On disputes within EA
Hi Aurora. Re EA disputes, I think things do go much deeper than just arguing over the most important problem or the best strategy. IIRC, I've never written anything that simply argued "AI risk is more important than global poverty", or something along those lines. I could if I wanted to, it just hasn't felt that important. (Actually, part of why I started my Facebook group was to discuss a wider range of problems and strategies than EA does, even though I expect to disagree with many of them.) Social stuff here is vague and fuzzy, especially seeing it from the inside, but it appears to me that much of EA (certainly far from all, but much) is driven by deference to authority and dominance/submission dynamics; in Haidt's theory the "authority" axis, rather than the "care" axis. This is something that me and many other nerd types find fundamentally dark, in a different way than just thinking effort is being wasted.
I don't expect this to be a knockdown argument, but just as an illustration, let's suppose that everyone agrees that global poverty is the most important problem, and donating to AMF is the best strategy for addressing it. (I don't in fact believe this, but let's just suppose.) If we've decided that this is literally the world's most important problem, and we're going to spend a significant fraction of our lifetime time and money to fix it, one obvious question to ask is "how much progress are we making?". That is, of all the people in Africa whom nets could be distributed to, how many is AMF reaching at its current level of funding? To my knowledge, literally no one anywhere has ever tried to calculate this number. (At a very rough guess, I think it's somewhere around 20%.) This seems extremely strange. It's as if the US entered World War II, with war bonds and ration stamps and Victory Gardens and the draft and all the rest of it, but nobody anywhere ever asked whether the US was winning, even many years later. (In actual WWII, battle news was almost always the top headline, per the NYT archives: https://www.nytimes.com/store/front-page-new-york-times-reprint-nskeep.html.)