Comment by enaaem

Decision making is risk management, and is often simpler than making predictions. You need to think in terms of utility functions and you need to stay far away from the catastrophic regions. You often don't need super accurate predictions for that.

One example is pandemics. The thing is pandemics are extremely fat tailed in terms of fatalities. You need a lot of data to fit any outbreak model and no one has time for that. But the reaction to diseases out breaks is very simple. Is it super deadly and contagious? If yes, shut everything down! This is essentially what Asian countries like Vietnam and Taiwan did during Covid and they handled it way better than the West. I remember arguing about outbreak models, which is a completely useless activity.