Episode Summary
Executive Summary: Rob Wiblin interviews Nick Beckstead about why the far future matters, how to think about population ethics, and where philanthropy can most reduce catastrophic risk. Beckstead defends prioritizing AI safety, biosecurity, and community-building in effective altruism, while reflecting on his career, grantmaking judgments, and the tradeoffs between broad vs narrow interventions and fast vs slow progress.
Main Topics: Why the far future dominates moral concern: Beckstead argues that because the future could contain immense numbers of valuable lives over millions or billions of years, preventing catastrophic loss and improving civilization’s trajectory may outweigh most present-focused priorities. Population ethics and objections to longtermism: The conversation explores person-affecting views, discounting future welfare, and the claim that we cannot meaningfully influence the future. Beckstead explains why he assigns low credence to these objections. Priority global catastrophic risks: Beckstead highlights advanced AI and biosecurity/pandemic preparedness as the highest-priority risk areas due to high expected harm, neglect, and some tractability. Broad vs narrow interventions: They compare broad interventions like improving judgment, institutions, and economic growth with narrow targeted work like AI strategy or specific biosecurity measures, with Beckstead increasingly favoring targeted approaches. Effective altruism community strategy: Beckstead wants the EA community to become more full-time, specialized, and career-oriented, emphasizing high-impact roles over side-project participation or earning to give alone. Career path and advice for high-impact work: He offers practical guidance for students and early-career people interested in Open Phil, AI safety, biosecurity, animal welfare, and government roles relevant to longtermist causes. Grantmaking, science, and personal development: The interview covers Open Phil grants (Target Malaria, Impossible Foods, Ed Boyden’s lab), Beckstead’s decision-making methods, and how philosophy, Bayesian thinking, and skepticism shaped his judgment.
Key Arguments: The far future likely contains most of the possible value, so even small changes in civilization’s survival or trajectory can matter enormously. Time discounting is inappropriate for direct moral value; it may make sense for capital or investment, but not for judging the goodness of future lives. Person-affecting views struggle to explain why preventing a terrible future is valuable while creating happy future people is not; they also produce implausible implications for world-destruction cases. We can influence catastrophic risk through coordinated work; thousands of smart people working on hard scientific and policy problems can shift outcomes. AI safety needs both technical alignment research and strategic/political thinking about deployment and governance. Biosecurity and pandemic preparedness are underfunded and can benefit from technical progress (countermeasures, surveillance) and policy work. The EA community should professionalize: more people should go full-time on high-impact problems rather than treating them as hobbies or online discussion topics. Broad interventions like growth or forecasting matter, but targeted interventions may be better when we have a strong inside view about a specific risk and can act on it. Philanthropy can help but often plays a limited role compared with direct scientific, policy, or organizational work; being in the trenches may create more breakthroughs than funding alone.
Data Points: Years since grants/recommendations: About 1.5 years - Beckstead says many Open Phil grants he recommended are too recent to judge clearly. Open Phil science bets: 3 major examples - He names Target Malaria, Impossible Foods, and Ed Boyden’s MIT lab among the biggest science-program bets. Potential future timescale: Hundreds of millions to billions (possibly trillions) of years - Used to illustrate the vast scale of possible future value. Open Phil size: 20s - Wiblin corrects an earlier rough estimate, saying Open Phil is now in the 20s of staff. EA community role capacity: Not 2,000 people - Beckstead argues the community does not have a scalable use for thousands of people in AI strategy or technical safety. Powerbroker audiobook length: 60 hours - Beckstead mentions The Power Broker as a 60-hour audiobook he is 40 hours through. Career-history starting point: 2006-2009 - He describes choices in undergrad/grad school and early exposure to EA ideas around this period. Clean meat investigation effort: A couple hundred hours - Open Phil’s exploratory analysis of cultured meat tractability was limited in scope.
Pivotal Quotes: "almost all of the potential value is in the distant future" — Nick Beckstead: Explaining why longtermism is so important. "we're in favor of making people happy, not making happy people" — Nick Beckstead: Summarizing the intuition behind person-affecting views and his critique of them. "I would love to see people in the effective altruist community thinking through [AI strategy]" — Nick Beckstead: On the need for more strategic work beyond technical AI alignment.
Implications: The episode reinforces longtermism as a practical agenda: fund neglected high-risk areas, professionalize EA talent pipelines, and pair technical research with strategy. For listeners, the clearest paths are AI safety, biosecurity, EA operations, and policy roles.