I like the EA section. I think grouping people by specific goals/interests and preparing questions for those goals is the right way. If I cared about EA, then being able to predict which charities will start/stop being effective, before they actually implement whatever changes they’re considering, would allow me to spend money more efficiently. It would be good not only to have an accurate personal model, but also to see other people with better models make those predictions, and know how reliable they really are.
Likewise, we could have something about AGI, e.g. “which AGI safety organization will produce the most important work next year”, so that we can fund them more effectively. Of course, “most important” is a bit subjective, and, also, there is a self-fulfilling component in this (if you don’t fund an organization, then it won’t do anything useful). But in theory being able to predict this would be a good skill, for someone who cares about AGI safety.
Problem is, I don’t really know what else we commonly care about (to be honest, I don’t care about either of those much).
I think the world politics do have transferability in calibration
I would also like this to be true, but I wonder if it really is. There is a very big difference between political questions and personal questions. I’d ask if someone has measured whether they experience any transfer between the two, but then I’m not even sure how to measure it.
It might be nice to have a set of twenty EA questions, a set of twenty ongoing-academic-research questions, a set of twenty general tech industry questions, a set of twenty world politics questions for the people who like them maybe, and run multiple contests at some point which refine predictive ability within a particular domain, yeah.
It’d be a tough time to source that many, and I feel that twenty is already about the minimum sample size I’d want to use, and for research questions it’d probably require some crowdsourcing of interesting upcoming experiments to predict on, but particularly if help turns out to be available it’d be worth considering if the smaller thing works.
I like the EA section. I think grouping people by specific goals/interests and preparing questions for those goals is the right way. If I cared about EA, then being able to predict which charities will start/stop being effective, before they actually implement whatever changes they’re considering, would allow me to spend money more efficiently. It would be good not only to have an accurate personal model, but also to see other people with better models make those predictions, and know how reliable they really are.
Likewise, we could have something about AGI, e.g. “which AGI safety organization will produce the most important work next year”, so that we can fund them more effectively. Of course, “most important” is a bit subjective, and, also, there is a self-fulfilling component in this (if you don’t fund an organization, then it won’t do anything useful). But in theory being able to predict this would be a good skill, for someone who cares about AGI safety.
Problem is, I don’t really know what else we commonly care about (to be honest, I don’t care about either of those much).
I would also like this to be true, but I wonder if it really is. There is a very big difference between political questions and personal questions. I’d ask if someone has measured whether they experience any transfer between the two, but then I’m not even sure how to measure it.
It might be nice to have a set of twenty EA questions, a set of twenty ongoing-academic-research questions, a set of twenty general tech industry questions, a set of twenty world politics questions for the people who like them maybe, and run multiple contests at some point which refine predictive ability within a particular domain, yeah.
It’d be a tough time to source that many, and I feel that twenty is already about the minimum sample size I’d want to use, and for research questions it’d probably require some crowdsourcing of interesting upcoming experiments to predict on, but particularly if help turns out to be available it’d be worth considering if the smaller thing works.