Here's why Sam Altman says OpenAI's GPT-5 falls short of AGI

OpenAI's CEO, Sam Altman, has made it clear that while the newly launched GPT-5 represents a significant milestone in artificial intelligence, it does not meet the criteria for artificial general intelligence (AGI). During a press call prior to the release, he emphasized that AGI remains a theoretical concept where AI can reason and understand like humans. Altman explained that the core mission of OpenAI is to develop AGI that serves all of humanity. He acknowledged that while GPT-5 exhibits impressive capabilities, it still lacks several critical aspects that define true AGI. "This model shows a level of general intelligence, yet we still seem to be missing a few essential components that are crucial to AGI," he stated. One key limitation highlighted by Altman is GPT-5's inability to learn independently from new experiences once it is deployed. He noted, "A significant factor is that this model does not continuously learn from the new data it encounters, which I believe is a hallmark of AGI. Nevertheless, it represents a remarkable improvement in intelligence and capability." The question of how to define AGI and the timeline for its development is a hot topic among AI experts. Some, including Yann LeCun, Meta's chief AI scientist, suggest that true AGI may still be decades away. However, Altman believes that GPT-5 is a step toward that goal. Reflecting on OpenAI's journey, he remarked, "If I could go back five years prior to the launch of GPT-3 and see where we are now, I would recognize that we are considerably closer to something resembling AGI." In previous statements, Altman has expressed that the founders of OpenAI embarked on this journey nearly nine years ago with the conviction that AGI could be achieved and could have profound implications for humanity. While the pursuit of AGI remains a priority, Altman also mentioned that OpenAI is already contemplating the concept of superintelligence—an advancement where AI capabilities far exceed human intelligence. He believes that such tools could revolutionize scientific discovery and innovation, leading to unprecedented levels of abundance and prosperity.