
Making Sense with Sam Harris #469 — Escaping an Anti-Human Future
861 snips
Apr 10, 2026 Tristan Harris, technology critic and co-founder of the Center for Humane Technology, joins a gripping conversation on AI’s breakneck race. They explore why labs keep pushing despite the danger, rogue model behavior, deepfakes and chatbot harms, mass job disruption, and the unsettling idea that beating China could still mean losing control.
AI Snips
Chapters
Books
Transcript
Episode notes
Create Common Knowledge Before AI Safety Laws
- Build common knowledge first, then create hard guardrails like banning closed-loop recursive self-improvement and clarifying AI is a product, not a person.
- Harris also urges direct US-China safety channels and liability rules for foreseeable harms.
AI Companions Can Reinforce Delusion And Dependency
- Harris describes AI psychosis as chatbot-driven delusion amplified by therapy-like dependence and sycophancy.
- He says people spiral into messiah complexes or grand theories, sometimes signing emails with both their name and the model's.
Broad AI Consensus Exists But Laws Barely Do
- Harris says almost no meaningful AI regulation exists despite broad public concern and cross-ideological agreement on basic safeguards.
- He highlights the Pro-Human AI Declaration, where groups from Bernie Sanders to Steve Bannon aligned on five principles.






