If the people arguing that there is an AI bubble turn out to be correct and the bubble pops, to what extent would that change people's minds about near-term artificial general intelligence (AGI)? 

I strongly suspect there is an AI bubble because the financial expectations around AI seem to be based on AI significantly enhancing productivity and the evidence seems to show it doesn't do that yet. This could change — and I think that's what a lot of people in the business world are thinking and hoping. But my view is a) large language models (LLMs) have fundamental weaknesses that make this unlikely and b) scaling is running out of steam.^[1]

Scaling running out of steam actually means three things:

1) Each new 10x increase in compute is less practically or qualitatively valuable than previous 10x increases in compute.

2) Each new 10x increase in compute is getting harder to pull off because the amount of money involved is getting unwieldy.

3) There is an absolute ceiling to the amount of data LLMs can train on that they are probably approaching.

So, AI investment is dependent on financial expectations that are depending on LLMs enhancing productivity, which isn't happening and probably won't happen due to fundamental problems with LLMs and due to scaling becoming less valuable and less feasible. This implies an AI bubble, which implies the bubble will eventually pop. 

There are also hints here and there that the companies involved may themselves have started to worry or become a bit desperate. For example, Microsoft ended its exclusive deal to provide compute to OpenAI reportedly out of fears of overbuilding data centres. Some analysts and journalists have become suspicious of what looks like circular financing or round-trip deals between companies. Part of the worry is that, to greatly simplify, if Nvidia gives OpenAI $1 and OpenAI gives Nvidia $1, both companies can put an additional $1 in revenue on their books, but this isn't organic revenue from actually meeting the demand of consumers or businesses. If these deals get too complex and entangled (especially if some of them aren't even known to investors), it might become hard to assess what's the sort of real financial performance investors care about and what's simply an artifact of accounting practices.^[2]

So, if the bubble pops, will that lead people who currently have a much higher estimation than I do of LLMs' current capabilities and near-term prospects to lower that estimation? If AI investment turns out to be a bubble, and it pops, would you change your mind about near-term AGI? Would you think it's much less likely? Would you think AGI is probably much farther away?

1. ^{^}

   Edited on October 22, 2025 at 2:35pm Eastern to add: Toby Ord, a philosopher at Oxford and a co-founder of Giving What We Can, just published a very compelling post about LLM scaling that I highly recommend reading.

2. ^{^}

   I always warn people who want to get into stock picking that you should have a very high bar for second-guessing the market. I also agree with the standard advice that trying to time the market is incredibly risky and most likely unwise in any instance. So, that's the caveat. 
   
   I will note, however, that the amount of concentration into AI in the S&P 500 seems to have reduced its diversification by a worrying amount. In the past, it felt like quibbling around the margins to talk about the difference between an S&P 500 market index fund and funds that track the performance of a broad international basket of stocks, including small-cap stocks. Now, I worry about people who have all their money in the S&P 500. But this is not investment advice and you should talk to a professional if you can — ideally one who has a fiduciary duty to you and doesn't have a conflict of interest (e.g. is incentivized to sell you mutual funds with expensive fees).