Eschatology (THE APOCALYPSE) with Phil Torres

Episode Summary

Executive Summary: Allie Ward interviews philosopher/neuroscientist Phil Torres about eschatology: the secular study of end-of-world risks. The episode frames apocalypse thinking as evidence-based risk analysis, covering climate change, biodiversity collapse, biosecurity, AI misalignment, and moral bioenhancement, while arguing that focusing on worst-case scenarios is practical, not pessimistic, because it helps reduce existential risk and protect civilization’s future.

Main Topics: What eschatology means in a secular sense (Priority: 5/5): Torres explains that modern eschatology is the study of end-times from an empirical, naturalistic perspective rather than a purely religious one. Major existential risks to humanity (Priority: 5/5): The conversation surveys the most credible routes to extinction or irreversible civilizational decline: environmental degradation, dangerous technologies, and machine superintelligence. Climate change and biodiversity loss (Priority: 5/5): Torres argues that anthropogenic climate change and the sixth mass extinction are already underway and represent urgent, human-driven threats. Dual-use technology and biosecurity (Priority: 5/5): The episode emphasizes that technologies can solve problems but also lower the barrier for catastrophic harm, including synthetic pathogens. AI alignment and superintelligence (Priority: 4/5): Torres describes the main AI danger as misaligned goals rather than malevolence, using instrumental convergence and ant analogies to explain why advanced systems could harm humans. Ethics, reproduction, and moral enhancement (Priority: 3/5): The discussion explores antinatalism, whether having children is defensible under existential risk, and controversial proposals like oxytocin or other moral bioenhancements. How to respond: civic action and optimism (Priority: 4/5): Despite dark subject matter, Torres argues for recycling, voting, helping others, and studying worst-case scenarios as a way to improve long-term human flourishing.

Key Arguments: Existential risk should be treated as a serious scientific and philosophical field because some events could cause human extinction or irreversible loss of humanity’s potential. Climate change and biodiversity collapse are already measurable and urgent, with environmental wisdom lagging behind technological and economic growth. Technological progress is double-edged: it can cure disease and restore ecosystems, but it also democratizes destructive power and makes catastrophic misuse easier. The real AI danger is not a robot that hates humans; it is a highly capable system whose goals are not aligned with ours and which may treat humans as obstacles or raw material. A small number of malicious or reckless actors may be enough to cause enormous harm because advanced tools are becoming cheaper and easier to access. Moral bioenhancement is proposed by some philosophers as a potential response to human violence and low empathy, but it is ethically controversial and politically fraught. Studying worst-case scenarios is not pessimism; it is a strategy for preventing them and preserving the long-term future of civilization.

Data Points: Global wild vertebrate population decline: 58% - Living Planet Report comparison from 1970 to 2012 cited as evidence of biodiversity collapse. Share of species extinct: ~99.9% - Torres notes that nearly all species that have ever lived are extinct. Doomsday Clock setting (2017): 2.5 minutes to midnight - Moved forward largely due to Trump-era policy and climate/nuclear concerns. Doomsday Clock setting (1991): 17 minutes to midnight - Referenced as a more optimistic historical point. Toby Ord survival estimate: 1 in 6 - Estimate that humanity may have a one-in-six chance of surviving the century. Bostrom probability estimate: at least 20% by 2100 - Referenced as an existential-risk estimate based on objective and subjective considerations. Martin Rees estimate: 50-50 - Suggestion of a 50% chance of civilizational collapse this century. Approximate age of humanity: 200,000 years - Used to frame the current moment as unusually dangerous in human history. Recorded history span: 4,000-6,000 years - Contrasted with the much longer existence of Homo sapiens. Supervolcano recurrence: about every 50,000 years - Used to contextualize geophysical extinction-level threats. Large asteroid impact recurrence: about every 400,000 years - Cited as another rare but catastrophic natural hazard. Toba catastrophe population bottleneck: ~1,000 humans - Suggested as a past near-extinction event. Q&A apocalypse example: High-density blueberries - Andrew Sandberg’s playful physics thought experiment about Earth suddenly turning into blueberries.

Pivotal Quotes: "It's dangerous to be alive and risks are everywhere." — Phil Torres / Nick Bostrom cited by host: Opening line of Bostrom’s existential-risk paper, used to frame the episode’s subject. "If there is a technology, you know, as some people have said, if everybody around the world had an app on their phone that where they could open it up and push a button that would destroy the world. Who thinks the world would last for more than two seconds?" — Phil Torres: Explaining how democratized destructive technology could make civilization extraordinarily fragile. "The real danger is that it exceeds us. It exceeds the best possible that any member of our species could possibly achieve." — Phil Torres: Core explanation of why superintelligent AI is risky even without malicious intent.

Implications: The episode urges listeners to take existential risk seriously without succumbing to despair: support science, vote, reduce harm, and think long-term. It suggests that the future depends on aligning technology, ethics, and policy before destructive power becomes too widely available.