Effective Altruism: An Introduction – 80,000 Hours (April 2021) Five: Prof Will MacAskill on moral uncertainty, utilitarianism & how to avoid being a moral monster
16 snips
Apr 12, 2021 AI Snips
Chapters
Transcript
Episode notes
How Effective Altruism Began
- Will MacAskill describes EA as a confluence of GiveWell, LessWrong and Giving What We Can.
- He recounts founding projects like Giving What We Can and 80,000 Hours and their unexpected early growth.
Treat Moral Uncertainty Like Risk
- Moral uncertainty should mirror empirical uncertainty: use probabilities and expected value reasoning.
- Acting on a single favored moral theory resembles the 'football fan' model and is irrational under uncertainty.
Why Moral Uncertainty Is Messy
- Moral uncertainty creates hard problems: intertheoretic comparisons, non-cardinal theories, and fanaticism.
- Small credences in views with infinite or incommensurable values can break expected value calculations.