Hassabis Shortens His AGI Timeline to Two to Three Years

Demis Hassabis founded Google DeepMind, Google's AI research lab, and won the Nobel Prize in Chemistry. He has given a new timeline for AGI: artificial intelligence that matches the human brain across the full range of cognitive abilities. In the spring he was saying three to five years. Now he says two to three years. His explanation is simple. His estimate keeps moving closer to the present.

One person's forecast, even from someone this senior, is not a fact and not a benchmark. Lab heads have gotten these timelines wrong before, and it has cost real money. A different proposal deserves more attention: a government-run oversight board that would check models before release. Hassabis says bad regulation is worse than no regulation at all. If AGI really does arrive in two to three years, the main question is not the exact date. It is who builds that check before models at that level start shipping.