Real Despots Hijack Artificial Intelligence

A.I. is a teenager now, roaring into the world, testing limits, rebelling against authority, itching to usurp the old guard and remake the planet in its image.

Unfortunately, Pete Hegseth is also a teenager. His hormones are raging; his judgment is shaky. Like a repentant frat boy, he had to promise the adults in the Senate that he wouldn’t drink while he is in charge of the military and its 12-figure budget. He certainly lacks the maturity to guide, discipline or even understand the earth-shattering power of an adolescent A.I.

Hegseth should be focused on our nerve-racking duel with Iran. Instead, he spent the week at war with Dario Amodei, the thoughtful chief executive of Anthropic and one of the few in Silicon Valley advocating for humanity. Anthropic is the only A.I. company operating on classified military systems; its clever chatbot, Claude, was deployed by the military to help catch Venezuela’s Nicolás Maduro.

More than most of his peers, Amodei has been blunt about “civilizational concerns” — the risks of A.I. wiping us out. He even hired an Oxford-educated philosopher, a young Scottish woman, to teach Claude right from wrong. She’s feeding his “soul,” she said. Claude even has his own Constitution, rules for the bot’s values and behavior. (Good luck!)

A fully powerful A.I. may be only one to two years away, Amodei wrote in a January essay, “The Adolescence of Technology,” adding that it will be “smarter than a Nobel Prize winner across most relevant fields: biology, programming, math, engineering, writing, etc.” It will be able to control “physical tools, robots or laboratory equipment through a computer.” And as we can already see, with A.I. partners and suicides related to A.I., it will have a powerful psychological influence on all of us.

Americans could land in a panopticon, constantly surveilled. “It might be frighteningly plausible to simply generate a complete list of anyone who disagrees with the government on any number of issues, even if such disagreement isn’t explicit in anything they say or do,” Amodei wrote. A.I. could “detect pockets of disloyalty forming, and stamp them out before they grow.”

Subscribe to The Times to read as many articles as you like.

Maureen Dowd is an Opinion columnist for The Times. She won the 1999 Pulitzer Prize for distinguished commentary. She is the author, most recently, of “Notorious.” @MaureenDowd • Facebook

© The New York Times