Don't you end up having a problem with recursion? The people who use AI to predict the rational actors in the economy are actors themselves, and need to be included in their own model. Then you have others trying to predict what that model will say, etc.

What rational actors? Could you share some sources showing that individual economic choices are mostly driven by a cost-benefit analysis to determine whether an option is right for them? As opposed to being driven by habit, emotion, and social customs?

The rational choice for food would not lead to an obesity and diabetes epidemic in the wealthiest country. Something doesn't add up.

AI is even better at predicting mood and emotion. See social media.

Ok, so you couldn't give sources for your first claim, so now you're moving the goal posts.

Let's try again. Can you give sources showing that AI effectively models the economic decisions of non-rational actors? Anyone who can predict the economy should be able to quickly become very wealthy.

I never said we can perfectly model rational actors. I did say we can now do it better than before.

Fair point that you're just making a prediction about future capabilities. I'm not convinced but time will tell.