Here's to hoping it doesn't take long to burst, but I'm afraid nvidia still has too much money to keep the bubble going until 2027
this post was submitted on 04 Nov 2025
108 points (97.4% liked)
Fuck AI
4558 readers
1565 users here now
"We did it, Patrick! We made a technological breakthrough!"
A place for all those who loathe AI to discuss things, post articles, and ridicule the AI hype. Proud supporter of working people. And proud booer of SXSW 2024.
founded 2 years ago
MODERATORS
Nvidias stock price is frankly insane.
I would short it, but if I went by all my instincts I’d have been shorting Tesla for 5 years now.
And the liability OpenAI holds is terrifying.
2008 happened because so staging like 13T of money (built on fraudulent mortgage values) that never should have existed disappeared, with these inflated prices we’re looking at a milder situation but probably still a big hit.
Given how the rest of the economy is probably already in mild recession I think this one will hurt, even if mild.
This all could have been sustainable, but Zuck and Altman and co are all pushing this way too far. Altman really needs someone to slap his hand as he keeps trying to take more and more.
Th economy is in a recession, and the money the hyperscalers have been throwing at AI is kinda like a bailout. Just wait for it all to stop.
It's not even a bailout, it's a financial potemkin village we've built to obscure how bad the recession actually is, and how much of a long-term depression it is likely to be.
I think the real problem is do they actually have that much money?
"have" is such a plastic concept.
They know it is a bubble, because they profit from it. My biggest fear is that they will get governments to bail them out once it starts bursting.
iT’s A sTrAtEgIc BeT, AGI
"Markets can remain irrational longer than you can remain solvent"
That's because it allows someone to win a lot of money without producing anything in the real world.
Who knows though.
Everyone can see it coming, but no one can say when it is coming. Could pop tomorrow, could last for another two years. That's the problem with bubbles.
Everyone notices the housing market is a bubble but it still keeps going. Money printer go brrrr.
That's what people said from 2005, took years for the straw to break the camel's back.
3 years? That's not a lot of time. This AI bubble probably has about that much life left in it.
There is one difference between housing and stocks. When housing market crashes, you are still left with a house to live in. When stock crashes... well, nothing.
I think it's no longer riding in the columns of hot air going skyward. It may even be starting it's float down, but something still needs to pop it.
I wish, but people who have been using AI to cheat through highschool/college will not suddenly study two years worth of material over the month.
People will convince themselves that they alone use it responsibly instead. The bubble will only be actually burst when a worse opportunity comes up.
Like how AI burst NFTs and how NFTs burst Metaverse.
On the bright side people may not find it ok to answer a friend's hypothethical by stopping them mid-conversation to ask chatgpt anymore.
..people who have been using AI to cheat through highschool/college will not suddenly study two years worth of material over the month..
Oddly enough, I know several younger people - men and women - who are no longer using AI to write up their work for school, because as they claim, the results are getting poorer. In one case the AI generated a set of conclusions that were entirely unrelated to the main point of the paper. The guy ended up backpedaling on the whole project, which he'd only started on and had to use several weekends to sort it.
Where I would use AI was to find the source materials and then verify those sources to make sure they were organic and not synthetic, then go forward on my own...
But I'm also an old fart, so what do I know of how higher-ed rolls today.
I have a fold of screenshots labelled "AI slop", specifically, very fabricated, incorrect answers to scientific questions.
Meanwhile, high school teacher are telling kids this is the technology of the future. Don't take career advice from someone who ended up a high school teacher based on career advice.
people who have been using AI to cheat through highschool/college will not suddenly study two years worth of material over the month.
They might if the technology suddenly becomes prohibitively expensive to use. The cost of LLMs is hugely subsidized right now because AI companies know that if they passed the real cost of compute on to their customers, no one would continue to use their products. And the problem is getting worse. New models are using MORE computation, not less, and AI companies are continuing to burn mountains of cash 'hyper scaling' instead of trying to figure out how to bring costs down.
And that's fine for them while they are still investors for them to grift... but the VC money won't last forever. Eventually investors are going to expect some sort of return on investment, and most big investors are already massively over-leveraged on AI, so every quarter there's less and less VC cash to go around.
So some day soon when the investment well runs dry, AI companies will have no choice but to start passing their costs onto their customers, and we'll all finally see just how grossly expensive this technology really is.
Idk, the capitalist class is salivating at the prospect of replacing workers with ai. I think they'll continue to throw money at it. Besides when have we ever passed on an opportunity to destroy the environment for the chance at a buck or two?
Crypto seems headed into a correction and that's usually a frontrunner of tech more generally. When the pullback begins to affect "AI", the bubble will probably be poppin...
On the other hand I can't see the future. It could be slow, unpredictable, never, etc. I think it's possible the "AI" bubble could metastasize rather than pop.
Didn't the big short guy just reveal a huge bet against nvidia?
Hmm. Is it good to agree with him?