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
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
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.
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
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.
> but dear God would it be satisfying to watch him be proven right on the financials.
I don't get why there are so many people who need Zitron to be wrong. It reminds me of back in the days when people still liked Elon and every criticism of him or Tesla must be a short-seller spreading lies to move the price.
I think it's pretty easy to slot Ed's predictions into your possibility space even if you're still betting on a different outcome being more likley. Something else is going on.
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.
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.
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.
So many numbers.. such a cold explanation.. what this misses though is that the world is not ruled by logic, but rather by people and if the right people want these companies to succeed, they will succeed. At least some of them will because they are instrumental to implement the continental fortresses (including surveillance and police state) to deal with the climate and ecological collapse that the elites know is coming.
41 comments
[ 0.22 ms ] story [ 38.0 ms ] thread> 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.
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
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.
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.
I don't get why there are so many people who need Zitron to be wrong. It reminds me of back in the days when people still liked Elon and every criticism of him or Tesla must be a short-seller spreading lies to move the price.
I think it's pretty easy to slot Ed's predictions into your possibility space even if you're still betting on a different outcome being more likley. Something else is going on.
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.
It’s a long read with well-sourced numbers that make the point well.
But, OpenAI investors and management loosing money and power would be the one part of it I would enjoy.