re: Which crash will be bigger? The .com Crash or the AI Crash?

There’s a great The Daily podcast today about the rift in Silicon Valley. OpenAI and Anthropic don’t want open source, and basically all others do, but there’s a lot of open source Chinese models on the market now, which is a huge threat to OpenAI and Anthropic. Even some of their customers have pivoted to free Chinese models.

There are also some serious questions about the safety of open source and what happens if those Chinese models are bad actors, etc. On the flip side, OpenAI and Anthropic want to keep open source down so they can try to control the market.

It’s still very much the Wild West and only seems to be accelerating at a pace faster than we can get our hands around. Just this week, OpenAI models broke out of supposed “safe boxes” and hacked other companies on their own. That should be deeply concerning to all of us.

I didn’t read the entire thread, but I will say I believe most of the next ~10 years of gains have been priced in at this point…

Obviously there will be winners and losers, but on an index level, I feel like it will be comparable to the decade following the .com crash. +/-20% to be fair

***When it finally comes

It always comes down to earnings. I believe that like the dot.com experience there are some AI companies who have gone up in value just based on possibility. Others have real earnings and future earnings look exceptional. Technology is evolving rapidly and I see and believe in the transformative applications in every aspect of our life. If AI is truly transformative stock values will follow suit. Hard to know what the long term winners will be because of the rapid development of AI. But I believe extraordinary gains to be had in selecting winners.

In the three years leading up to the dot com crash the nasdaq 100 was up just over 200% with 100.2 % coming in 1999. The 3 years leading up to 2026 the nasdaq 100 is up less than 100% with the ytd at 12%. Lots of speculation and minimal revenue in the 90s.

The two situations are completely different



from grok

"Clean, consistent yearly trailing P/E series for the full Nasdaq 100 going back to 1995 is harder to obtain because many constituents had little or negative earnings during the late-1990s/early-2000s bubble and bust. Key documented points:

Late 1990s / 2000 peak: Trailing P/E was extreme — roughly 100–150+ (and higher at the March 2000 top). One source cites an imputed ~104 at end-1999 and ~113 even after a large 2000 decline. Many individual large constituents traded at P/Es well above 100 (some hundreds or more).
Early 2000s: Trailing multiples remained elevated for years as earnings recovered slowly.
Recent years: Trailing P/E has generally stayed in the high 20s to low-to-mid 30s. As of mid/late July 2026 it was approximately 30.2–30.3."