DeepMind Co-Founder Warns AI Safety
Google DeepMind co-founder Shane Legg warned that rapidly advancing artificial intelligence must not outpace safety controls. He launched an institute to explore artificial general intelligence deployment and broader public discussions.
Xurve View
Insights:
Xurve View

Google DeepMind co-founder Shane Legg warned during Wednesday morning's launch of a new institute that artificial intelligence capabilities must not outrun safety controls. The launch comes as increasingly capable AI agents perform autonomous software tasks, raising risks of runaway self-improvement. Legg maintains a 50% probability forecast for achieving artificial general intelligence by 2028.











