Altman: People should not cede control of the future to superintelligence

A three-sentence post

OpenAI CEO Sam Altman posted a short message on X on Nov. 4 (01:40 UTC). He wrote that people should always be in control of the future. OpenAI, he said, is against the notion that it would be acceptable for people to be relegated to doing things that “AI systems think makes sense to reward in humans.” People should not cede control of the future to superintelligence, he added, and that requirement should ultimately constrain AI progress.

By the morning of Nov. 6, the post had more than 10.3 million views and over 5,400 replies. François Chollet, the creator of Keras and a co-founder of the ARC Prize, quoted it with “Co-sign.” In his view, the only future worth building is one in which AI remains a tool in human hands, used to elevate the human condition.

Sam Altman@sama

People should always be in control of the future.

OpenAI is against the notion that it would be acceptable for people to be relegated to a role of doing things that “AI systems think makes sense to reward in humans”.

People should not cede control of the future to superintelligence, and that requirement should ultimately constrain AI progress.

Where the phrase comes from

The phrase Altman put in quotation marks appears word for word in “Machines of Loving Grace,” the essay Anthropic CEO Dario Amodei published in October 2024. In its section on work and meaning, Amodei argues that once AI becomes broadly effective and cheap enough, today’s economic setup will stop making sense. Among possible answers, he lists a large universal basic income and a capitalist economy of AI systems. In that second scenario, AI systems give out resources to humans based on “what the AI systems think makes sense to reward in humans,” a judgment “ultimately derived from human values.”

Amodei presented this as one possibility among several, not as a proposal. In the same section, he wrote that all of these solutions have “tons of possible problems” and that exploitative or dystopian directions are also possible and have to be prevented. Altman did not name Amodei or Anthropic; the link rests on the identical wording.

The ADIPEC keynote

Two days earlier, on Nov. 2, Altman made a similar point at the opening of the ADIPEC energy conference in Abu Dhabi. According to The National, he told the paper’s editor-in-chief, Mina Al-Oraibi, that there “should not be a handoff to AI” and that the technology must be “a tool built for people.” On this point, he said, there “must be no room for debate.” He added: “I can’t believe that this differentiates us from other AI companies, but it does.” In the same talk, he said AI had gone from trailing the smartest scientists at the start of the year to being “mostly smarter than them.”