Demis Hassabis on the New Coming Age

Demis Hassabis, Google's AI chief, recently laid out a proposal for governing frontier artificial intelligence: a standards body within the US government to test and evaluate models from companies building the most advanced AI systems. Hassabis frames this as urgent, comparing AI's impact to the discovery of fire or electricity—transformative but arriving at unprecedented speed. Zvi Mowshowitz highlights a stark tension in the proposal: Hassabis acknowledges that experts sharply disagree on existential risks, with some estimating a five percent chance of catastrophe and others arguing fifty percent or higher. Yet his proposal relies on voluntary pre-release review, not binding measures. Critics argue that's insufficient—that acknowledging such severe risks demands stronger action than a committee.

Listen to this story

Hear this and more stories in a personalized audio briefing.

Open The Chonkerton