This is something I think about, only my framing is that of predictionism; what I mean is society's occupation with predicting things.

This is important because predictions are both 1) necessary to make value judgments of the present and 2) borderline impossible for many things. So you have people making value judgments that hinge on things they have no right to know.

I also classified predictions into three categories, based on difficulty. The easiest being periodic things like movements of planets. The second being things that have been known to happen and might happen again in the future, like war. And the third are novel phenomenas that have never happened before, like superintelligence. Even the second one is hard, the third is impossible.

There are so many predictions that fall in this third category that people are making. But no matter how many 'models' you make, it all falls into the same trap of not having the necessary data to make any kind of estimate of how successful the models will be. It's not the things you consider, it's the things you don't consider. And those tend to be like 80% of the things you should.