Thread by @patio11

A very interesting conversation about forecasting the future (with the Superforecaster guy), where it tends to pick up predictive alpha, the sociological makeup of people who repeatedly make good predictions, etc.

conversationswithtyler.com/episodes/phili…

An ambient belief I have had for many, many years:

Life (for a person, a group, humanity) is an iterated test, with lots of confounding factors. Some problems on the test announce "I am a high school math problem. Seriously."

We can observe that not all strategies test equally.

We are currently faced with an iterated test, with a very tight feedback loop, where *a lot* of the problems announce "I am a high school math problem" or "Choose your own adventure: med school, epidemiology undergrad, or paid attention in 5th grade science or European History."

This test is distinguished from most tests by the iteration rate being higher, the answers being marked in a far more obvious fashion, the confounding factors being lesser because most of them operate on a longer timescale than the iteration speed, and the stakes being higher.

So if there was ever a method for doing well on tests or a stat on a character sheet which gives a bonus to your roll, that is going to be both important as a matter of societal concern and also quite a bit more obvious than it normally is from the constant test that is life.

Epistemic status: This is a model I use to interpret things I have seen in the world over the course of my life, which I have found to repeatedly generate predictions which work well for me, but I acknowledge my reasoning may be corrupted.

If you had asked me to write a character sheet for myself in high school, I would have put *all* the points into whatever skill governed the roll for doing well on tests, and I have spent much of my adult life sometimes regretting that effort allocation, which through mid-20s.

(On recent evidence, I regret it less.)

* was same