In this post he built a model that estimates that preventing people from starving in close to half the countries of the world after a food shock would be net negative for the long-term future.
I no longer endorse the methodology of that post. Currently, I think:
The nearterm extinction risk from nuclear war is so low that it is better to assess interventions which aim to decrease famine deaths in a nuclear winter based on standard cost-benefit analysis (CBA), where saving lives is assumed to be good.
From a neartermist perspective, one can assume saving lives is good despite the meat eater problem.
I plan to post about both of the above in the coming weeks or months.
Which, in his defense, is technically not exactly advocating for genocide.
I have changed my mind about many things, but I have always strongly opposed killing people. Deciding not to save statistical lives is very different from killing people:
By donating 10 % of one’s income to GiveWell’s top charities, which save a life for around 5 k$, one could save many lives through one’s career.
However, not doing the above is not as bad as being a serial killer.
I also find it interesting that many people support the continuation of wars, thus plausibly increasing deaths, while claiming that can be good longterm (e.g. by supposedly preventing authoritarian countries from gaining power). I generally oppose supporting the continuation of wars based on the simple heuristic that killing people is very bad on priors (relatedly).
Thanks for the context, Juan!
I no longer endorse the methodology of that post. Currently, I think:
The nearterm extinction risk from nuclear war is so low that it is better to assess interventions which aim to decrease famine deaths in a nuclear winter based on standard cost-benefit analysis (CBA), where saving lives is assumed to be good.
From a neartermist perspective, one can assume saving lives is good despite the meat eater problem.
I plan to post about both of the above in the coming weeks or months.
I have changed my mind about many things, but I have always strongly opposed killing people. Deciding not to save statistical lives is very different from killing people:
By donating 10 % of one’s income to GiveWell’s top charities, which save a life for around 5 k$, one could save many lives through one’s career.
However, not doing the above is not as bad as being a serial killer.
I also find it interesting that many people support the continuation of wars, thus plausibly increasing deaths, while claiming that can be good longterm (e.g. by supposedly preventing authoritarian countries from gaining power). I generally oppose supporting the continuation of wars based on the simple heuristic that killing people is very bad on priors (relatedly).