I have mixed feelings about that. I do think that what we've been talking about, global poverty, is certainly a more immediate cause that we can with greater certainty know what we're doing and know that we can have good consequences. I also think that animal welfare, and especially trying to combat factory farming of animals, which causes suffering on a vast scale, is another very practical and immediate concern that we can know we're having good consequences.
Exact quote, beginning
Peter Singer
moral philosopher and author
In a podcast published in 2026, Peter Singer said work to reduce extinction risk from super-intelligent AI was harder to assess than work on global poverty or animal welfare, while acknowledging the value of preventing extinction.
Retired Princeton University professor; philosopher and authorPublished April 20, 2026 · date of remarks not establishedMindscape, episode 351
Context and checks
Surrounding words
Quite a number of people in the effective altruism movement are interested in preventing our species becoming extinct, including trying to reduce the risk that AI will take us over in some way. 0:52:41.9 PS: I have mixed feelings about that. I do think that what we've been talking about, global poverty, is certainly a more immediate cause that we can with greater certainty know what we're doing and know that we can have good consequences. I also think that animal welfare, and especially trying to combat factory farming of animals, which causes suffering on a vast scale, is another very practical and immediate concern that we can know we're having good consequences. Whereas trying to reduce the risk of extinction, and especially the risk of extinction being caused by super-intelligent AI, is much more speculative that you're going to actually succeed. But of course the counter-argument goes, well, we're not just talking about 8 billion people when we're talking about reducing the risk of extinction.
What this quote does not say
- The printed 3242.4-second end is the next transcript timestamp and therefore an enclosing upper bound, not a word-level end time.
- The qualification concerns tractability and cause prioritization, not a denial that super-intelligent AI could cause extinction.
How this was checked
- Quote matched character-for-character against the captured page.
Reviewed by
an AI reviewer that read the whole source · an automatic character-by-character check
