In Greek mythology, Prometheus is a Titan who is sentenced to eternal torture by Zeus after stealing fire from Olympus and gifting it to humanity. He is a democratizing figure who breaks out a source of great power from its confines among the gods to give it to the everyday folk.
In Silicon Valley, OpenAI CEO Sam Altman views himself in a similar light.
In a new interview with Vanity Fair, Altman was asked to describe his personal mission and how he would like to be remembered in the history of AI.
“And what I would really like, if the history books ever write anything about me at all, it’d be great if they don’t, is just the work that OpenAI did really made sure that AI was a fundamental thing that every person got hugely empowered with,” Altman said, comparing the technology to fire or electricity. “It was not something that happened to them. It was something that people were given.”
When the interviewer concludes that the figure he is describing is essentially Prometheus, Altman admits, “closer to that than other things you could come up with for sure.”
The OpenAI chief says that many people in Silicon Valley “really believed only a small priesthood should get access to AI,” ultimately casting a group of Silicon Valley insiders as “benevolent dictators” who determine how the general public benefits from AI development. But Altman and OpenAI allegedly believed that giving the “superpower” of AI to people would instead achieve “tremendously more good than bad.”
Altman’s answer taps into longstanding fears of the formation of an AI-powered superclass versus a group of tech-illiterate working-class folk whose jobs get automated by a super-advanced stage of artificial intelligence: a so-called “permanent underclass.” Some of the people who subscribe to this prophecy claim that if these newer, higher-tech artificial intelligence models are kept under wraps only to be used by a select group of people or are made available at prices too high for the average segment of society to afford, then it will result in a group of tech elites gaining immense power over others. OpenAI’s rival Anthropic reignited this paranoia earlier this year with the release of Fable, the watered-down public-use version of its allegedly scarily advanced model Mythos, which was only made available to select government entities and corporations.
The permanent underclass narrative ties into the discussion over AI regulation as well, where some AI industry leaders argue that so-called overregulation will block people out and lead to a concentration of power in a handful of top AI companies. In the interview, Altman said that OpenAI is in a gray area when it comes to the question of regulation.
“We are willing to accept some things going wrong to make sure that this is a widely held and used and democratic technology,” Altman said. For example, open-source models will cause a “tidal wave of cybersecurity problems,” Altman argues, but still should not be banned.
“We’re going to, as a society, I think, have to just accept some fairly severe cyber incidents from open models in exchange for the liberty that comes with that,” Altman said.
