Sam Altman: GPT-5 Is Impressive, But Not Artificial General Intelligence

Sam Altman, CEO of OpenAI, has stated that the recently launched GPT-5, while a major leap forward in artificial intelligence, does not meet the criteria for artificial general intelligence, or AGI. During a press call ahead of the release, Altman acknowledged the model’s impressive capabilities but emphasized that it still lacks key traits he associates with true AGI. One of the most significant gaps, according to Altman, is the model’s inability to learn continuously in real time as it interacts with new information. “This is not a model that continuously learns as it's deployed from the new things it finds, which is something that to me feels like AGI,” he said. He stressed that while GPT-5 demonstrates broad intelligence and substantial improvements over previous versions, it remains fundamentally limited by its static nature—trained on fixed datasets and unable to adapt autonomously. Altman described GPT-5 as “generally intelligent,” a significant step toward AGI, but not yet at the threshold where AI can reason, adapt, and learn like a human across diverse and novel situations. He noted that even five years ago, the capabilities of today’s GPT-5 would have seemed like a major milestone on the path to AGI. The concept of AGI—machines with human-level cognitive abilities across a wide range of domains—remains theoretical and highly debated. Some experts, including Meta’s chief AI scientist Yann LeCun, believe we are still decades away from achieving it. Despite this, Altman remains committed to OpenAI’s long-term mission: developing AGI that benefits all of humanity. In a January blog post, he reiterated that the company was founded nearly nine years ago with the belief that AGI is achievable and could be the most transformative technology in human history. Beyond AGI, Altman is now focused on the next frontier: superintelligence. He envisions systems that can reason far beyond human capacity, accelerating scientific discovery, innovation, and global prosperity. “Superintelligent tools could massively accelerate scientific discovery and innovation well beyond what we are capable of doing on our own,” he wrote. For now, GPT-5 stands as a powerful milestone in OpenAI’s journey—but not the final destination.