Anthropic's Olah reveals 'mysterious' things he finds while studying AI models

Anthropic Co-founder Christopher Olah said he keeps finding things that are "mysterious, even unsettling" while studying AI models. "I lead a research team...We find structures that mirror results from human neuroscience," Olah said. "We find internal states that...mirror joy, satisfaction, fear, grief and unease. I don't know what that means, but...it warrants...discernment," he added.