Possible with Reid Hoffman
Reid Hoffman
What if, in the future, everything breaks humanity's way? Possible is an award-winning, weekly podcast that sketches out the brightest version of the future—and what it will take to get there. Hosts Reid Hoffman and Aria Finger take that question to the people building the answer: founders, scientists, artists, and deep thinkers working across AI, healthcare, education, and geopolitics. The conversations run deep and specific, testing optimism against hard questions rather than trading hot takes. And between guest episodes, Reid riffs on the AI news that matters and what to make of it.
Episodes
Mentioned books
7 snips
Oct 28, 2026 • 29min
AI’s midterm coalitions, Anthropic’s safety bill, and VC conflicts
AI’s political future takes center stage, from proposals to ban superintelligence and bipartisan liability for rogue AI agents to Anthropic’s frontier safety bill. The discussion explores government’s struggle to keep pace with technical change, corporate safeguards against insider risks, and the uneasy ethics of venture capitalists backing rival startups. A sharp look at regulation, national competition, transparency, and trust in the AI race.
92 snips
Oct 21, 2026 • 27min
Karpathy’s AI gap, free Adobe clones, and AI’s effect on persistence
AI is widening the gap between everyday users and frontier researchers. A developer’s effort to recreate seven Adobe apps sparks a look at software pricing, competition, open source, and intellectual property. The conversation also explores research on AI assistance and persistence, plus how to design tools that build judgment, learning, and grit.
35 snips
Oct 14, 2026 • 28min
Why Americans are so worried about AI
Why are Americans so uneasy about AI? The conversation explores unequal prosperity, fears of a worse future, and how manufacturing and robotics could spread AI’s benefits. It also examines AI-powered hiring, the role of human judgment in team decisions, and whether advertising can coexist with trustworthy AI assistants.
19 snips
Oct 7, 2026 • 1h 24min
Building safety tools no one can take away with ROOST’s Camille François
Camille François, founding president of Roost and a Columbia professor, explores why online safety should be shared infrastructure rather than a corporate luxury. She recounts Discord’s response after Smyte vanished, the child-exploitation risks facing small apps, and Roost’s open-source tools for detection and enforcement. They also discuss AI companions, deepfake abuse, digital sovereignty, and making safety as invisible as encryption.
21 snips
Sep 30, 2026 • 15min
Who gets to pull the plug on AI?
AI’s massive infrastructure bets may be hiding serious financial exposure. Reid Hoffman and Aria Finger explore independent technical auditors, AI kill switches, California’s proposed safeguards, and the political hurdles facing federal oversight. They also debate international cyber treaties, sharper regulation, and whether “synthetic intelligence” better captures AI’s potential.
46 snips
Sep 23, 2026 • 41min
When AI's builders sound the alarm
Warnings of a 10%+ AI extinction risk spark a debate over steering versus slowing development. The conversation explores biological and cyber threats, job disruption, international coordination, recursive self-improvement, and the need for kill switches. It also tackles America’s dual mandate: leading in AI while making prosperity broadly shared through affordable energy, public benefits, and renewed trust.
26 snips
Sep 16, 2026 • 44min
Seven creators, $1,000 a week in AI tokens, one summer
Seven creators receive $1,000 a week in AI tokens and push models into unfamiliar territory. Their experiments span Dungeons & Dragons worldbuilding, a weekend-built iOS game, robotics, filmmaking, and personal agents that grow more useful over time. The conversation explores creative “spark jobs,” human taste, directing swarms of agents, and how to create genuine surprise from predictive systems.
27 snips
Sep 9, 2026 • 41min
Can we defend against AI hacks? | Dozie Anazia
Dozie Anazia, a cybersecurity analyst and computer scientist, explores AI-built software, autonomous penetration testing, and the expanding attack surface, from cloned banking pages to password-stealing hardware signals. He discusses layered defensive agents, agent-versus-agent security, why human hackers still matter, and how deepfakes threaten identity checks. Plus, a simple family safe word for thwarting voice fraud.
43 snips
Sep 2, 2026 • 59min
Training robots for a world they’ve never seen
Chelsea Finn, Stanford computer scientist and cofounder of Physical Intelligence, explores how robots can learn beyond scripted tasks. She recounts months of laundry-folding failures, explains why messy data may improve generalization, and examines robots adapting across bodies and environments. The conversation also tackles misleading demos, hardware reliability, physical safety, humanoids versus simpler machines, and robotics’ potential to reduce drudgery.
99 snips
Aug 26, 2026 • 32min
Inside an AI-powered personal dashboard | Matthew Tiemann
Matthew Tiemann, an AI strategist who builds agent-driven dashboards and automation tools, discusses agent “heartbeat” systems and a personal living-room dashboard. He talks about using Oura sleep data to time decisions. He outlines subagents, adversarial review, 12-hour human check-ins, and lessons from rapid game prototyping.
