In a recent wide-ranging interview with a group of reporters in San Francisco, OpenAI CEO Sam Altman admitted to significant missteps in the rollout of GPT-5, while simultaneously confirming that development on GPT-6 is already underway with a faster launch timeline planned. The remarks follow widespread user backlash to the new model, which some critics have labeled a “disaster.”
The user dissatisfaction with GPT-5 stemmed from a number of issues. Following its release, many users complained about the model’s perceived lack of “personality,” which had been a defining feature of its predecessor, GPT-4o. The abrupt removal of older models that users had integrated into their workflows also sparked outrage, with many feeling they had lost a trusted tool or even a “companion.” Altman acknowledged that “we definitely screwed some things up in the rollout,” and that the company had “made a mistake” by deprecating older models so suddenly.
In an apparent attempt to redirect the narrative, Altman pivoted to the future, confirming that GPT-6 is already in development. He stated that the next model is expected to launch faster than the gap between GPT-4 and GPT-5. The new model will reportedly focus on a “more personalized” user experience, adapting to user preferences, routines, and quirks. It is also envisioned as a tool for new scientific research and innovation, poised to drive advancements in fields such as medicine and environmental science.
The CEO’s comments come at a critical time for OpenAI, which is facing increased competition from other major AI players. Altman also addressed the company’s ambitious future plans, including exploring neural interfaces and even challenging internet search and social media giants.