U.S. Sen. Mark Kelly has unveiled a white paper titled “AI for America” that lays out a plan to avoid disruptions in the labor market and in energy and water supplies while building public trust.
Kelly appeared virtually at an Arizona State University panel discussion in Tempe on Friday to elaborate on the proposal, along with other experts.
“Every time we’ve had a major innovation, whether it was building the first electrical grids or sending people to the moon or the beginning of the internet, we have in our country tended to put people first. And AI should not be different,” Kelly said.
“But I think there’s also a lot of risk here. We could wind up leaving people behind without a clear path to a good-paying job. And on top of it, the infrastructure that we need to support these AI data centers is going to put a real strain on utility companies.”
Kelly proposes that the AI technology industry pay into an “AI Horizon Fund” that would pay for workforce training and for energy and water infrastructure investments.
“If companies are going to make billions of dollars building these AI tools, and they’re going to use American infrastructure to do this and use American talent and American innovation, then it just makes sense that they help strengthen the foundations that made all of this possible in the first place,” he said.
ASU President Michael Crow, who moderated the discussion, noted that new technology — even going back to writing and printing presses — has always been enormously disruptive. But AI is different.
“AI is both, for the first time ever in these kinds of technologies, both disruptor and enabler,” Crow said. “The tool itself can be unbelievably disruptive to work and labor and all kinds of other things, like learning, but also at the same time, it’s a tool that can empower learning, empower workers to do new things, to move in new directions. The tool itself has a dual existence.
“How do we use this tool in the most constructive way?”
Julie Su, senior fellow at the Century Foundation and former acting U.S. Secretary of Labor in the Biden administration, said that the first question must be, “How will this make life better for working people?”
Su said that in other technology innovations, like the assembly line, workers were part of the discussion, which resulted in protections like minimum wage and the 40-hour work week.
“It wasn’t just the innovation that happened. It was that we developed a structure around it so that we could both innovate and protect working people,” she said.
Jason Goldman, who was the first chief digital officer of the White House during the Obama administration, previously worked at Google and Twitter and said that technology companies don’t naturally think ahead to the social implications of their inventions.
“My experience working inside Silicon Valley is when you’re at the beginning of a boom like this, there’s a zeal in that you believe this technology is going to build a better future. You believe it’s going to be good for people,” he said. “There’s a fever, and it encourages you to look at problems that arise as bugs in the system that you can deal with later as opposed to features that you didn’t want to happen.”
Kelly said that while tech companies can’t predict all outcomes, they can contribute to the AI Horizon Fund to pay for workforce training at universities like ASU, plus community colleges and unions.
“Of course there’s going to be some job disruption, but AI also provides the opportunity to create new jobs,” he said.
“When there are new opportunities, we have to have the training programs in place and somebody has to pay for that. And I don’t think it should be on the backs of the American taxpayer to foot the entire bill.”
Crow said that people want jobs with dignity that contribute to society, and universal, lifelong learning can keep them connected to that.
“So that throughout your entire life you can access what you need to be able to adapt to the quickening pace of technological change,” he said.
In addition, the AI Horizon Fund would help pay for water and energy infrastructure improvements needed to build all the data centers that power AI.
Su said she likes that the AI for America framework says that companies cannot just think about return on investment.
“These data centers need things that are traditionally thought of as public goods. And so the question becomes why should public goods be completely upended by the needs of private companies without the private companies having to play a bigger role in meeting that need themselves?” she said.
And AI can help solve the problems it creates, Goldman said.
“If you’re going to create a system in which there’s going to be this tremendous energy demand, the system you’re building should help us get to that being possible,” he said.
“We should be able to incent the outcomes that we want as opposed to abstract problem solving.”
