The AI valuation bubble is now getting silly(theguardian.com)
theguardian.com
The AI valuation bubble is now getting silly
https://www.theguardian.com/technology/nils-pratley-on-finance/2025/oct/08/the-ai-valuation-bubble-is-now-getting-silly
10 comments
> The development of AI will clearly have society-changing effects, just as the arrival of the internet and mobile communication did.
Or, maybe the economics can't be made to pencil out, and AI is abandoned (again).
> The impossibility lies in knowing the speed of adoption, and which lumps of capital will earn extraordinary returns and which will end up getting torched.
Almost certainly all of it will end up getting torched.
Or, maybe the economics can't be made to pencil out, and AI is abandoned (again).
> The impossibility lies in knowing the speed of adoption, and which lumps of capital will earn extraordinary returns and which will end up getting torched.
Almost certainly all of it will end up getting torched.
… as long as we don’t get worried about that 4T crypto market…
Crypto transactions mostly just transfer money from one gamblers account to another's whereas AI is resulting in money going into real things like data centers.
the reason I don't really stress about valuations any more is the m2 US money supply has grown 20% in the last 5 years, cumulative inflation over last 5 years is ~24%, assets have no ceiling because fiat has no floor. -- you hear 4t now but thats 3t from 5 years ago.
Crypto is very difficult to value in terms of cash flows, whether BTC should be $120K or not is a big question mark. Regular companies do have standard valuation techniques that are currently silly.
By ‘standard valuation techniques’ you mean: regular companies produce value and wealth and represent what a society can create whereas crypto is like ‘a shiny object sitting there with people trying to gauge how much someone is willing to pay in the future on the premise it will be worth more’ ? :-D
I’m not very deep in the art or collectible business.
I’m not very deep in the art or collectible business.
> asking where today’s AI bubble – because that’s what it clearly is – fits in a 1990s timeline
Yet evidence is:
- CAPE is high - NVIDIA has a very large market cap - enormous capital investment in AI and relatively few companies
I assume also a chorus of “AI actually doesn’t make you more productive!” And “AI capex and opex vastly outweighs realized profits!”
Seems a little less than “clearly”.
All of these are very much RISK factors, yet you need to assume the market is being irrational and assume that AI is NOT going to have the impact the market thinks it will. Personally for me I don’t understand: pretty clear trend in capabilities without a clear and insurmountable roadblock, so I totally get the “I think it’s a bubble” argument, it’s just that I think people underestimate what’s to come
Yet evidence is:
- CAPE is high - NVIDIA has a very large market cap - enormous capital investment in AI and relatively few companies
I assume also a chorus of “AI actually doesn’t make you more productive!” And “AI capex and opex vastly outweighs realized profits!”
Seems a little less than “clearly”.
All of these are very much RISK factors, yet you need to assume the market is being irrational and assume that AI is NOT going to have the impact the market thinks it will. Personally for me I don’t understand: pretty clear trend in capabilities without a clear and insurmountable roadblock, so I totally get the “I think it’s a bubble” argument, it’s just that I think people underestimate what’s to come
"This time is different"
It’s more like people pretend it’s “obvious” that the market is going to do something or that AI developments are not going to support the spend, it’s just not obvious to me. I think people overweight probability on a bubble. Saying it’s “clearly a bubble” is hubris.
Bank of England flags risk of 'sudden correction' in tech stocks inflated by AI
https://news.ycombinator.com/item?id=45516265