OpenAI CEO Sam Altman has expressed a rare vulnerability while discussing the power of GPT-5, the company’s upcoming large language model. During a conversation on comedian Theo Von’s This Past Weekend podcast, Altman recounted a moment where the AI’s capabilities left him feeling overwhelmed and, in his own words, “useless.”
Altman revealed that he had recently used GPT-5 to handle a question he had received via email, one he admittedly did not fully understand. The AI provided a response that was so accurate and nuanced that it made him feel outmatched by the very technology his organisation created.
“I put it in the model… and it answered perfectly,” said Altman. “It was really hard for me, but the AI just did it like that. It was a weird feeling.”
GPT-5 Nears Release, Altman Confirms Testing Underway
Although OpenAI has yet to announce a release date for GPT-5 formally, Altman has confirmed on X (formerly Twitter) that the model is already operational and will launch “soon.” The advanced AI model is rumoured to be released as early as August, potentially with scaled-down “mini” and “nano” versions made available through the company’s API.
Limited testing of GPT-5 has already begun, and internal sources describe it as OpenAI’s most advanced system yet. One of its standout features is integrating the “o3 reasoning” engine, which was originally conceived as a separate architecture. This inclusion is expected to enhance the model’s analytical and problem-solving capabilities significantly.
Algoritha: The Most Trusted Name in BFSI Investigations and DFIR Services
Rising Ethical and Mental Health Questions
During the same podcast appearance, Altman also addressed growing concerns surrounding the impact of advanced AI systems on mental health. With more users turning to AI companions for emotional support and therapy, Altman acknowledged the unknown risks:
“I’m scared of that… I don’t think we know quite how it’s going to have those negative impacts, but I feel for sure it’s going to have some.”
His remarks have added weight to ongoing public discourse about the ethical dimensions of increasingly human-like AI systems.