I guess signalling non-agency is tactical level; protective camouflage, poker bluffing etc. Agenty thinking as above is essentially strategic, winning with moves that are creative, devious, hard to predict or counter, going meta, gaming the system. Pretending to be a loyal citizen of Oceania is a good tactic while you covertly work towards other goals.
For cultural agency, the Wikipedia page on locus of control's one place to start. And there was the Power Distance Index in Gladwell's Outliers.
Just finished reading K.J. Parker's Devices and Desires. What struck me at first was "Eh, no, medieval people didn't think like that," but after mentally shifting gears to thinking of it as an author tract like HPMOR, with modern characters in a quasi-historical setting, it was much more enjoyable.
Upvote. The Drake Equation and SETI seem at least as relevant as, say, Pascal's Mugging. GIGO, sure, but a standard dismissal in statistics is to say there's not enough data, more research needed. Isn't this where Bayes is supposed to win over frequentism, when it comes to imperfect or incomplete information?
Babyeater FAI would be very different, but could still give us big hints on how to make human FAI. It's the standard science process, instead of reinventing the wheel, stand on the shoulders of giants and learn what other smart people who've come before have figured out.
Armchair psych, agreed. The one bit was patients performed poorly on the Cognitive Reflection Test, but yeah that's not news. I'm sure most people in most circumstances would perform poorly on the CRT. System 1 is like original sin, it's the default state that takes effort to overcome.