OpenAI CEO Sam Altman has described the company's GPT-5.4 model as his favorite to interact with. However, he acknowledged that OpenAI still needs to address three key weaknesses in the technology. The comments highlight ongoing improvements in AI conversational abilities.
Sam Altman, CEO of OpenAI, recently shared his positive impressions of the GPT-5.4 model, stating it is his 'favorite model to talk to.' This remark underscores the advancements in making AI interactions feel more natural and engaging.
Despite this praise, Altman admitted that the model is not without flaws. He pointed out that OpenAI must still fix three specific weaknesses, though details on these issues were not elaborated in the available information. Such acknowledgments reflect the iterative nature of AI development at OpenAI, where enhancements in user experience coexist with areas requiring further refinement.
The statement comes amid OpenAI's continued evolution of its GPT series, building on previous models to improve responsiveness and utility. Altman's balanced view suggests confidence in the direction of GPT-5.4 while emphasizing the need for ongoing work to resolve persistent challenges.
