TechFlame reports that although ChatGPT has introduced a new GPT-5 model, there is still a possibility that it will go wrong. A senior OpenAI executive once again emphasized this this week. ChatGPT director Nick Terley stated in an interview: “In terms of reliability, there is a huge discontinuity between reliability and complete reliability.” He further explained, “Unless we can prove that ChatGPT is more reliable than human experts in all fields, not just some, we will continue to recommend that you check the answers carefully.” (gold 10)