Evaluator · round one
Writes Don't Worry About the Vase; founder of Balsa Research; SFF recommender in four rounds.
$8.82M of Zvi's recommendations funded
$5.42M granted through Zvi's order
My top focus is AI safety, and anything I think moves the needle there in good directions, but I am open to a wide variety of cause areas and approaches.
I write my blog how I think. If you want to allocate based on the kind of thinking you see on my blog or when you talk to me, fund through me. If not, don't do that.
I believe in using my unique position, including local knowledge, to identify opportunity. Lightcone Commons gives me the opportunity to fund things that are relatively illegible, and to rely more on my firsthand knowledge, versus my role at SFF where we funded relatively legible things.
I believe we still have to give people hope for the future and have a distributed portfolio of the things I care about. This includes things that would ultimately prove irrelevant if superintelligence arrives soon.
No matter what I say on the lists below, anything sufficiently good will get me interested.
That said, in the 1st round of Lightcone Commons the vast majority of applications and good applications focused on aspects of AI safety, so my funding list was highly concentrated there.
While AI and making it go well is my top priority, things I am interested in that are not AI include but are not limited to (although much of this was absent from our applications):
Things in AI I am interested in include but are not limited to:
Things I am relatively not interested in at this time (but am not fully 0% interested in), where I expect anything that would pass my funding bar to already be funded otherwise, although my bar is not infinitely high if something looks neglected by others and I know why and like it anyway, include:
I use my own technical evaluations, and discard anything I don't think is promising.
