Throughout the “AI bubble” talk in 2024 and 2025, I consistently argued that we were nowhere near the peak of the AI bubble. So far, that view has held up, as valuations are significantly higher today than they were in 2024 and 2025. If you look at the way the dotcom bubble unfolded, dotcom didn't take off until after Netscape IPOed in 1995. The market had 5 more years of growth until the collapse. And even after col…
I think we're a lot closer to the peak than when Netscape IPO'd relative to the dotcom bust for a few reasons: * big banks are trying to get out of their data center loan commitments, even selling that debt at a discount. From the article: > According to the Financial Times, major lenders are already scrambling to offload pieces of massive data center loans through private transactions, risk transfers and synthetic s…
This isn't necessarily a sign that they don't believe in the data centre loans, it's more than banks are basically required to avoid concentrated risk, because of the regulations we (mostly correctly) imposed upon them post GFC.
Now, personally I'm not convinced there's enough demand for AI services that these datacentres make sense, but we'll see I guess.