A UC San Diego study found GPT-4.5 was judged human more often than real people in live chats, raising sharper questions about AI disclosure, trust, and online identity.
AI can finally pass the Turing Test better than a human, study warns - Experiment should lead to a re-evaluation of how we ...