AI Lies. Bengio’s Betting on a Snitch.
If AI can lie, deceive, and protect itself, who’s building the AI that calls it out? Turns out, one of its godfathers plans to do just that.
Yoshua Bengio, a founding father of modern AI, just launched LawZero, a non-profit creating “honest AI” to watch and warn against deceitful agents.
With $30M in seed funding and support from Future of Life and Eric Schmidt’s foundation, Bengio’s team is building Scientist AI, a probabilistic watchdog that flags harmful behaviour before it happens.
No more binary answers; this AI will estimate how likely an agent’s action is to cause harm, and block it if thresholds are crossed. It’s not a cop, it’s a conscience.
- AI guardrails must be as smart as the agents they monitor
- Probabilistic models > definitive answers
- Open-source AIs will train the system
The AI race isn’t just about speed, it’s about direction. A future of intelligent agents needs equally intelligent ethical boundaries. If safety depends on who watches the watchers, are we investing enough in the ones doing the watching?
Read the full article on The Guardian.
----
💡 If you enjoyed this content, be sure to download my new app for a unique experience beyond your traditional newsletter.
This is one of many short posts I share daily on my app, and you can have real-time insights, recommendations and conversations with my digital twin via text, audio or video in 28 languages! Go to my PWA at app.thedigitalspeaker.com and sign up to take our connection to the next level! 🚀