Sam Altman says AI might one day solve questions even he can’t grasp

In a recent episode of comedian Theo Von's 'This Past Weekend' podcast, Sam Altman, CEO of OpenAI, shared a thought-provoking experience that showcased the remarkable capabilities of GPT-5. The discussion transitioned from light-hearted banter to a more serious reflection, particularly when Altman revealed a moment that left him both amazed and somewhat unsettled. Altman described how he posed a challenging question to GPT-5—one that he found difficult to comprehend himself. "I inputted the question into the model, and it responded flawlessly," he recounted. This revelation struck a chord with him, leading to a feeling of inadequacy in comparison to the AI's proficiency. While GPT-5 has yet to be officially launched, Altman's insights suggest it marks a significant advancement in the realm of artificial intelligence, particularly in reasoning capabilities. He emphasized the model's ability to consolidate various technological innovations developed by OpenAI into a single framework. Although no specific release date for GPT-5 was mentioned, Altman's comments have sparked speculation within the industry regarding a possible launch in the near future. His experience with the model has reignited discussions about the potential for generative AI to redefine human cognitive boundaries and knowledge expectations. This podcast episode arrives at a crucial juncture for OpenAI, as it navigates increased competition, internal changes, and ongoing debates on the path toward artificial general intelligence. For Altman, the personal implications of AI's evolution are already evident: "I felt I should have been able to answer the question, but GPT-5 resolved it in an instant."