Elon Musk said Thursday that artificial intelligence could surpass human intelligence as early as this year, arguing that progress toward artificial general intelligence is accelerating faster than humans are ready.
The comments by the CEOs of Tesla, SpaceX and xAI came during a wide-ranging conversation with BlackRock CEO Larry Fink at the World Economic Forum in Davos on Thursday.
“By the end of this year, we may have an AI that is smarter than any human being,” Musk said. “By next year at the latest,” he added, AI could be “smarter than all humans combined” by around 2030 or 2031.
Musk's comments put him among a group of tech CEOs who say AGI will arrive within years, not decades, and are increasingly concerned about workforce disruption, governance and economic concentration. He said the economic impact of AI will not be based solely on software, but will also depend on the introduction of humanoid robots capable of performing large-scale physical tasks.
“If we had ubiquitous AI and ubiquitous robots that were essentially free or close to free, the world economy would explode,” he said.
Musk also reiterated his view that humanoid robots will eventually outnumber humans.
“My prediction in a benign future scenario is that we will have so many robots and AIs created that they will meet all of our human needs,” Musk said.
Musk added that Tesla has already started using an early version of its humanoid robot, Optimus, in factories to perform simple tasks and plans for more complex tasks by the end of 2026. The company also plans to sell the humanoid robot to the public by the end of next year, once safety and reliability goals are met, Musk added.
Some researchers had previously questioned Musk's schedule, citing unresolved safety, cost and engineering challenges.
Gary Marcus, a cognitive scientist and professor emeritus of psychology and neuroscience at New York University, previously said, “Elon has a track record of making overly optimistic predictions about AI.'' decryption. “At this point, no one knows how to make humanoid robots that are safe, reliable, and generally useful at any price, so it's just fantasy to imagine that humanoid robots will sell 200 times more in the near future.”
Musk acknowledged the risks posed by the widespread use of humanoid robots, but said advances in AI and robotics are becoming more complex. The main constraint in implementation is electricity, he explained, emphasizing the need to transition to solar power.
“The Sun is by far the largest source of energy. If you look outside the Earth, the Sun rounds up to 100% of the total energy. The Sun is 99.8% of the mass of the solar system. Jupiter is about 0.1% and the rest is miscellaneous,” he said. “Even if we burned Jupiter with a thermonuclear fusion reactor, the amount of energy produced by the Sun would still reach 100%.”
He argued that the deployment of large-scale solar power will determine how quickly AI systems can scale. But Musk said advanced AI and robotics must be developed carefully.
“We need to be very careful with AI. We need to be very careful with robotics,” he said. “We don't want to feel like we're in a James Cameron movie, The Terminator.”
Despite these concerns, Musk concluded by encouraging optimism about the future of AI and humanoid robots.
“When it comes to quality of life, it's actually better to be an optimist and wrong than to be a pessimist and right,” he says.
