Thanks. That sounds good, but it is an experimental program, not something you’d observe on Less Wrong.
I expect that you could get more complex results than yes or no. Like with some primes or some observers preparing a retreat would help, with others it wouldn’t, and in some contexts you’d lose status and credibility directly for trying to prepare a retreat.
True. We are interested in communities where truth-tracking is high status, so that cuts down the number of contexts. We would also probably need to evaluate it against other ways of coping with being incorrect (disassociation e.g. Eliezer(1999), apology etc) and see whether it is a good strategy on average.
Thanks. That sounds good, but it is an experimental program, not something you’d observe on Less Wrong.
I expect that you could get more complex results than yes or no. Like with some primes or some observers preparing a retreat would help, with others it wouldn’t, and in some contexts you’d lose status and credibility directly for trying to prepare a retreat.
True. We are interested in communities where truth-tracking is high status, so that cuts down the number of contexts. We would also probably need to evaluate it against other ways of coping with being incorrect (disassociation e.g. Eliezer(1999), apology etc) and see whether it is a good strategy on average.