Biological Anthropology, with an adviser whose latest book is in philosophy of mind, the next book on information theory, the previous book on—of all things—biological anthropology, and most of his career was as a semioticist and neuroscientist.
My previous adviser was a physicist in the philosophy of physics who turned into a philosopher of mind. My main sources of inspiration are Bostrom and Russell, who defy field borders. So I’m basically studying whatever you convince me makes sense in the intersection of interestingly complex and useful for the world. Except for math, code and decision theory, which are not my comparative advantage, specially not among EAs.
I would tend to focus on AGI-related topics, though you may have specific alternate ideas that are compelling for reasons that I can’t see from a distance. In addition to AGI safety, studying political dynamics of AGI takeoff (including de novo AGI, emulations, etc.) could be valuable. I suggested a few very general AGI research topics here and here. Some broader though perhaps less important topics are here.
Biological Anthropology, with an adviser whose latest book is in philosophy of mind, the next book on information theory, the previous book on—of all things—biological anthropology, and most of his career was as a semioticist and neuroscientist. My previous adviser was a physicist in the philosophy of physics who turned into a philosopher of mind. My main sources of inspiration are Bostrom and Russell, who defy field borders. So I’m basically studying whatever you convince me makes sense in the intersection of interestingly complex and useful for the world. Except for math, code and decision theory, which are not my comparative advantage, specially not among EAs.
Thanks for asking for suggestions. :)
I would tend to focus on AGI-related topics, though you may have specific alternate ideas that are compelling for reasons that I can’t see from a distance. In addition to AGI safety, studying political dynamics of AGI takeoff (including de novo AGI, emulations, etc.) could be valuable. I suggested a few very general AGI research topics here and here. Some broader though perhaps less important topics are here.