Future of Life Institute Podcast

Suzy Shepherd joins the podcast to discuss her new short film "Writing Doom", which deals with AI risk. We discuss how to use humor in film, how to write concisely, how filmmaking is evolving, in what ways AI is useful for filmmakers, and how we will find meaning in an increasingly automated world.   Here's Writing Doom:   https://www.youtube.com/watch?v=xfMQ7hzyFW4   Timestamps: 00:00 Writing Doom  08:23 Humor in Writing Doom 13:31 Concise writing  18:37 Getting feedback 27:02 Alternative characters 36:31 Popular video formats 46:53 AI in filmmaking49:52 Meaning in the future

Andrea Miotti joins the podcast to discuss "A Narrow Path" — a roadmap to safe, transformative AI. We talk about our current inability to precisely predict future AI capabilities, the dangers of self-improving and unbounded AI systems, how humanity might coordinate globally to ensure safe AI development, and what a mature science of intelligence would look like.   Here's the document we discuss in the episode:   https://www.narrowpath.co  Timestamps: 00:00 A Narrow Path 06:10 Can we predict future AI capabilities? 11:10 Risks from current AI development 17:56 The benefits of narrow AI  22:30 Against self-improving AI  28:00 Cybersecurity at AI companies  33:55 Unbounded AI  39:31 Global coordination on AI safety 49:43 Monitoring training runs  01:00:20 Benefits of cooperation  01:04:58 A science of intelligence  01:25:36 How you can help

Tamay Besiroglu joins the podcast to discuss scaling, AI capabilities in 2030, breakthroughs in AI agents and planning, automating work, the uncertainties of investing in AI, and scaling laws for inference-time compute. Here's the report we discuss in the episode:  https://epochai.org/blog/can-ai-scaling-continue-through-2030  Timestamps: 00:00 How important is scaling?  08:03 How capable will AIs be in 2030?  18:33 AI agents, reasoning, and planning 23:39 Automating coding and mathematics  31:26 Uncertainty about investing in AI 40:34 Gap between investment and returns  45:30 Compute, software and data 51:54 Inference-time compute 01:08:49 Returns to software R&D  01:19:22 Limits to expanding compute

Ryan Greenblatt joins the podcast to discuss AI control, timelines, takeoff speeds, misalignment, and slowing down around human-level AI. You can learn more about Ryan's work here: https://www.redwoodresearch.org/team/ryan-greenblatt  Timestamps: 00:00 AI control  09:35 Challenges to AI control  23:48 AI control as a bridge to alignment 26:54 Policy and coordination for AI safety 29:25 Slowing down around human-level AI 49:14 Scheming and misalignment 01:27:27 AI timelines and takeoff speeds 01:58:15 Human cognition versus AI cognition

Tom Barnes joins the podcast to discuss how much the world spends on AI capabilities versus AI safety, how governments can prepare for advanced AI, and how to build a more resilient world.   Tom's report on advanced AI: https://www.founderspledge.com/research/research-and-recommendations-advanced-artificial-intelligence   Timestamps: 00:00 Spending on safety vs capabilities 09:06 Racing dynamics - is the classic story true?  28:15 How are governments preparing for advanced AI?  49:06 US-China dialogues on AI 57:44 Coordination failures  1:04:26 Global resilience  1:13:09 Patient philanthropy  The John von Neumann biography we reference: https://www.penguinrandomhouse.com/books/706577/the-man-from-the-future-by-ananyo-bhattacharya/

Samuel Hammond joins the podcast to discuss whether AI progress is slowing down or speeding up, AI agents and reasoning, why superintelligence is an ideological goal, open source AI, how technical change leads to regime change, the economics of advanced AI, and much more.   Our conversation often references this essay by Samuel: https://www.secondbest.ca/p/ninety-five-theses-on-ai   Timestamps: 00:00 Is AI plateauing or accelerating?  06:55 How do we get AI agents?  16:12 Do agency and reasoning emerge?  23:57 Compute thresholds in regulation28:59 Superintelligence as an ideological goal 37:09 General progress vs superintelligence 44:22 Meta and open source AI  49:09 Technological change and regime change 01:03:06 How will governments react to AI?  01:07:50 Will the US nationalize AGI corporations?  01:17:05 Economics of an intelligence explosion  01:31:38 AI cognition vs human cognition  01:48:03 AI and future religions 01:56:40 Is consciousness functional?  02:05:30 AI and children

Anousheh Ansari joins the podcast to discuss how innovation prizes can incentivize technical innovation in space, AI, quantum computing, and carbon removal. We discuss the pros and cons of such prizes, where they work best, and how far they can scale. Learn more about Anousheh's work here: https://www.xprize.org/home  Timestamps: 00:00 Innovation prizes at XPRIZE 08:25 Deciding which prizes to create 19:00 Creating new markets 29:51 How far can prizes scale?  35:25 When are prizes successful?  46:06 100M dollar carbon removal prize 54:40 Upcoming prizes 59:52 Anousheh's time in space

Mary Robinson joins the podcast to discuss long-view leadership, risks from AI and nuclear weapons, prioritizing global problems, how to overcome barriers to international cooperation, and advice to future leaders. Learn more about Robinson's work as Chair of The Elders at https://theelders.org  Timestamps: 00:00 Mary's journey to presidency  05:11 Long-view leadership 06:55 Prioritizing global problems 08:38 Risks from artificial intelligence 11:55 Climate change 15:18 Barriers to global gender equality  16:28 Risk of nuclear war  20:51 Advice to future leaders  22:53 Humor in politics 24:21 Barriers to international cooperation  27:10 Institutions and technological change

Emilia Javorsky joins the podcast to discuss AI-driven power concentration and how we might mitigate it. We also discuss optimism, utopia, and cultural experimentation. Apply for our RFP here:   https://futureoflife.org/grant-program/mitigate-ai-driven-power-concentration/Timestamps: 00:00 Power concentration  07:43 RFP: Mitigating AI-driven power concentration 14:15 Open source AI  26:50 Institutions and incentives 35:20 Techno-optimism  43:44 Global monoculture  53:55 Imagining utopia

Anton Korinek joins the podcast to discuss the effects of automation on wages and labor, how we measure the complexity of tasks, the economics of an intelligence explosion, and the market structure of the AI industry. Learn more about Anton's work at https://www.korinek.com  Timestamps: 00:00 Automation and wages 14:32 Complexity for people and machines 20:31 Moravec's paradox 26:15 Can people switch careers?  30:57 Intelligence explosion economics 44:08 The lump of labor fallacy  51:40 An industry for nostalgia?  57:16 Universal basic income  01:09:28 Market structure in AI