Americans are understandably anxious about artificial intelligence, and politicians are eager to assuage their fears before the midterm elections. When confronting the perils of lost jobs and skyrocketing utility bills, some of my colleagues feel compelled to hit the brakes on President Trump’s all-gas approach to AI. Sen. Bernie Sanders (I., Vt.), for example, recently called for a moratorium on data-center construction.
We must resist that urge. No amount of regulation will stop AI development. Slowing innovation will ensure that China writes the rules for the next century.
If the Democrats retake Congress in November, we must advance policies that acknowledge the automation-related job loss and rising utility bills Americans endured long before the arrival of ChatGPT. AI certainly could exacerbate those maladies, but we can better address them by taking advantage of the AI infrastructure build-out to spur overdue policy change.
Consider the job market: Whatever disruption it endures, studies suggest AI will also generate millions of new positions. We must train a workforce for those AI-enabled jobs of the future, yet federally funded programs too often fail to deliver workers with the skills that employers need. Tax credits could create incentives for partnership between employers and universities in teaching, curricula and hiring. Employers would directly communicate to schools the skills they seek in an AI-driven economy, providing graduates a clearer path to employment.
We must also encourage employers to work with unions, high schools and nonprofits to invest in vocational training and trade apprenticeships. Massive data-center and chip-factory projects expose severe shortfalls in skilled labor. Ford Motor has a shortfall of 5,000 highly paid mechanics, while persistent electrician shortages inflate housing construction costs. Employer tax credits for investing in and hiring from these training partnerships could expand jobs across all these industries—and they have bipartisan appeal.
Electricity rates are subject to inflationary pressure from the rapid growth of energy-gobbling data centers. We can alleviate the problem without spending taxpayer dollars by offering AI companies what they want most: speed. In the hypercompetitive race to build data centers and chip plants, every delay in gas-turbine delivery, local permits, electric-grid interconnection and environmental clearance slows AI innovation and raises costs for investors. Congress can establish a single permitting process to accelerate AI data-center construction and grid interconnection on terms that protect local ratepayers and their communities.
Congress would fast-track approvals for companies building microgrids or self-funding the grid and battery improvements needed for the additional power load. They could reduce pollution if Congress enabled direct procurement of industrial-scale renewables and restored tax incentives for energy storage and other promising alternatives.
Effectively navigating an AI-powered future requires bipartisan solutions that neither hit the brakes on the industry nor stomp on the accelerator. Instead, policymakers need to grab the steering wheel.
