So how can we practice bringing order out of scientific chaos?

It might be that most of us were born too early to become students of the Bayesutsukai, because we've already been exposed to too many answers or hints. There's plenty of existing science that we don't know, of course, but we'll never have the experience of discovering evolution ourselves. Maybe we should be training six-year-olds.

We can get some practice on made-up worlds; a bigger version of Zendo. I'm sure this is better than no training, but we are not products of these made-up worlds, so I don't think it would be as effective for teaching us to be cognitive scientists.