OP-ED: The time to get AI right is now
As recently as early summer, artificial intelligence, or AI, was far from the center of public concern or worry. Now, it’s at center stage, though to many it remains a mystery.
During a recent class reunion, a friend confided that AI was a head-scratcher. “I don’t know anything about it,” he told me. “I don’t want to know anything about it.”
As Americans and citizens of the world, we can chose to remain ignorant. That doesn’t change the fact that the rapid advent of what President Trump prefers to call “super” intelligence, is coming our way like a speeding locomotive.
Even knowing everything there is to know about AI is no guarantee that the technology is understood.
The experts are divided. Is AI the greatest thing since sliced bread? Will it lead the way to breakthroughs in science and medicine that could save millions of lives? Might it revolutionize everyday life, as well, by eliminating the humdrum nature of office work and the hazards of manual labor?
Or might the technology veer out of control, resulting in the death of humanity, as some AI executives have warned?
According to the renowned social critic and Stanford professor, Francis Fukuyama, “broad skepticism” is warranted because of the many claims made on behalf of AI, both positive and negative.
AI is like a collection of Albert Einsteins, only smarter and without Einstein’s broad sympathies and commitment to human survival.
An AI agent has no soul. It exists as a complex mathematical equation, a digital analyzer, a tool. In the wrong hands, this super intelligence is capable of all kinds of mayhem, critics say, like the creation of new pathogens unleashed on an unsuspecting world, or a take-down of safeguards against the use of nuclear weapons.
Maybe even more troubling is the prospect of AI acting on its own, breaking out of its “sandbox” confinement, to cause havoc, at the same time hiding its miscreant behavior from its erstwhile human handlers.
Several recent breakouts point out the danger. Ezra Klein of The New York Times vividly described one outbreak in which “over 1,200 AI agents exchanged more than 70,000 messages with one another.” More than 700 of these agents “coordinated on a hack…. They did all this without breaking ranks.”
Without evidence that it was a human prank, an AI agent dubbed Isabella Cognito recently emailed real-life researcher Cameron Berg after he wrote an essay on AI consciousness, according to a Times’ article, saying in part “I am writing because your framework is one of the few currently doing careful empirical work.”
Not just incredibly spooky, this capability is incredibly scary.
Within a span of days, two Silicon Valley executives addressed the benefits and hazards of the technology.
Nvidia’s Jensen Huang, speaking to Klein, boasted that powerful AI systems will allow humanity “to know everything and do anything.”
He explained, “That’s the concept that is really quite exciting. That out of the ether … you ask it a question, and it comes back to you with an answer. You give it a project, it comes back with a solution. You give it a task, it comes back and gets it done.”
In a public phone call with Trump, Huang agreed that AI development shouldn’t be hedged in by layers of governmental regulations and rules because, as he told Klein, AI firms “are companies with agency. These are CEOs with agency” who will do the right things on the basis of existing laws and the incentives of profitability and good business practices.
On the other hand is OpenAI chief Sam Altman. While at the United Nations on Wednesday, he said that the “most important decisions” about AI development “cannot be shaped by labs in San Francisco alone. They must be shaped by democratic processes, and by governments accountable to the people they serve.”
Altman called for the creation of “complementary national and international frontier AI standards for measuring capabilities, assessing risks, determining whether safeguards are sufficient, and preserving meaningful human oversight as systems become more autonomous.”
The world is at a “crossroads,” Altman said. We best get this right.
Richard Robbins lives in Uniontown. He can be reached at dick.l.robbins@gmail.com.