Thanks, this is very insightful and I think very true. I saw Saylor on Lex Friedman's podcast and he was criticizing economists for focusing too much on aggregate stats like inflation, when they should be building agent-based models instead. But it felt like a decades-old outdated criticism because economists can and do build agent based models these days using heavy computation and data. (Of course, the value of the models are still questionable because human behavior isn't always stable over time.)

Anyway, thanks for this. Helps me understand where he's coming from a bit better.