There's a weird gleefulness about AI being a bubble that'll pop any day now both in this thread and in the world at large. It's kinda weird and I find most of the predictions about the result of such a bubble popping to sound highly exaggerated.
I think it’s because a lot of people feel like AI is being pushed from the top a lot with all kinds of spectacular claims / expectations, while it’s actually still difficult to apply to a lot of problems.
The AI bubble bursting would be somewhat of an “I told you so!” moment for a lot of people.
And there’s also a large group that’s genuinely afraid to lose their job because of it, so for that group it’s very much understandable.
Whether or not the predictions of the bubble popping are exaggerated or not, I cannot tell; it feels like most companies investing in AI know that the expectations are huge and potentially unrealistic (AGI/ASI in the next few years), but they have money to burn and the cost of missing out on this opportunity would be tremendous. I know that this is the position that both Meta and Google shared with their investors.
So it all seems to be a very well calculated risk.
I do agree that there seems to be a bubble, imo it's largely in the application space with the likes of cursor being valued at $23B+, but I don't see GPU purchases going down anytime soon and I don't even see usage going down. If these overhyped apps fail then it seems like something else will take their place. The power that LLMs provide is just too promising. It seems like those predicting things like a global economic crisis or the new big model-providers like OpenAI going to 0 just think that AI is like NFTs with no real intrinsic value.
I work for a company making AI related products. I've asked multiple Executive VP's if we are in an AI bubble. They all dance around the question.
I started working in 1997 and saw the dot com bubble and collapse. If you had asked me in 1999 I thought it would go on forever. I used to get 5 recruiter calls a day. In 2001 I got zero recruiter calls the entire year and it took me 10 months to find a new job.
I saw people at Cisco that had exercised stock options and held on to it at $80 and then it dropped to $20 and they had huge AMT tax bills that they couldn't pay even if they sold everything they had including their house.
My warnings of the AI bubble popping are not just an "I told you so" but also a warning to people to diversify out of their company stock and diversify in general.
