#208 — Existential Risk

Episode Summary

Executive Summary: Sam Harris introduces a conversation with philosopher Toby Ord about effective altruism and existential risk, framing COVID as a warning shot rather than an existential threat. The discussion links moral impartiality, charitable effectiveness, and the future of humanity, emphasizing how technological power has outpaced wisdom and created a historically brief but crucial “precipice” period.

Main Topics: Podcast framing and format clarification: Harris explains subscriber access, distinguishes the podcast’s two formats, and clarifies that recent solo and topical episodes are meant to serve different purposes. Effective altruism and moral impartiality: Ord explains how his work, influenced by Singer and others, focuses on helping as much as possible by prioritizing interventions that produce the most benefit per dollar, regardless of emotional appeal or geographic distance. Automating generosity and donor psychology: The conversation stresses that people should use commitment devices and career choices to overcome weak intuitions, burnout, and sentimentality so their giving remains sustained and maximally effective over time. Existential risk and the concept of the precipice: Ord defines existential risk as threats that could extinguish humanity or permanently block its future, arguing that modern civilization has entered a narrow historical window where power has surged faster than wisdom. Natural vs. anthropogenic risks: Ord contrasts low-probability natural threats like asteroids and supernovas with much larger human-caused risks such as nuclear war, engineered pandemics, and other technological catastrophes. Moral significance of the future: A major theme is that losing humanity would erase potentially millions of future generations, making even low-probability risks morally enormous because they threaten vast unrealized value. Population ethics and Parfit’s influence: The discussion invokes Derek Parfit’s work on how to weigh extinction versus partial catastrophe, highlighting the moral importance of the last portion of humanity because it determines whether the future exists at all.

Key Arguments: Physical distance should not significantly affect moral concern; the same logic that compels helping a nearby drowning child should extend to helping distant strangers. Donation effectiveness varies enormously; some interventions are about 10,000 times more effective than others, so where money goes can matter more than whether one gives. Effective altruism is not only about being generous but about making generosity actually count through data, long-term commitments, and careful allocation of resources. People should automate charitable behavior because human motivation is unstable; long-term pledges prevent burnout and maximize lifetime impact. Earning to give can be the best career path for highly suited people, because high income can be converted into much larger humanitarian impact. Natural existential risks are real but likely much smaller than anthropogenic ones; human technology has made catastrophe more plausible than the universe’s background threats. The current era is a short-lived historical bottleneck: humanity now has the power to destroy itself without having yet developed equivalent wisdom to control that power. The future may contain the overwhelming majority of human value, so extinction or civilizational collapse would not just be bad now but would erase enormous possible good. Existential risk matters not only because of current deaths, but because it ends the entire future, the achievements yet to come, and the intergenerational project humanity inherited. Even non-consequentialists should care about effective altruism when the options do not violate side constraints, since everyone can still agree on reducing harm efficiently.

Data Points: Podcast episode 207: Racism and police violence episode - Harris references a prior controversial solo episode and its annotated transcript. Global pandemics historically: 1918 flu killed about 3% of the world - Ord compares COVID-19 with past pandemics to show current events are serious but not unprecedented. Black Death mortality: About 25% to 50% of Europe; roughly 10% of the world - Used as historical perspective on pandemic severity. Charitable giving pledges: More than 4,000 people - Giving What We Can members who have pledged to donate at least 10% of income. Money pledged: More than $1 billion - Total amount pledged by members of Giving What We Can. Money already given: More than $100 million - Actual donations made through the pledge community. Effectiveness gap in health interventions: About 10,000 times - Ord cites DCP2 analysis showing best global health interventions vastly outperform worst ones. Random intervention ratio: About 100 times more effective on average - From Ord’s statistical work on health aid interventions. Low-cost life saving estimate: Around $1,000 US - Approximate cost of saving a life in poor countries discussed in Singer’s analogy. Asteroid impact probability: About 1 in 120,000 in the next century - Astronomers’ estimate for a kilometer-or-larger asteroid impact. Asteroid extinction risk: About 1 in 1,000,000 in the next century - Ord’s rough estimate for existential catastrophe from asteroid impact. Supernova existential risk: About 1 in 1,000,000,000 in the next century - Ord’s estimate for a nearby supernova causing existential catastrophe. Natural existential risk upper bound: Lower than about 1 in 200 per century - Bound inferred from the fact that Homo sapiens has survived about 200,000 years. Most likely natural-risk bound: Below about 1 in 2,000 per century - Ord’s more confident estimate after multiple lines of reasoning. Anthropogenic existential risk: About 1 in 6 over the next century - Ord’s rough estimate for human-caused existential catastrophe. Human history: About 2,000 centuries of Homo sapiens - Ord emphasizes the vast stretch of human history relative to the present. Species lifespan comparison: Around 1 million years - Ord notes many species last roughly this long, suggesting humanity could too. Earth’s habitable future: About 500 million to 1 billion years - Potential future duration if humanity survives and remains on Earth. Funding threshold example: 10% of income - Core Giving What We Can pledge level discussed throughout the exchange. Personal giving example: Over a quarter of everything earned so far - Ord describes his own cumulative giving.

Pivotal Quotes: "I see this as a warning shot." — Toby Ord: Ord describes COVID-19 as a signal that humanity remains vulnerable to catastrophic events. "the power to destroy ourselves without the wisdom to ensure that we don't" — Toby Ord: Ord summarizes the central danger of the modern era in his “precipice” thesis. "if we’re gone, things will just meander on their own course" — Toby Ord: Ord explains the moral and civilizational significance of preserving human agency and moral progress.

Implications: Listeners are urged to think more impartially, give more effectively, and take existential threats seriously. The long-term future may depend on whether humanity can align its growing power with better institutions, wiser norms, and sustained risk reduction.