Human extinction would mean the end of humanity’s achievements, culture, and future potential. According to some ethical views, this would be a terrible outcome for humanity. But what are the public’s beliefs about human extinction? And how much do people prioritize preventing extinction over other societal issues? Across five empirical studies (N = 2,147; U.S. and China), we find that people consider extinction prevention a societal priority and deserving of greatly increased societal resources. However, despite estimating the likelihood of human extinction to be 5% this century, people believe that the chances would need to be around 30% for it to be the very highest priority (U.S. medians). In line with this, people consider extinction prevention to be only one among several important societal issues. We also find that people’s judgments about the relative importance of extinction prevention appear relatively fixed and hard to change by reason-based interventions.
The other societal issues they scored as more important were: healthcare, poverty, education, homelessness. I think this should be a pretty large update vs. how much public pressure there will be to manage x-risks from AI, vs. much less effective, and potentially even misguided, interventions. More explicitly, I expect it will be a long time before the public’s stated preferences for AI safety from polls turn into actionable policies / votes; and when they do, a lot of the time those policies will probably not be that effective / will be misguided, as seems to already be happening to some degree: https://x.com/Thomas_Woodside/status/2025414535570493446.
From ’Lay beliefs about the badness, likelihood, and importance of human extinction’
https://www.nature.com/articles/s41598-026-39070-w.
The other societal issues they scored as more important were: healthcare, poverty, education, homelessness.
I think this should be a pretty large update vs. how much public pressure there will be to manage x-risks from AI, vs. much less effective, and potentially even misguided, interventions. More explicitly, I expect it will be a long time before the public’s stated preferences for AI safety from polls turn into actionable policies / votes; and when they do, a lot of the time those policies will probably not be that effective / will be misguided, as seems to already be happening to some degree: https://x.com/Thomas_Woodside/status/2025414535570493446.