Sam Altman, CEO of OpenAI, has called for a slowdown in the rapid advancement of artificial intelligence, citing two primary ways the technology could develop in detrimental directions. His remarks come in the wake of a stark warning from an AI whistleblower.
In an X post on Sunday night, Altman stated, "There are two ways AI could go very badly and that we must avoid." The first concern is the potential loss of control over AI's future, which he deemed "unacceptable." Altman emphasized that AI must always serve humanity and that alignment and safety measures must keep pace with the development of AI capabilities.
The second major concern raised by Altman is the risk of excessive power concentration. He warned that a highly powerful AI could be wielded by a single entity to impose its worldview on others, leading to a "dystopian" future.
Altman clarified that "pacing" does not mean halting progress but rather slowing it down. He argued that the competitive pressure within the United States should not lead to recklessness or allow capabilities to outstrip safety monitoring. "No amount of American competitive pressure should justify recklessness, or let capabilities get ahead of alignment and monitoring," he posted. He added that "progress has been rapid and will continue to be. But it should be slower than it otherwise could be," stating that safety issues could come at a significant cost.
Altman's plea follows the resignation of Jacob Coxon, a former researcher at OpenAI and Anthropic, who expressed concerns that companies are not acting responsibly. Coxon warned that these companies are "racing straight to self-improving superintelligence and gambling with our lives."
Coxon urged the public not to underestimate AI's power, suggesting that these systems could soon become superhuman, capable of hacking anything and revolutionizing fields overnight. He claimed that while executives publicly present a more measured view, private discussions reveal similar anxieties.
Earlier warnings about the potential existential threat of AI have also been voiced by Geoffrey Hinton, often referred to as the "Godfather of AI." Hinton has cautioned that developing superintelligence without a scientific consensus on its safe and controllable development could be catastrophic, potentially leading to human extinction.