Computerphile

AI Sandbagging (2025x20)

:

Following the theme of AI research and safety, Aric Floyd talks about how some Large Language Models might follow the all too human trait of sandbagging - "lying" about their true capabilities.