A class of problems I’d expect to have to ward off with a rationality Q&A site are explicit or implicit requests for personal advice… following the MathOverflow example, perhaps “Questions of interest to other rationalists” would be useful.
Another problem is that voting on the answers might start to be an opinion poll. It might just become a vehicle for majoritarianism in the community.
My first thought to treat this would be that answers should be “Robust to variation in priors”. I believe Jaynes had similar advice for the philosophy of science. Anyway, it would prevent people mixing in particular with the answers. The goal would be to have answers provide “update procedures” rather than the posteriors of others’ updates. “A good answer is a good update procedure”?
This is a good question and I don’t have a good answer yet.
(I agree with you that creating competition with LW is not desirable.)
A class of problems I’d expect to have to ward off with a rationality Q&A site are explicit or implicit requests for personal advice… following the MathOverflow example, perhaps “Questions of interest to other rationalists” would be useful.
Another problem is that voting on the answers might start to be an opinion poll. It might just become a vehicle for majoritarianism in the community.
My first thought to treat this would be that answers should be “Robust to variation in priors”. I believe Jaynes had similar advice for the philosophy of science. Anyway, it would prevent people mixing in particular with the answers. The goal would be to have answers provide “update procedures” rather than the posteriors of others’ updates. “A good answer is a good update procedure”?