William MacAskill on Effective Altruism, Moral Progress, and Cultural Innovation

Episode Summary

Executive Summary: Will MacAskill argues for a nuanced, partly pluralist longtermism: prioritize existential risk reduction, civilization’s values, and high-leverage institutions, while rejecting extreme utilitarian edge cases and some expected-value gambles. He defends concern for animals, future people, and culture, but warns against over-optimization and speculative “crazy town” reasoning.

Main Topics: Moral theory: utilitarianism, pluralism, and limits (Priority: 5/5): MacAskill says he is neither a pure utilitarian nor a classical pluralist, instead combining multiple moral considerations with probabilistic weightings. He defends utilitarian reasoning at mid-scale but resists applying it uncritically to very small or very large domains. Animal welfare and wild animal suffering (Priority: 5/5): He discusses hedonistic vs preference utilitarianism, wild-animal welfare, and why the moral status of animals in nature is uncertain. He is open to research and cautious intervention rather than sweeping ecosystem destruction. Longtermism, existential risk, and hingey eras (Priority: 5/5): A major theme is that the far future contains vastly more value than the present, so reducing extinction risk and steering civilization toward good values matters enormously. He argues we may be living in an unusually influential historical window. Decision theory, expected value, and paradoxes (Priority: 4/5): MacAskill rejects some extreme expected-value conclusions, such as 51/49 world-doubling gambles, and treats paradoxes like Pascal’s wager, the St. Petersburg paradox, and the repugnant conclusion as evidence that moral reasoning has scope limits. Effective altruism as cultural innovation (Priority: 4/5): He and Cowen agree EA is not just about numerical optimization or cost-effectiveness; it is also a cultural project that fosters truth-seeking, impartiality, and better institutions. MacAskill acknowledges culture may be underweighted in EA circles. Practical application: careers, philanthropy, and institutions (Priority: 4/5): The conversation moves to concrete advice for lawyers, donors, universities, opera, and world-government constitutions. MacAskill favors leveraged, future-shaping interventions over generic giving or low-impact prestige philanthropy. Personal choices, reproduction, and next projects (Priority: 3/5): He explains why he currently does not plan to have children, given time constraints and alternative meaning-producing projects, and outlines future plans: The Future Fund, more books, or possibly a new college/university model.

Key Arguments: The best moral theory may be simple at the meta-level, but our practical moral reasoning should combine multiple views and degrees of confidence rather than forcing one rigid framework. Wild animal suffering is a real possibility, but because nature is poorly understood, the right response is more research and caution, not reckless intervention like “paving over the jungle.” Even if one accepts longtermist premises, there are strong reasons to prioritize existential risk, value formation, and institutions over local or short-run optimization. Extreme expected-value reasoning breaks down in high-stakes, low-probability cases; some purportedly rational gambles are too brittle to endorse. The repugnant conclusion and population ethics show that moral philosophy hits genuine limits, but that does not imply abandoning all consequentialist reasoning—just restricting its scope. Human-originating civilization is not obviously better than alien-originating civilization in principle, but empirically humans remain the likely sole path to a flourishing future if we go extinct. Democracy, liberalism, trust, technology, and growth are not just economic variables; they may be among the best available levers for a good long-term future. Effective altruism attracts smart young people partly because it combines intellectual difficulty, moral seriousness, and a socially “safe” altruistic niche for highly educated people.

Data Points: Book publication date: August 16 - The new book What We Owe the Future is said to publish on August 16. Future Fund money moved: about $140 million - MacAskill says he has been helping Sam Bankman-Fried launch the Future Fund and it has moved roughly this amount this year. Hypothetical gamble odds: 51% / 49% - Tyler proposes a demon bet that doubles the world with 51% probability and deletes it with 49% probability. Alternative gamble odds: 90-10 - MacAskill says he would need the probabilities to be much wider before taking the gamble. Population ethics scale example: 10 trillion lives - He uses this as an example of many wonderful lives in the repugnant conclusion discussion. Fertility example: 10% increase in world population - He says a good fertility program might raise population by this amount, but future-value stakes dwarf it. Future lives scale: trillions upon trillions - Used to emphasize that the long-term future dominates present population considerations. Growth projection: 2% per year for 10,000 years - He argues that sustained growth at this rate would imply absurdly vast output, showing why growth must eventually slow. Economic output example: 10 to the 87 worlds' worth - His estimate of output from 2% annual growth over 10,000 years. Atom comparison: 10 to the 67 atoms within 10,000 light years - Used to illustrate the extreme scale mismatch between projected output and available physical matter. Historical window: roughly 200 years - He says we have had anything close to current technological progress only for about this long.

Pivotal Quotes: "I think we should have some degree of belief in a variety of moral views and take a compromise between them." — Will MacAskill: Explaining why he is not a pure utilitarian and how he approaches moral reasoning. "I worry a bit this is verging into the absurd." — Will MacAskill: Said while discussing wild animal welfare, ants, and how much confidence we should have in speculative ethical claims. "We should be more ambitious than that with our model reasoning." — Will MacAskill: His reply to the idea that utilitarian reasoning should simply be abandoned at large scales.

Implications: Listeners should take longtermism seriously but avoid simplistic optimization. The episode argues for leveraging institutions, culture, and caution under deep uncertainty rather than chasing clean but brittle moral formulas.