3 ms·Large language models pass three-party Turing test, outscoring actual humans3 points by maxloh 22d ago