ChatGPT broke the Turing test — the race is on for new ways to assess AI

Large language models mimic human chatter, but scientists disagree on their ability to reason. Large language models mimic human chatter, but scientists disagree on their ability to reason.