Have you guys already done a voting survey (with whatever system seems good – STAR voting?) sent out to (1) the population of talent you want to take with you for AI research, (2) rationalists who aren't mission critical but are still valuable as interstitial social elements?
If not – at some point that's going to happen, right? This discussion seems most useful if it exists to inform a large number of people who might move what the options are, so that their preferences can be assessed numerically.
I moved to Seattle from the Bay Area, and while I love the weather and relatively sane rent and general environment around here, I think 14% of the US population having a mild form of SAD should be one of the dominating factors in decisionmaking. I have mild SAD, which I find acceptable because I'm not standing in any civilizational bottlenecks, but I would move if I worked at an EA organization and considered my work very important.
If you guys consider SAD to be a solved problem via more dakka, then I retract this and mostly recommend the area.
That's what I do.
My spouse has been learning Korean for two years on Duolingo, which also uses spaced repetition, with approximately no days missed, and learning has been slow. I think Anki is much worse as a primary language learning tool because it doesn't remix particles to give you new sentences to chew on.
Where I find it shines for me (I review my cards ~2/week):
Hmm, I looked you up on Facebook and apparently you sent me a friend request god-knows-when (which I presumably ignored because I didn't know you), which I have just accepted.