U.S. Sen. Bernie Sanders, I-Vermont, has established himself as one of the most vocal anti-artificial intelligence voices in American politics, and his latest proposal is his most radical yet.
Sanders and progressive U.S. Rep. Greg Casar of Texas are now crafting legislation that would ban artificial "superintelligence" outright, with text expected in the coming weeks.
Sanders and Casar's proposal also has the stated goals of "pausing advanced AI development," establishing a new Cabinet-level federal agency and directing U.S. international policy to "ban superintelligence around the world."
Bernie Sanders' proposal is the most extreme AI-related legislation to date
It isn't clear how much support anti-AI legislation would garner, certainly while Republican majorities remain mostly rallied behind Trump's pro-AI posture.
But the Sanders proposal is the most extreme AI-related legislation to date.
Americans have grown increasingly favorable to slowing AI development, though to this point the main target has been data centers, the physical infrastructure required to develop and run advanced AI models.
Not only is Sanders' legislation the most extreme anti-AI proposal yet, but it also centralizes more power in the federal government to regulate AI development than any bill before it. Sanders wants to hand anti-AI advocates new federal tools to slow it down, even as it risks setting America back in the race to keep pace with China.
The rationale Sanders and Casar want to present is that superintelligent AI is an extremely dangerous prospect that demands regulation. They set the penalties for developing advanced AI models at the same level as those for illegally building nuclear weapons, treating the pursuit of the technology itself as akin to building a warhead.
Concerns about the development of these models aren't completely unfounded. Artificial intelligence has the potential to change mankind as we know it, and there are plenty of projections for how that could go wrong.
AI capable of improving itself could also lead to miraculous innovations: revolutions in how we scan for cancers, discover new drugs and countless other possibilities for immense good. For the economy as a whole, it promises to streamline processes across industries, automate rote work and let human workers focus on more productive tasks.
However, critics worry about the speed of these developments, fearing AI could reach the point of self-improvement before the guardrails needed to keep it controlled are in place. The rampant Skynet doomsday predictions à la "Terminator" go a bit over the top, but there is merit to building guardrails around a technology that could soon eclipse human intelligence.
What Sanders says is a guise. He just wants to limit growth.
There is a difference between regulating AI and calling for progress to grind to a halt.
Sanders' reflexes are the latter, and at their core, anti-growth. He has spent much of this year rallying against data centers, mostly on overblown concerns about water and electricity usage. Sanders even cosponsored legislation with Democratic Rep. Alexandria Ocasio-Cortez of New York for a national data center moratorium, which would seriously limit America's ability to continue developing and using AI.
Sanders reveals his true motivation when he turns to job losses. At a rally in April, after citing a statistic that AI could displace "half of all entry-level white collar jobs in the next one to five years," the progressive senator said, "That is what the oligarchs are pushing. So what do we do about it?"
Apparently, what Sanders wants to do about it is regulate the industry into nonviability, whether through bans, red tape or other measures designed to choke off growth.
Should we have banned the development of the internet to preserve the millions of jobs it rendered obsolete? Should we have blocked the automobile to save the carriage drivers, stable hands and horse breeders?
Fears of how technological change would upend life and the economy were present in both cases. Yet we pushed forward, because that was the right course.
The obvious difference, the speed at which this technology is arriving, doesn't change the dynamic. Protecting industries by outlawing innovation doesn't save the jobs likely to be lost; it only delays the moment workers have to start adjusting to the new reality. There is something uniquely American about pushing forward anyway, and something uniquely anti-American about holding ourselves back from the next frontier.
This mentality doesn't just arise from AI.
"What do you think is going to happen to cab drivers, and Uber drivers, and truck drivers, and bus drivers, and all of those many, many millions of workers who are now involved in transportation?" Sanders asked at that same rally, about the development of self-driving cars.
The reality is that Sen. Sanders fears the tech revolution we're in the midst of – not because he's scared it will fail, but because he's scared it will succeed. It's true that AI and other technological developments will render certain occupations obsolete, but history hasn't been kind to those who tried to stop progress for that reason.
America is poised to become a different economy in the coming decades, built largely on artificial intelligence. That's a scary thought to many, and I understand why that uncertainty makes people reflexively oppose such change. But fear of growth is not the right answer. Americans are unhappy with economic stagnation, and the best way out of it is through growth, even if that growth changes the world we live in.
Dace Potas, a law student, is an opinion columnist for USA TODAY. Follow him on X: @DacePotas
This article originally appeared on USA TODAY: Bernie Sanders' AI ban is most extreme to date | Opinion
