaughter. Allowing anthropic to go first might be in their best interest.
I guess it really shows how people like to ignore that there's just not that much money in the economy
We all get 1TB ram in our PC as the default
your data get's sold
This. I also think there'll be more local AI models to fill gaps and people see how we've become needlessly dependent and we start using AI responsibly.
I would watch it happen with glee. I dont know exactly what all the other bad side effects it would had. I assume it will sux majorly for a lot of people and would not like that.
But, OpenAI investors and management loosing money and power would be the one part of it I would enjoy.
Just another article re-iterating the same bearish takes on these companies with sprinklings of quirky uses of profanity to let you know the author is a human. If you're not interested in reading the same thing you've been reading for the last few years, feel free to move on.
Could we end up with a merger between them and Anthropic?
https://www.wheresyoured.at/what-happens-if-openai-dies/#ant...
> I think it’s very unlikely this happens, and if it does, it would be ruinous for everybody involved.
The full snippet is in TFA.
hahahaah thank you, I asked too soon, realized later it was covered in there lol
This seems to be an increasingly likely outcome. But In the aftermath, like the dot com crash, the debt, and assets are distributed in the liquidation. So it's not the death of AI, or even of ChatGPT. Doctorow wrote about this, but with less bombast than Zitron. https://doctorow.medium.com/https-pluralistic-net-2025-09-27...
A less bombastic quote from the cited Doctorow: " AI is the asbestos we are shoveling into the walls of our society and our descendants will be digging it out for generations". I am a big fan of Cory, but as a developer I tend to think that AI will continue to aid us in refining what we shovel as well as what was once already shoveled -- refining it from a carcinogen into a benign and load-bearing :D foundation. The metaphor is catchy, but may not be apt, unless somehow we unilaterally decide that use of AI itself is the cancer and we all act to eliminate it. I think a dismal fate for OpenAI and Anthropic is likely quite separate from the fate of AI itself, particularly for software development.
I think if you look at how AI is being used anywhere outside of computer science the quote makes a bit more sense.
AI is being used to dehumanize people's jobs, provide bad customer service, a bunch of people are being fired for dubious benefit, people are being turned into accountability sinks (the AI is in charge of decisions, but the human takes the blame).
Doctorow is also funnier than Zitron.
Another excellent Zitron article.
The actual numbers he presents are well sourced and, sadly, correct.
Even if AI is the future of computing, we will have a 9.x economic earthquake headed our way.
Read his articles carefully. Don't be distracted by the salt, the swearing is just there to help the numbers go down easier.
Work it out with a pencil. For those of us with the usual innumeracy about billions, trillions, etc, it can be helpful to draw diagrams on graph paper.
Also, @dang, unflag this please, it is well within the HN ambit.
[dead]