Lila, see my comment above. Also notice the current Status quo is, as AlasdairGives mentioned, rich white people who are receiving the money through institutions, which sometimes have to pay other rich white people fees and other costs, people who are not doing EA work. Overall my best guess is that donating to institutions is about twice worse than donating to individuals in terms of cost per employee and task. It could be more.
It is important that at least one or two institutions remain affiliated with high status entities. But we no longer need to guarantee this. With Singer on LYCS, Musk and Freeman suporting Superintelligence related NGOs, FHI at Oxford and CSER and FLI burgeoning, we have solved the status question almost to satisfaction. Signalling is not our problem, efficiency is. And 90k can pay one person within institutions, and three or more outside institutions. As long as the feedback and trust mechanisms are good, the next marginal dollar should go to direct donations to individuals. Furthermore since those dollars are donations, and are not conditional on any work, they can be obtained by those who are not currently in their country of citizenship, which enormously facilitates moving for those whose efforts are better allocated elsewhere.
My concern is that feedback and trust mechanisms aren’t good. Even the best of us I think would struggle to produce quality work without a boss and coworkers and deadlines. If organizations are actually just using gratipay to pay their employees without taxes, there are some legal concerns. People don’t take kindly to allegations of tax-dodging, and if something like this were to get out, it would probably hurt donations.
Lila, see my comment above. Also notice the current Status quo is, as AlasdairGives mentioned, rich white people who are receiving the money through institutions, which sometimes have to pay other rich white people fees and other costs, people who are not doing EA work. Overall my best guess is that donating to institutions is about twice worse than donating to individuals in terms of cost per employee and task. It could be more.
It is important that at least one or two institutions remain affiliated with high status entities. But we no longer need to guarantee this. With Singer on LYCS, Musk and Freeman suporting Superintelligence related NGOs, FHI at Oxford and CSER and FLI burgeoning, we have solved the status question almost to satisfaction. Signalling is not our problem, efficiency is. And 90k can pay one person within institutions, and three or more outside institutions. As long as the feedback and trust mechanisms are good, the next marginal dollar should go to direct donations to individuals. Furthermore since those dollars are donations, and are not conditional on any work, they can be obtained by those who are not currently in their country of citizenship, which enormously facilitates moving for those whose efforts are better allocated elsewhere.
My concern is that feedback and trust mechanisms aren’t good. Even the best of us I think would struggle to produce quality work without a boss and coworkers and deadlines. If organizations are actually just using gratipay to pay their employees without taxes, there are some legal concerns. People don’t take kindly to allegations of tax-dodging, and if something like this were to get out, it would probably hurt donations.