Advertise on podcast: Future of Life Institute Podcast

Carina Prunkl is a researcher at Inria. She joins the podcast to discuss how to assess the capabilities and risks of general-purpose AI. We examine why systems can solve hard coding and math problems yet still fail at simple tasks, why pre-deployment tests often miss real-world behavior, and how faster capability gains can increase misuse risks. The conversation also covers de-skilling, red teaming, layered safeguards, and warning signs that AIs might undermine oversight. LINKS: Carina Prunkl personal website CHAPTERS: (00:00) Episode Preview (01:04) Introducing the report (02:10) Jagged frontier capabilities (05:29) Formal reasoning progress (12:36) Risks and evaluation science (19:00) Funding evaluation capacity (24:03) Autonomy and de-skilling (31:32) Authenticity and AI companions (41:00) Defense in depth methods (48:34) Loss of control risks (53:16) Where to read report PRODUCED BY: https://aipodcast.ing SOCIAL LINKS: Website: https://podcast.futureoflife.org Twitter (FLI): https://x.com/FLI_org Twitter (Gus): https://x.com/gusdocker LinkedIn: https://www.linkedin.com/company/future-of-life-institute/ YouTube: https://www.youtube.com/channel/UC-rCCy3FQ-GItDimSR9lhzw/ Apple: https://geo.itunes.apple.com/us/podcast/id1170991978 Spotify: https://open.spotify.com/show/2Op1WO3gwVwCrYHg4eoGyP

Li-Lian Ang is a team member at Blue Dot Impact. She joins the podcast to discuss how society can build a workforce to protect humanity from AI risks. The conversation covers engineered pandemics, AI-enabled cyber attacks, job loss and disempowerment, and power concentration in firms or AI systems. We also examine Blue Dot's defense-in-depth framework and how individuals can navigate rapid, uncertain AI progress. LINKS: Li-Lian Ang personal siteBlue Dot Impact organization siteCHAPTERS: (00:00) Episode Preview (00:48) Blue dot beginnings (03:04) Evolving AI risk concerns (06:20) AI agents in cyber (15:52) Gradual disempowerment and jobs (23:26) Aligning AI with humans (29:08) Power concentration and misuse (34:52) Influencing frontier AI labs (43:05) Uncertain timelines and strategy (50:18) Writing, AI, and action PRODUCED BY: https://aipodcast.ing SOCIAL LINKS: Website: https://podcast.futureoflife.org Twitter (FLI): https://x.com/FLI_org Twitter (Gus): https://x.com/gusdocker LinkedIn: https://www.linkedin.com/company/future-of-life-institute/ YouTube: https://www.youtube.com/channel/UC-rCCy3FQ-GItDimSR9lhzw/ Apple: https://geo.itunes.apple.com/us/podcast/id1170991978 Spotify: https://open.spotify.com/show/2Op1WO3gwVwCrYHg4eoGyP

Emilia Javorsky is a physician-scientist and Director of the Futures Program at the Future of Life Institute.  She joins the podcast to discuss her newly published essay on AI and cancer. She challenges tech claims that superintelligence will cure cancer, explaining why biology’s complexity, poor data, and misaligned incentives are bigger bottlenecks than raw intelligence. The conversation covers realistic roles for AI in drug discovery, clinical trials, and cutting unnecessary medical bureaucracy. You can read the full essay at: curecancer.ai CHAPTERS: (00:00) Episode Preview (01:10) Introduction and essay motivation (06:30) Intelligence vs data bottlenecks (19:03) Cancer's complexity and heterogeneity (29:05) Measurement, health, and homeostasis (41:41) AI in drug development (50:13) Regulation, FDA, and innovation (01:02:58) Practical paths toward cures PRODUCED BY: https://aipodcast.ing SOCIAL LINKS: Website: https://podcast.futureoflife.org Twitter (FLI): https://x.com/FLI_org Twitter (Gus): https://x.com/gusdocker LinkedIn: https://www.linkedin.com/company/future-of-life-institute/ YouTube: https://www.youtube.com/channel/UC-rCCy3FQ-GItDimSR9lhzw/ Apple: https://geo.itunes.apple.com/us/podcast/id1170991978 Spotify: https://open.spotify.com/show/2Op1WO3gwVwCrYHg4eoGyP

Tech executives have promised that AI will cure cancer. The reality is more complicated — and more hopeful. This essay examines where AI genuinely accelerates cancer research, where the promises fall short, and what researchers, policymakers, and funders need to do next. You can read the full essay at: curecancer.ai CHAPTERS: (00:00) Essay Preview (00:54) How AI Can, and Can't, Cure Cancer (17:05) Reckoning with Past Failures (35:23) Misguiding Myths and Errors (59:15) AI Solutions Derive from First Principles or Data (01:31:31) Systemic Bottlenecks & Misalignments (02:08:46) Conclusion (02:14:35) The Roadmap Forward PRODUCED BY: https://aipodcast.ing SOCIAL LINKS: Website: https://podcast.futureoflife.org Twitter (FLI): https://x.com/FLI_org Twitter (Gus): https://x.com/gusdocker LinkedIn: https://www.linkedin.com/company/future-of-life-institute/ YouTube: https://www.youtube.com/channel/UC-rCCy3FQ-GItDimSR9lhzw/ Apple: https://geo.itunes.apple.com/us/podcast/id1170991978 Spotify: https://open.spotify.com/show/2Op1WO3gwVwCrYHg4eoGyP

Zak Stein is a researcher focused on child development, education, and existential risk. He joins the podcast to discuss the psychological harms of anthropomorphic AI. We examine attention and attachment hacking, AI companions for kids, loneliness, and cognitive atrophy. Our conversation also covers how we can preserve human relationships, redesign education, and build cognitive security tools that keep AI from undermining our humanity. LINKS: AI Psychological Harms Research Coalition Zak Stein official website CHAPTERS: (00:00) Episode Preview (00:56) Education to existential risk (03:03) Lessons from social media (08:41) Attachment systems and AI (18:42) AI companions and attachment (27:23) Anthropomorphism and user disempowerment (36:06) Cognitive atrophy and tools (45:54) Children, toys, and attachment (57:38) AI psychosis and selfhood (01:10:31) Cognitive security and parenting (01:26:15) Education, collapse, and speciation (01:36:40) Preserving humanity and values PRODUCED BY: https://aipodcast.ing SOCIAL LINKS: Website: https://podcast.futureoflife.org Twitter (FLI): https://x.com/FLI_org Twitter (Gus): https://x.com/gusdocker LinkedIn: https://www.linkedin.com/company/future-of-life-institute/ YouTube: https://www.youtube.com/channel/UC-rCCy3FQ-GItDimSR9lhzw/ Apple: https://geo.itunes.apple.com/us/podcast/id1170991978 Spotify: https://open.spotify.com/show/2Op1WO3gwVwCrYHg4eoGyP

Andrea Miotti is the founder and CEO of Control AI, a nonprofit. He joins the podcast to discuss efforts to prevent extreme risks from superintelligent AI. The conversation covers industry lobbying, comparisons with tobacco regulation, and why he advocates a global ban on AI systems that can outsmart and overpower humans. We also discuss informing lawmakers and the public, and concrete actions listeners can take. LINKS: Control AI Control AI global action page ControlAI's lawmaker contact tools Open roles at ControlAI ControlAI's theory of change CHAPTERS: (00:00) Episode Preview (00:52) Extinction risk and lobbying (08:59) Progress toward superintelligence (16:26) Building political awareness (24:27) Global regulation strategy (33:06) Race dynamics and public (42:36) Vision and key safeguards (51:18) Recursive self-improvement controls (58:13) Power concentration and action PRODUCED BY: https://aipodcast.ing SOCIAL LINKS: Website: https://podcast.futureoflife.org Twitter (FLI): https://x.com/FLI_org Twitter (Gus): https://x.com/gusdocker LinkedIn: https://www.linkedin.com/company/future-of-life-institute/ YouTube: https://www.youtube.com/channel/UC-rCCy3FQ-GItDimSR9lhzw/ Apple: https://geo.itunes.apple.com/us/podcast/id1170991978 Spotify: https://open.spotify.com/show/2Op1WO3gwVwCrYHg4eoGyP

Ryan Kidd is a co-executive director at MATS. This episode is a cross-post from "The Cognitive Revolution", hosted by Nathan Labenz. In this conversation, they discuss AGI timelines, model deception risks, and whether safety work can avoid boosting capabilities. Ryan outlines MATS research tracks, key researcher archetypes, hiring needs, and advice for applicants considering a career in AI safety. Learn more about Ryan's work and MATS at: https://matsprogram.org CHAPTERS: (00:00) Episode Preview (00:20) Introductions and AGI timelines (10:13) Deception, values, and control (23:20) Dual use and alignment (32:22) Frontier labs and governance (44:12) MATS tracks and mentors (58:14) Talent archetypes and demand (01:12:30) Applicant profiles and selection (01:20:04) Applications, breadth, and growth (01:29:44) Careers, resources, and ideas (01:45:49) Final thanks and wrap PRODUCED BY: https://aipodcast.ing SOCIAL LINKS: Website: https://podcast.futureoflife.org Twitter (FLI): https://x.com/FLI_org Twitter (Gus): https://x.com/gusdocker LinkedIn: https://www.linkedin.com/company/future-of-life-institute/ YouTube: https://www.youtube.com/channel/UC-rCCy3FQ-GItDimSR9lhzw/ Apple: https://geo.itunes.apple.com/us/podcast/id1170991978 Spotify: https://open.spotify.com/show/2Op1WO3gwVwCrYHg4eoGyP

Deric Cheng is Director of Research at the Windfall Trust. He joins the podcast to discuss how AI could reshape the social contract and global economy. The conversation examines labor displacement, superstar firms, and extreme wealth concentration, and asks how policy can keep workers empowered. We discuss resilient job types, new tax and welfare systems, global coordination, and a long-term vision where economic security is decoupled from work. LINKS: Deric Cheng personal websiteAGI Social Contract project siteGuiding society through the AI economic transitionCHAPTERS: (00:00) Episode Preview (01:01) Introducing Derek and AGI (04:09) Automation, power, and inequality (08:55) Inequality, unrest, and time (13:46) Bridging futurists and economists (20:35) Future of work scenarios (27:22) Jobs resisting AI automation (36:57) Luxury, land, and inequality (43:32) Designing and testing solutions (51:23) Taxation in an AI economy (59:10) Envisioning a post-AGI society PRODUCED BY: https://aipodcast.ing SOCIAL LINKS: Website: https://podcast.futureoflife.org Twitter (FLI): https://x.com/FLI_org Twitter (Gus): https://x.com/gusdocker LinkedIn: https://www.linkedin.com/company/future-of-life-institute/ YouTube: https://www.youtube.com/channel/UC-rCCy3FQ-GItDimSR9lhzw/ Apple: https://geo.itunes.apple.com/us/podcast/id1170991978 Spotify: https://open.spotify.com/show/2Op1WO3gwVwCrYHg4eoGyP

Oly Sourbut is a researcher at the Future of Life Foundation. He joins the podcast to discuss AI for human reasoning. We examine tools that use AI to strengthen human judgment, from collective fact-checking and scenario planning to standards for honest AI reasoning and better coordination. We also discuss how we can keep humans central as AI scales, and what it would take to build trustworthy, society-wide sensemaking. LINKS: FLF organization siteOly Sourbut personal siteCHAPTERS: (00:00) Episode Preview (01:03) FLF and human reasoning (08:21) Agents and epistemic virtues (22:16) Human use and atrophy (35:41) Abstraction and legible AI (47:03) Demand, trust and Wikipedia (57:21) Map of human reasoning (01:04:30) Negotiation, institutions and vision (01:15:42) How to get involved PRODUCED BY: https://aipodcast.ing SOCIAL LINKS: Website: https://podcast.futureoflife.org Twitter (FLI): https://x.com/FLI_org Twitter (Gus): https://x.com/gusdocker LinkedIn: https://www.linkedin.com/company/future-of-life-institute/ YouTube: https://www.youtube.com/channel/UC-rCCy3FQ-GItDimSR9lhzw/ Apple: https://geo.itunes.apple.com/us/podcast/id1170991978 Spotify: https://open.spotify.com/show/2Op1WO3gwVwCrYHg4eoGyP

Nora Ammann is a technical specialist at the Advanced Research and Invention Agency in the UK. She joins the podcast to discuss how to steer a slow AI takeoff toward resilient and cooperative futures. We examine risks of rogue AI and runaway competition, and how scalable oversight, formal guarantees and secure code could support AI-enabled R&D and critical infrastructure. Nora also explains AI-supported bargaining and public goods for stability. LINKS: Nora Ammann siteARIA safeguarded AI program pageAI Resilience official siteGradual Disempowerment websiteCHAPTERS: (00:00) Episode Preview (01:00) Slow takeoff expectations (08:13) Domination versus chaos (17:18) Human-AI coalitions vision (28:14) Scaling oversight and agents (38:45) Formal specs and guarantees (51:10) Resilience in AI era (01:02:21) Defense-favored cyber systems (01:10:37) AI-enabled bargaining and trade PRODUCED BY: https://aipodcast.ing SOCIAL LINKS: Website: https://podcast.futureoflife.org Twitter (FLI): https://x.com/FLI_org Twitter (Gus): https://x.com/gusdocker LinkedIn: https://www.linkedin.com/company/future-of-life-institute/ YouTube: https://www.youtube.com/channel/UC-rCCy3FQ-GItDimSR9lhzw/ Apple: https://geo.itunes.apple.com/us/podcast/id1170991978 Spotify: https://open.spotify.com/show/2Op1WO3gwVwCrYHg4eoGyP

David Duvenaud is an associate professor of computer science and statistics at the University of Toronto. He joins the podcast to discuss gradual disempowerment in a post-AGI world. We ask how humans could lose economic and political leverage without a sudden takeover, including how property rights could erode. Duvenaud describes how growth incentives shape culture, why aligning AI to humanity may become unpopular, and what better forecasting and governance might require. LINKS: David Duvenaud academic homepageGradual DisempowermentThe Post-AGI WorkshopPost-AGI Studies DiscordCHAPTERS: (00:00) Episode Preview (01:05) Introducing gradual disempowerment (06:06) Obsolete labor and UBI (14:29) Property, power, and control (23:38) Culture shifts toward AIs (34:34) States misalign without people (44:15) Competition and preservation tradeoffs (53:03) Building post-AGI studies (01:02:29) Forecasting and coordination tools (01:10:26) Human values and futures PRODUCED BY: https://aipodcast.ing SOCIAL LINKS: Website: https://podcast.futureoflife.org Twitter (FLI): https://x.com/FLI_org Twitter (Gus): https://x.com/gusdocker LinkedIn: https://www.linkedin.com/company/future-of-life-institute/ YouTube: https://www.youtube.com/channel/UC-rCCy3FQ-GItDimSR9lhzw/ Apple: https://geo.itunes.apple.com/us/podcast/id1170991978 Spotify: https://open.spotify.com/show/2Op1WO3gwVwCrYHg4eoGyP

Stephen Adler is a former safety researcher at OpenAI. He joins the podcast to discuss how to govern increasingly capable AI systems. The conversation covers competitive races between AI companies, limits of current testing and alignment, mental health harms from chatbots, economic shifts from AI labor, and what international rules and audits might be needed before training superintelligent models.  LINKS: Steven Adler's Substack: https://stevenadler.substack.comCHAPTERS:(00:00) Episode Preview(01:00) Race Dynamics And Safety(18:03) Chatbots And Mental Health(30:42) Models Outsmart Safety Tests(41:01) AI Swarms And Work(54:21) Human Bottlenecks And Oversight(01:06:23) Animals And Superintelligence(01:19:24) Safety Capabilities And Governance PRODUCED BY: https://aipodcast.ing SOCIAL LINKS: Website: https://podcast.futureoflife.org Twitter (FLI): https://x.com/FLI_org Twitter (Gus): https://x.com/gusdocker LinkedIn: https://www.linkedin.com/company/future-of-life-institute/ YouTube: https://www.youtube.com/channel/UC-rCCy3FQ-GItDimSR9lhzw/ Apple: https://geo.itunes.apple.com/us/podcast/id1170991978 Spotify: https://open.spotify.com/show/2Op1WO3gwVwCrYHg4eoGyP

Tyler Johnston is Executive Director of the Midas Project. He joins the podcast to discuss AI transparency and accountability. We explore applying animal rights watchdog tactics to AI companies, the OpenAI Files investigation, and OpenAI's subpoenas against nonprofit critics. Tyler discusses why transparency is crucial when technical safety solutions remain elusive and how public pressure can effectively challenge much larger companies. LINKS: The Midas Project WebsiteTyler Johnston's LinkedIn Profile CHAPTERS: (00:00) Episode Preview(01:06) Introducing the Midas Project(05:01) Shining a Light on AI(08:36) Industry Lockdown and Transparency(13:45) The OpenAI Files(20:55) Subpoenaed by OpenAI(29:10) Responding to the Subpoena(37:41) The Case for Transparency(44:30) Pricing Risk and Regulation(52:15) Measuring Transparency and Auditing(57:50) Hope for the Future PRODUCED BY: https://aipodcast.ing SOCIAL LINKS: Website: https://podcast.futureoflife.org Twitter (FLI): https://x.com/FLI_org Twitter (Gus): https://x.com/gusdocker LinkedIn: https://www.linkedin.com/company/future-of-life-institute/ YouTube: https://www.youtube.com/channel/UC-rCCy3FQ-GItDimSR9lhzw/ Apple: https://geo.itunes.apple.com/us/podcast/id1170991978 Spotify: https://open.spotify.com/show/2Op1WO3gwVwCrYHg4eoGyP

William MacAskill is a senior research fellow at Forethought. He joins the podcast to discuss his Better Futures essay series. We explore moral error risks, AI character design, space governance, and persistent path dependence. The conversation also covers risk-averse AI systems, moral trade between value systems, and improving model specifications for ethical reasoning. LINKS:- Better Futures Research Series: https://www.forethought.org/research/better-futures- William MacAskill Forethought Profile: https://www.forethought.org/people/william-macaskill CHAPTERS: (00:00) Episode Preview (01:03) Improving The Future's Quality (09:58) Moral Errors and AI Rights (18:24) AI's Impact on Thinking (27:17) Utopias and Population Ethics (36:41) The Danger of Moral Lock-in (44:38) Deals with Misaligned AI (57:25) AI and Moral Trade (01:08:21) Improving AI Ethical Reasoning (01:16:05) The Risk of Path Dependence (01:27:41) Avoiding Future Lock-in (01:36:22) The Urgency of Space Governance (01:46:19) A Future Research Agenda (01:57:36) Is Intelligence a Good Bet? PRODUCED BY: https://aipodcast.ing SOCIAL LINKS: Website: https://podcast.futureoflife.org Twitter (FLI): https://x.com/FLI_org Twitter (Gus): https://x.com/gusdocker LinkedIn: https://www.linkedin.com/company/future-of-life-institute/ YouTube: https://www.youtube.com/channel/UC-rCCy3FQ-GItDimSR9lhzw/ Apple: https://geo.itunes.apple.com/us/podcast/id1170991978 Spotify: https://open.spotify.com/show/2Op1WO3gwVwCrYHg4eoGyP

Karl Koch is founder of the AI Whistleblower Initiative. He joins the podcast to discuss transparency and protections for AI insiders who spot safety risks. We explore current company policies, legal gaps, how to evaluate disclosure decisions, and whistleblowing as a backstop when oversight fails. The conversation covers practical guidance for potential whistleblowers and challenges of maintaining transparency as AI development accelerates. LINKS: About the AI Whistleblower InitiativeKarl Koch PRODUCED BY: https://aipodcast.ing CHAPTERS: (00:00) Episode Preview (00:55) Starting the Whistleblower Initiative (05:43) Current State of Protections (13:04) Path to Optimal Policies (23:28) A Whistleblower's First Steps (32:29) Life After Whistleblowing (39:24) Evaluating Company Policies (48:19) Alternatives to Whistleblowing (55:24) High-Stakes Future Scenarios (01:02:27) AI and National Security SOCIAL LINKS: Website: https://podcast.futureoflife.org Twitter (FLI): https://x.com/FLI_org Twitter (Gus): https://x.com/gusdocker LinkedIn: https://www.linkedin.com/company/future-of-life-institute/ YouTube: https://www.youtube.com/channel/UC-rCCy3FQ-GItDimSR9lhzw/ Apple: https://geo.itunes.apple.com/us/podcast/id1170991978 Spotify: https://open.spotify.com/show/2Op1WO3gwVwCrYHg4eoGyP DISCLAIMERS: - AIWI does not request, encourage or counsel potential whistleblowers or listeners of this podcast to act unlawfully. - This is not legal advice and if you, the listener, find yourself needing legal counsel, please visit https://aiwi.org/contact-hub/ for detailed profiles of the world's leading whistleblower support organizations.