OpenAI CEO Sam Altman said artificial intelligence could follow two catastrophic paths: humans could lose control of its future, or power could become highly concentrated. According to Sina Finance, he wrote on X that AI must always serve people and that alignment and safety tools must advance faster than model capabilities.
Altman also said a scenario in which one country or one lab gains too much power would be extremely dystopian. He added that avoiding those risks requires taking a narrow middle path.
Anthropic CEO Dario Amodei also warned over the weekend that AI model capability growth should be slowed to give risk controls time to catch up. Altman later said he agreed with Amodei's view.
U.S. President Donald Trump pushed back on those safety concerns on Sunday, saying the U.S. leads China in AI and that winning the AI race means winning the future.
The worries weighed on AI stocks on Monday, with Sandisk down 5% premarket and Nvidia down 3% premarket. Dan Ives, founding partner at Wedbush Securities, told Yahoo Finance that the debate could briefly pressure the market at the open, but that investors would soon focus again on the industry's planned $5 trillion in AI spending over the next five years.