I got the impression that their new, general-purpose pool would still be fairly longtermist, but it’s possible they will have to make sacrifices.
To clarify it’s not that I don’t think they would be “longtermist” it’s more that I think they may have to give to longtermist options that “seem intuitively good to a non-EA”, e.g. giving to an established organisation like MIRI or CHAI, rather than give to longtermist options that may be better on the margin but seem a bit weirder at first glance like “buying out some clever person so they have more time to do some research”.
That pretty much gets to the heart of my suspected difference between Longview and LTFF—I think LTFF funds a lot of individuals that may struggle to get funding from elsewhere whereas Longview tends to fund organisations that may struggle a lot less—although I do see on their website that they funded Paul Slovic (but he seems a distinguished academic so may have been able to get funding elsewhere).
I think that looking at their track record is only partially representative. They used to follow a structure where they would recommend donation opportunities to particular clients. Recently they’ve set up a fund that works differently; people would donate to the fund, then the fund will make donations at their will. My guess is that this will help a bit around this issue, but not completely. (Maybe they’ll even be extra conservative, to prove to donors that they will match their preferences.)
Another (minor) point is that Longview’s donations can be fungible with LTFF. If they spend $300K on something that LTFF would have otherwise spent money on, then the LTFF would have $300K more to spend on whatever it wants. So if Longview can donate to, say, only 90% of interesting causes, up to $10Mil per year, the last 10% might not be that big of a deal.
To clarify it’s not that I don’t think they would be “longtermist” it’s more that I think they may have to give to longtermist options that “seem intuitively good to a non-EA”, e.g. giving to an established organisation like MIRI or CHAI, rather than give to longtermist options that may be better on the margin but seem a bit weirder at first glance like “buying out some clever person so they have more time to do some research”.
That pretty much gets to the heart of my suspected difference between Longview and LTFF—I think LTFF funds a lot of individuals that may struggle to get funding from elsewhere whereas Longview tends to fund organisations that may struggle a lot less—although I do see on their website that they funded Paul Slovic (but he seems a distinguished academic so may have been able to get funding elsewhere).
I think that looking at their track record is only partially representative. They used to follow a structure where they would recommend donation opportunities to particular clients. Recently they’ve set up a fund that works differently; people would donate to the fund, then the fund will make donations at their will. My guess is that this will help a bit around this issue, but not completely. (Maybe they’ll even be extra conservative, to prove to donors that they will match their preferences.)
Another (minor) point is that Longview’s donations can be fungible with LTFF. If they spend $300K on something that LTFF would have otherwise spent money on, then the LTFF would have $300K more to spend on whatever it wants. So if Longview can donate to, say, only 90% of interesting causes, up to $10Mil per year, the last 10% might not be that big of a deal.