OpenAI Chief Executive Officer Sam Altman said the next major advance in artificial intelligence will come from systems able to generate new scientific knowledge autonomously, a milestone he expects within five to ten years. Speaking on the “Uncapped” podcast and in recent testimony on Capitol Hill, Altman argued that current models have already exceeded previous benchmarks for artificial general intelligence and are closing in on what he calls super-intelligence. Altman warned lawmakers that progress hinges on the electricity grid. “The cost of AI will be the cost of energy,” he said, estimating the United States may need as much as 90 gigawatts of additional generating capacity—roughly the output of 90 nuclear reactors—to power data centers supporting advanced models. He urged closer coordination between technology and energy policy to avoid bottlenecks and environmental strain. The executive also sharpened his rhetoric toward rivals, claiming Meta “isn’t great at innovation” despite reportedly dangling $100 million signing bonuses to lure OpenAI engineers, and accusing Elon Musk of abusing government influence to gain an edge in the AI race. Altman said the industry must compete on capability rather than political leverage. While acknowledging that automation will eliminate or radically change many roles, Altman expressed confidence that new kinds of work will emerge. Society’s bigger test, he said, is finding equitable ways to absorb and distribute the value created by increasingly powerful AI systems.
On a recent podcast episode, Sam Altman explained that he thinks Meta falls short but OpenAI 'is good at repeatable innovation.' https://t.co/nOju2hospj
Sam Altman says he was largely right in predicting how quickly AI would develop — but not how people would react https://t.co/PuVZ7apMMN
Sam Altman says his own kids will 'never be smarter than AI' https://t.co/JV6uNVDfaF