New benchmark tracks AI-human control failures
A new benchmark tests how AI and human operators fail together in shared control systems, pinpointing where and why things go wrong.
A new benchmark tests how AI and human operators fail together in shared control systems, pinpointing where and why things go wrong.