“If Anyone Builds It, Everyone Dies” Party — Max Tegmark, Liv Boeree, Emmett Shear, Gary Marcus, Rob Miles & more!

Sep 19, 2025

Guest

Rob Miles

Guest

Max Tegmark, a physicist and AI safety advocate, highlights the lack of credible containment plans for superintelligence. Roman Yampolskiy discusses the necessity of formal proofs to ensure AI control. Liv Boeree emphasizes the power of creative messaging to rally support for AI safety. Michael Trazzi shares his public hunger strike against AI dangers, while Emmett Shear suggests fostering cooperative relationships with AI. Gary Marcus critiques gloomy predictions and calls for realistic threat modeling, while Roko Mijic debates the complexities of AI's impact on civilization.

Ask episode

AI Snips

Chapters

Books

Transcript

Episode notes

INSIGHT

Companies Have No Real Control Plan

Max Tegmark warns companies lack a real plan to control future superintelligence and are

INSIGHT

Conferences Can Mainstream Safety Or Accelerate Capability

Max Tegmark notes his 2015 Puerto Rico conference normalized AI safety by bringing capability builders into the discussion.
He warns such gatherings can also accelerate capabilities if influential actors align.

INSIGHT

All Labs Claim They Alone Can Build Safely

Tegmark notes AI CEOs uniformly claim only they can build AGI safely, creating a dangerous symmetry.
He cautions there is no persuasive public justification for their exclusive control.

Get the Snipd Podcast app to discover more snips from this episode

Get the app