MIT professor reportedly says AI proof piece seems ‘radically novel’

MIT professor @rrwilliams reportedly describes a piece he thinks appears in both OpenAI’s and Anthropic’s mathematical solutions as seemingly “radically novel,” yet understandable. He says most components have traceable origins and researchers have a “fairly good understanding” of the OpenAI proof. The unusual piece’s origin and meaning remain unclear in his account.