Why GPT-5 is making OpenAI CEO Sam Altman feel ‘useless’

OpenAI CEO Sam Altman previously said his company needs to “figure out a different open source strategy.”

OpenAI CEO Sam Altman has voiced deep unease over the capabilities of GPT-5, the company’s upcoming AI model set to launch in August 2025. Speaking on a podcast, Altman admitted that testing the system left him feeling “useless” and compared its creation to the Manhattan Project — the World War II program that developed the first atomic bomb.

“There are moments in science when people look at what they’ve created and ask, ‘What have we done?’” Altman said. He recalled watching GPT-5 solve a problem that he could not, triggering what he called a “personal crisis of relevance.”

Altman’s historical reference highlights his belief that GPT-5, while not a weapon, could have irreversible consequences. He said his reaction was not due to fear of losing control but from realizing that AI may now surpass human cognitive capabilities in ways society may not be ready to manage.

He added that moments like this force a broader reckoning: “It’s not just about what AI can do; it’s about whether we’re ready for the answers.”

Technical specifications for GPT-5 remain undisclosed, but internal reports suggest major upgrades, including longer memory, improved multi-step reasoning, and advanced multimodal processing. Altman described GPT-4 in retrospect as “the dumbest model any of you will ever have to use again, by a lot.”

OpenAI has long aimed to build Artificial General Intelligence (AGI) — AI capable of performing any intellectual task a human can. Altman’s shift in tone signals that GPT-5 may represent a significant step toward that goal.

The launch comes amid pressure from Microsoft, which has invested $13.5 billion in OpenAI, to accelerate enterprise rollouts as competition from Google, Anthropic, and Meta intensifies. There is also investor interest in moving to a for-profit structure by the end of 2025.