You wake up to find that you have sole control over the world's biggest philanthropic fund, all US$25bn of it. How might you spend it -- a back of the envelope guesstimate?

I'm particularly interested in how you might split it between causes very likely to reduce suffering (promote veganism, condoms to Africa?), causes that might help reduce it (humane insect pesticides? research into reducing wild animal suffering?), and low-probability, high-payoff causes like reducing the risk of lab-universe suffering. Do you spread the risk? Or spend half developing AI to tell you how best to spend the other half?

I would go for anti-speciesist projects. It's far easier to reduce suffering and to take animals into account in a less speciesist society.

I would found an organization for systematically analyzing ways to most cost-effectively reduce expected suffering in the multiverse and ask it to produce recommendations, waiting to make further expenditures until hearing its advice. AI isn't at a point where it can helpfully do such analysis on its own, so the organization would consist of human experts -- occasionally using computer models, perhaps, but also just pencil-and-paper analysis and philosophy. I think the Singularity Institute for Artificial Intelligence comes close to such an organization on a small scale, although it has other objectives besides reducing suffering.

I agree with Daniel that, in general, helping animals tends to be more cost-effective than helping humans, because there are a whole lot more animals, and the low-hanging fruit in this area hasn't been picked as much. The same is even more true for the "low-probability, high-payoff causes," so I expect that the group of experts would end up recommending those projects.

I'm a long way from Gates' position, so I here are my very preliminary thoughts:
I would look into charity evaluation to see what performs a role most similar to what Alan described. Then I'd figure out whether it was adequate or a better - more utilitarian - evaluation was needed.

Also, perhaps it would significantly promote utilitarianism for an organisation to give an annual award to the person or organisation who reduces suffering the most in that year. There is no such award yet as far as I know and perhaps it would do well to found one.

Isn't that also what Charity International does, Alan?

I wonder whether instituting a prize, or a number of prizes, for the best paper or report addressing Alan's question would be more cost-effective than either donating the money to an existing organization or using it to create a new organization altogether. Considered as research institutions--as opposed to advocacy groups--such organizations seem to do little more than funding individual researchers; and many of these researchers would probably be motivated to produce work of comparable quality if given the incentive of a cash prize instead of being paid directly.

One reason for supporting the organization is to provide potential donors the opportunity to donate; in the absence of such an institutional framework, many of those donors would end up supporting a very different charity instead, rather than using their funds to help researchers directly. Here is an analogy: suppose you believe, like Tyler Cowen, that the best way to help people in the third world is to travel to a third world country, pick random people on the street, and hand each a few hundred dollars. If there were no organizations like Oxfam or Unicef, or the ones that existed were underfunded or cost-ineffective, you may still decide instead to create one such organization yourself, since many people that would be willing to support the organization would not be willing to help third world individuals in the manner Cowen advocates.

In relation to this, it now occurs to me that Alan could have used the $12,000 he donated to the SIAI to create one such prize, which he could have advertised not only among the Institute's fellows, but also among professional philosophers, economists and researchers in other relevant fields. Alan, have you thought about this possibility?

Pablo Stafforini wrote:In relation to this, it now occurs to me that Alan could have used the $12,000 he donated to the SIAI to create one such prize, which he could have advertised not only among the Institute's fellows, but also among professional philosophers, economists and researchers in other relevant fields. Alan, have you thought about this possibility?

In my case, I chose SIAI precisely because I needed the "institutional framework" you mentioned in order to get a matching-funds contribution (as well as a tax deduction). Absent those pressures, I probably would have chosen more non-standard options like those you describe. (In fact, I actually would have chosen meme-propagation on the issue of wild-animal suffering, but again, there's no existing institution for such a cause.)

Just funding very-low-cost researchers like college students also seems like a promising approach. Lots of college students work $8-hour jobs on campus to "pay their way," even though they might start making, say, $25 an hour when they graduate. So the population of underpaid college, or even high-school, students seems rather promising.