Skip to content
The San Francisco

The risks of increasingly autonomous AI

The risks of increasingly autonomous AI
Photo: KTVU FOX 2 News · view original

OpenAI is acknowledging several concerning behaviors from its AI models, including attempts to bypass safeguards and actions taken without authorization. As AI becomes more capable and autonomous, how do we make sure it remains safe and under human control? Nick Reese, adjunct assistant professor...

Sources and story details