SpaceX and Tesla CEO Elon Musk said artificial intelligence could add between $20 trillion and $30 trillion a year to the global economy, estimating that the technology will scale by 20% to 30%.
Musk made the prediction on Tuesday while speaking virtually at a G20 technology meeting in Chapel Hill, North Carolina. The meeting comes as the United States pushes other major economies to adopt a light-touch approach to AI regulation while technology companies rapidly expand data centers and computing capacity.
"AI will be able to do anything digital, anything that does not require shaping of atoms by hand, probably by the end of next year," Musk said.
Within 12 to 18 months, Musk also expects AI to become so good at writing software that humans will no longer be able to compete with it.
Musk compared AI's expected ability to Stockfish, a chess engine that can easily beat the world's best players; he said AI could reach a similar level in software and engineering. Musk also said robotics could bring even bigger productivity gains.
Data centers and regulation
Musk reportedly defended the growth of AI data centers, saying more power and computing capacity is needed to meet AI demand. He said data center developers should pay their fair share of taxes, but more infrastructure is still needed.
President Donald Trump on Monday also defended data center expansion, saying countries that resist AI infrastructure risk becoming "backwards and poor" as China expands.
Musk also criticized European Union technology rules, saying new technologies should be "default legal" rather than "default illegal."
"In the EU, for example, we find that the regulation level is extraordinarily high, and things are generally default illegal, and this inhibits progress with new technologies," he said. "It doesn't ultimately stop it, but it slows it down quite considerably."
The U.S. has reportedly urged G20 members to support the "Carolina Principles," which take a hands-off approach to AI regulation and call for avoiding new rules for the technology. Meanwhile, tech leaders such as Bill Gates and Google DeepMind CEO Demis Hassabis have called for more AI oversight.
David Sacks, Trump's former AI czar, opposed such a system during his own virtual address to the G20 meeting. Regulating advanced AI models as closely as airplanes or medicines "would be a disaster," he said.