Agreeing on that could be the starting point to discuss which factors were responsible for that completely different behaviour (and in which way that could be related to a potential AI crash), like:


I view it quite differently: Ithink 2026 is far more comparable to 2000 than to 2008.

Both 2000 and today were bull markets fueled by enormous enthusiasm around a transformative new technology—the internet then, AI now. The market behavior is nearly identical.

When the music stopped in 2000, money didn't leave the system the way it did in 2008–09. It rotated. Capital moved out of the wildly overvalued technology stocks and into reasonably priced value stocks and other parts of the market.

We've already seen small versions of that rotation during this cycle. If AI ultimately proves to be a bubble, I would expect a much more pronounced version of the same thing—and that's the environment in which Berkshire did exceptionally well in 2000–02. So did many value stocks.

If, instead, you mean deep systemic risk like 2008, then yes, everything goes down. Berkshire included.

I just don't think 2008 is the appropriate historical analogy for an AI bubble.