A MESSAGE FROM BLOOM ENERGY

AI growth is outpacing power infrastructure, making energy strategy a board-level concern. The 2025 Mid-Year Power Report highlights grid delays, rising onsite generation, and the strategic importance of power access.

Read report

Despite all the polling showing that Americans want quick, decisive regulatory action on artificial intelligence—and no shortage of hearings about the best legislative approach—Congress has yet to set real guidelines for the technology. Arizona Senator Mark Kelly is hoping to change that with his new “A.I. for America” road map, which calls for the creation of a dedicated federal fund to invest in infrastructure and worker retraining—paid for by the very A.I. companies that are straining U.S. infrastructure and threatening to replace American workers.

The idea, which Kelly calls “common sense,” is to ensure that the growth of the A.I. industry actually benefits ordinary citizens, while still “sustaining innovation and success” and all that. Other core elements of the plan involve educating K-12 students in A.I. and establishing guidelines—negotiated with labor union leaders—for how it should be used in workplaces. Kelly hopes to leverage the “enormous profits” that A.I. companies are ostensibly generating to fund the plan—even though, as I’ve reported ad infinitum, few Big Tech companies, besides chipmakers like Nvidia, are actually generating profits from A.I. right now.

In at least one important way, Kelly’s road map is aligned with President Trump’s A.I. Action Plan: It assumes an inevitable diffusion of artificial intelligence across society, despite its known flaws and safety concerns. Dr. Sarah Myers West, the co-executive director of AI Now, applauded Kelly’s focus on issues like environmental stewardship, cost-sharing, and community partnership, which she called a “much-needed antidote to the reckless A.I. infrastructure build-out.” But she still found it troubling, she told me, that lawmakers across the political spectrum have bought into the pursuit of A.I. supremacy. “There is a huge incentive for companies to persuade policymakers that the current trajectory is inevitable,” she told me. “These companies are turning to government to derisk their portfolios, and policymakers need to use that as leverage to act in the interest of the public rather than underwrite a precarious industry.”

On Tuesday, I got a few minutes to chat with Senator Kelly about his plan. As always, the following has been lightly edited for clarity.

Ian Krietzberg: Where did the idea of establishing some sort of federal, corporate-financed fund for A.I. originate?

Senator Mark Kelly: When I got to the Senate in 2020, we heard a lot about A.I., but there was no real widespread awareness of how fast things could change. Then ChatGPT came out, and people saw this was a rapidly moving technology. I became really concerned about what this means for workers. What if we wind up in a situation—five, 10 years from now—with millions of people losing their jobs because of A.I.? We’re not seeing those numbers yet, but I have anecdotal cases of a C.E.O. telling me that instead of hiring 20 analysts, they were going to hire two.

So I sat through most of those hearings on A.I., and we talked a lot about it, but then I didn’t see this thing moving along fast enough. Then when you see what the White House put out—a lot of it makes sense, and some of it is very reasonable—but what it’s lacking is, How are we going to pay for these things? Are we going to put all this on the American taxpayer and have government-only funded programs? Or would a better solution be a public-private partnership?

Companies are going to gain a lot—actually, the most—out of A.I., and they should be a partner with government in how we fund the workforce, training and retraining and upskilling, but also a partner in funding the infrastructure we need to build to make sure we stay ahead. My fear is, if we don’t do this, we’re going to find ourselves in a race with China that we’re going to lose, and that’s not an option.

A MESSAGE FROM BLOOM ENERGY

As AI adoption accelerates, power has become the defining constraint—and opportunity—for data center growth. Our latest 2025 Mid-Year Power Report reveals a dramatic shift in how industry leaders are planning for the future.

Read the report.

Big Tech and major A.I. players have been fighting regulatory legislation for years now. What’s been their response to your proposal?

Well, we’ve talked to not only tech companies, but A.I. companies. We’ve talked to universities, and folks formerly in government, and union leaders and labor [leaders], and over a period of months we’ve worked with them to reconfigure this proposal. So we took their input—not all of it, but some of it—across that whole range of people.

Under the current administration, which has taken a strong anti-regulatory stance on A.I., what are the odds of this initiative, or something like it, actually coming to fruition?

That conflict is always going to be there. The conflict between, do we want maximum innovation, or do we want to try to control the innovation to some extent, to maybe get a better outcome. That might not be what the innovators want—they want free rein. I get that, and I am very pro-innovation. I want this A.I. boom we could potentially experience to be as positive a thing as it can for our country. But that also means it needs to be positive for workers. So we’ve got to get ahead of this. We didn’t get ahead of social media, and now we’re still struggling to figure out, how did we get here, and why didn’t we put some guardrails in place to prevent some of the stuff we see on social media today?

How would the money in your A.I. Horizon Fund be equitably distributed? The projects you lay out in your plan are pretty diverse. For example, data centers aren’t equally distributed across the country, so you could imagine a disparate impact across states.

At the top level, you figure that out through a lot of collaboration. I think a good example is my SHIPS for America Act. We spent, like, a year talking to people and refining it, sitting down and taking the input seriously, then integrating it into the legislation. When we finally introduced the legislation in the Senate, we got a lot of support because we listened to people and incorporated their good ideas. We didn’t incorporate every idea. Some just wouldn’t work. That’s the same approach we have here.

Many researchers have pushed back against this idea that A.I. should be ubiquitous, or that it should be adopted in schools, for example. Your plan actually calls for more student exposure to A.I. So I’m curious how you think about this notion of the inevitability of artificial intelligence.

When I use ChatGPT, I see a remarkable technology. And we are, what, two years into this being publicly released? It is very, very powerful. As I talk to folks in industry, they think this is going to keep moving along at a rapid pace. I know what you’re saying. There are other people that are saying, Well, this isn’t going to turn out to be what we thought it was. But we’ve got to make sure we don’t find ourselves looking back five or 10 years in the future saying, Well, we really messed that up, because we’ve got 12 million people out of work, and energy costs are up 50 percent because we didn’t put in the additional capacity to the grid—and we should have done all these things upfront, because this thing did move along this fast.

I’m not saying that’s the worst case. It’s actually, in some ways, the best case, because it means the technology works. It helps us grow our economy, and ultimately, maybe if we manage it right, it creates more jobs. Every time we’ve had advancements in technology and innovation and productivity, we’ve had a bigger economy and more jobs. If we manage this well, I think we could have the same thing. But we could also screw it up. We need to have a plan for what we want this country to look like 10, 20, 30 years in the future.

That’s all for today. I’ll see you next week.

Ian

Need help? Review our FAQ page or contact us for assistance. For brand partnerships, email ads@puck.news.

You received this email because you signed up to receive emails from Puck, or as part of your Puck account associated with {{customer.email}}. To stop receiving this newsletter and/or manage all your email preferences, click here.

Puck is published by Heat Media LLC. 107 Greenwich St, New York, NY 10006