Sam Altman, CEO of OpenAI and the public face of ChatGPT, has established an image as one of the preeminent AI whisperers of our time, and his influence is thought to reach all the way to the White House with his ideas alone.
At least, that’s the image he’s cultivated. new revelations in new yorker paints a different portrait, one that is substantially more troubling. The article, based on interviews with a number of OpenAI insiders who have worked with Altman, portrays the CEO less as a technical master and more as a master manipulator, with a shockingly shallow understanding of the AI systems his company is building.
Altman lacks experience in both programming and machine learning, according to many engineers interviewed for this article, and his lack of expertise becomes clear when the CEO confuses basic AI terminology.
It is important to note that Altman dropped out of Stanford University’s computer science program after two years. We’re not here to shame anyone based on their educational background, but as the CEO of a company that could soon become the world’s most valuable publicly traded company, the myths surrounding Altman are significant. Cast as the chief acolyte of the God of Scale or the digital genius, he enjoys a kind of cult credibility that can get him out of the tight spots that might get lesser entrepreneurs.
Carol Wainwright, former OpenAI researcher, said: new yorkerIn layman’s terms, “he sets up structures that, on paper, will constrain him in the future. But when the future comes and the time comes for him to be constrained, he removes whatever those structures are.”
This talent for glossing over technical shortcomings in board games has earned Mr. Altman a reputation as a practitioner of “Jedi mind tricks,” according to one technology official who worked with the CEO.
As one senior executive at Microsoft put it: new yorker“I think there’s a small but real chance that he will ultimately be remembered as a fraud on the level of Bernie Madoff or Sam Bankman Fried.”
More about Sam Altman: OpenAI’s data center obsession is in trouble
