I think this is not only right but very, clearly, obviously right in such a way that the alternative feels a bit to me like a troll even though I know it isn’t. My perhaps myopic reason why is that if investing in such a manner as to produce outsize yield were that easy, while also mitigating the downside risk presupposed by the thought experiment zeroing out the distribution of opportunity costs, everyone nominally in a position to do what Will is saying should instead go and launch a competitor to Jane Street.
I mean, to steelman Will's point a bit, some very obvious investments to AI safety and AGI-pilled people include the likes of Anthropic and NVIDIA and MU, and funds like Situational Awareness and VARA; the last couple years of returns. You don't necessarily need yield beyond these to drastically outperform SP500.
(our point is that, as great as these gains will be, philanthropic giving is going to be much much greater)
Right and I agree with that, very bigly. My point is just that if it were self-evident that eg NVidia were just going to be a money printer going brrrrrrrrrrr for the next N years, more people would just go and buy X amount of e.g. NVidia.
I THINK NVidia (etc) are great investments. I HOLD as much exposure to this space as I can get my hands on. But that’s super different than just *postulating* as *deterministically known* that its yield is a) obvious/easy to model and b) is obviously/clearly higher than the also-extremely-difficult-to-model investment in Philanthropy Of Whatever Space-Specific Bespoke Form.
Which nets out to an ultimately pretty myopic observation that the thing to do is look at a prospective investment in front of you and evaluate the merits of it in a fairly idiosyncratic, subject-specific, domain-specific, operations-specific way. I just think a lot of the galaxy braining here has been very bad for philanthropic yields the last ~15 or so years, which is why I have a burr in my bonnet about this.
Ah, my bad, I was thinking about investing = donating, which obviously makes no sense. As in, the opposite was not deploying cash to non-profits, instead of not investing [in for-profits].
I think this is not only right but very, clearly, obviously right in such a way that the alternative feels a bit to me like a troll even though I know it isn’t. My perhaps myopic reason why is that if investing in such a manner as to produce outsize yield were that easy, while also mitigating the downside risk presupposed by the thought experiment zeroing out the distribution of opportunity costs, everyone nominally in a position to do what Will is saying should instead go and launch a competitor to Jane Street.
I mean, to steelman Will's point a bit, some very obvious investments to AI safety and AGI-pilled people include the likes of Anthropic and NVIDIA and MU, and funds like Situational Awareness and VARA; the last couple years of returns. You don't necessarily need yield beyond these to drastically outperform SP500.
(our point is that, as great as these gains will be, philanthropic giving is going to be much much greater)
Right and I agree with that, very bigly. My point is just that if it were self-evident that eg NVidia were just going to be a money printer going brrrrrrrrrrr for the next N years, more people would just go and buy X amount of e.g. NVidia.
I THINK NVidia (etc) are great investments. I HOLD as much exposure to this space as I can get my hands on. But that’s super different than just *postulating* as *deterministically known* that its yield is a) obvious/easy to model and b) is obviously/clearly higher than the also-extremely-difficult-to-model investment in Philanthropy Of Whatever Space-Specific Bespoke Form.
Which nets out to an ultimately pretty myopic observation that the thing to do is look at a prospective investment in front of you and evaluate the merits of it in a fairly idiosyncratic, subject-specific, domain-specific, operations-specific way. I just think a lot of the galaxy braining here has been very bad for philanthropic yields the last ~15 or so years, which is why I have a burr in my bonnet about this.
> Will MacAskill recently argued that large donors should invest money now, and wait until the intelligence explosion to give away their money.
I'm guessing you meant to say "argued that large donors should NOT invest money now"?
Ah no, he's saying they should invest it in stocks now and donate it later.
Ah, my bad, I was thinking about investing = donating, which obviously makes no sense. As in, the opposite was not deploying cash to non-profits, instead of not investing [in for-profits].