A surprising development in the artificial intelligence world has occurred that has even impressed Elon Musk, CEO of Tesla, xAI, Neuralink, and SpaceX.
The new version of ChatGPT based on OpenAI’s GPT-5 model has taken a different approach compared to previous models – instead of providing incorrect or hallucinated answers, it has now learned to say “I don’t know”
In this particular case, ChatGPT 5 admitted after 34 seconds of thinking that it had no information about the subject being queried.
This remarkable moment where the GPT-5 model declared “I don’t know – and I can’t reliably find out” has been praised as a significant leap forward for AI accountability.
In an era where humans are fighting against AI-driven misinformation, this development offers hope in the battle against false information.
It all began when a ChatGPT user named Kol Tregaskes shared a screenshot showing the AI’s surprising response to a query Elon Musk responded to the original post saying “impressive,” thus praising the AI chatbot for its ability to admit mistakes instead of directing users to incorrect responses.
OpenAI has reportedly reduced the chances of incorrect responses to just 10 percent in this latest GPT-5 version.
Despite these advancements, OpenAI’s head of ChatGPT, Nick Turley, has cautioned users against relying on the AI model as their primary information source.
