I don’t understand the incentives of the “oracular funder.” What are some examples of organisations/individuals that would be oracular funders? Where would they have gotten their money? What would make them want to participate in this idea? What are their alternatives (in how they spend their money)? Would it be fair to say that if they decided not to take part in this, then the idea wouldn’t work?
Not the author, but this is my understanding assuming the idea holds:
OpenPhil/FTX/etc currently spend a lot of time & effort on evaluating grants. The idea here is that it’s easier to evaluate a project after it’s finished, e.g. “the area where bednets were given out had X% lower mortality, lives are valued at Y$, so this bednet intervention was worth Z$”—but current status quo is that you have to predict the value of X when evaluating the grant, and make a prediction on the likelihood that it works at all. (technically you’d predict a probability distribution over different values of X), which is much harder than measuring X after the intervention.
Investors (which are VC’s and other people) take over the role of predicting the future, and oracular funders (openphil/FTX/etc) get an easier job and can make more accurate donations—making a larger impact than they would do otherwise. And good charity entrepenurs are rewarded more accurately, making more good for the world than if money was wasted on charities that at first glance seemed like a good idea but in fact has bad expected value.
I don’t understand the incentives of the “oracular funder.” What are some examples of organisations/individuals that would be oracular funders? Where would they have gotten their money? What would make them want to participate in this idea? What are their alternatives (in how they spend their money)? Would it be fair to say that if they decided not to take part in this, then the idea wouldn’t work?
I appreciate your clarification!
Not the author, but this is my understanding assuming the idea holds:
OpenPhil/FTX/etc currently spend a lot of time & effort on evaluating grants. The idea here is that it’s easier to evaluate a project after it’s finished, e.g. “the area where bednets were given out had X% lower mortality, lives are valued at Y$, so this bednet intervention was worth Z$”—but current status quo is that you have to predict the value of X when evaluating the grant, and make a prediction on the likelihood that it works at all. (technically you’d predict a probability distribution over different values of X), which is much harder than measuring X after the intervention.
Investors (which are VC’s and other people) take over the role of predicting the future, and oracular funders (openphil/FTX/etc) get an easier job and can make more accurate donations—making a larger impact than they would do otherwise. And good charity entrepenurs are rewarded more accurately, making more good for the world than if money was wasted on charities that at first glance seemed like a good idea but in fact has bad expected value.