It would certainly be relevant to the selection algorithm that is used to select which people are surveyed.
I am not using any well-defined selection algorithm. I am simply writing everyone who is famous (many citations, good reputation, genuine insights, academic degree etc.), everyone listed on Wikipedia, everyone I can find by googling certain keywords (AIXI, reinforcement learning etc.), authors of studies and papers, people who have been suggested to me (e.g. by lesswrong) and everyone those people link to on their sites.
(And also people who “ask” me to post their answers ;-)
ETA I think that the possibility to suggest people is kind of a reassurance that I do not cherry-pick people. Just tell me who and I will ask. I post every answer I get as long as they gave their permission to publish it.
I am not using any well-defined selection algorithm. I am simply writing everyone who is famous (many citations, good reputation, genuine insights, academic degree etc.), everyone listed on Wikipedia, everyone I can find by googling certain keywords (AIXI, reinforcement learning etc.), authors of studies and papers, people who have been suggested to me (e.g. by lesswrong) and everyone those people link to on their sites.
(And also people who “ask” me to post their answers ;-)
ETA I think that the possibility to suggest people is kind of a reassurance that I do not cherry-pick people. Just tell me who and I will ask. I post every answer I get as long as they gave their permission to publish it.