#243 – 'Godfather of AI' Yoshua Bengio: "I now see a path" to safe superintelligent AI

About this episode

Yoshua Bengio discusses his vision for ensuring AI is honest and incapable of deception through a new training architecture.

The co-inventor of modern AI and the most cited living scientist believes he's figured out how to ensure AI is honest, incapable of deception, and never goes rogue. Yoshua Bengio – Turing Award Winner and founder of LawZero – is disturbed by the many unintended drives and goals present in today's AIs, their willingness to lie, and ability to tell when they're being tested. AI companies are trying to stamp out these behaviours in a 'cat-and-mouse game' that Yoshua fears they're losing. --- Our new book is "a ridiculously in-depth guide to finding a fulfilling career that does good" and is out now! Order from your local bookstore, or online at https://80k.info/career-guide --- But Yoshua is optimistic: he believes the companies can win this battle decisively with a single rearrangement to how AI models are trained, and has been developing mathematical proofs to back up the claim. The core idea is that instead of training AI to predict what a human would say, or to produce responses we'd rate highly, we should train it to model what's actually true. Yoshua argues this new architecture, which he calls 'Scientist AI,' is a small enough change that we could keep almost all the…

People in this episode

Guest: Yoshua Bengio

Topics covered

• AI safety
• superintelligent AI
• training models
• honesty in AI
• deception in AI

Keywords

• AI
• superintelligence
• Scientist AI
• deception
• training methods

Mentioned in this episode

Organizations: LawZero

Products: Claude, ChatGPT

More episodes of 80,000 Hours Podcast

• How AI could create the world’s biggest problems (article by Zershaaneh Qureshi) · June 11, 2026 · 1h 30m
• What it's really like to run AGI safety at Google DeepMind (and where I disagree with 'doomers') | Rohin Shah · June 2, 2026 · 2h 48m
• What makes for a dream job? | Benjamin Todd · May 28, 2026 · 28 min
• We’re updating our career advice for the strangest time in history | Benjamin Todd, author of 80,000 Hours · May 26, 2026 · 1h 7m
• Can AIs already start 'rogue deployments' inside AI companies? (Landmark new METR report) · May 20, 2026 · 20 min
• '95% of AI Pilots Fail': The hidden agenda behind the viral stat that misled millions · April 28, 2026 · 10 min