So here’s a comment from a VP thread instead of something original.
Another reason science is a poor guide is that, done properly, it takes too long. We are in a very strange moment in time where everything is changing very quickly: genetic mutations, magnetic north, weather cycles, ecological conditions, etc. This is a poorly understood source of error in all but the most immovable sciences (e.g. physics). By analogy, imagine that the slow shift in the fundamental physical constants of cosmology were instead quick and erratic. This would introduce unaccounted errors in experimental methods.
I noticed about five years ago that the meteorologists were suddenly doing a fantastic job of predicting the 5-day forecast. All that data-crunching appears to have paid off. But now I’m noticing huge systemic biases in their mistakes (last year and a half or so). They still do an incredible job, considering, but they were constantly guessing low on temperatures in Minnesota’s this year. That sort of thing isn’t supposed to happen. And the precipitation is throwing them off much more than usual, although at first glance the errors at least appear to be properly random.
Or maybe I’m just being normal crazy.