What AI Means for American Manufacturing

My great-grandfather, Fred Stoelting was a blacksmith from southern Indiana who came of age in the late 19th Century. During his career, the invention of rubber threatened to disrupt the blacksmithing industry. But rather than panic, he pivoted his business from forging wagons and horseshoes from iron to repairing and re-tiring wagons with rubber.  “The good Lord took [iron] away and put rubber in its place,” he told a local newspaper in 1948. Soon after, Fred retired from the smithing shop, where his grandson (my father) had played model airplanes on the shop’s dirt floor as a child. My father went on to build a well-paying career of his own in Indiana as an aeronautical engineer. One line of work was gone, but new ones took its place.  Today, as AI promises to complete the tasks once done by workers, history is repeating itself. In manufacturing-oriented states like Indiana, many hardworking Americans are especially anxious. It is challenging enough for digital workers to adapt. But if AI shuts down a plant, it can be especially difficult to pivot an entire local economy.  My family history has taught me about the American capacity to adapt. And as the governor of Indiana, I focused on job creation. That’s why I’m confident that American manufacturing will not only survive the AI era, but that AI will actually boost American manufacturing.  Modern manufacturing Indiana ranks at the top in manufacturing jobs and revenue per capita among the 50 states, and as governor, I spent my days pursuing investors who could help ensure that our high rankings continued. Their biggest concerns (and therefore mine) were Indiana’s supply of skilled workers and electrical power, in that order.   Now, he marriage of AI, robotics, and advanced human-machine interfaces constitutes a new industrial revolution. Far from threatening U.S. manufacturing, AI is our best hope of sustaining that sector’s emerging renaissance, which a recent study by Denoitte projects will create a net 3.8 million new jobs by 2033.  Though future factories will employ fewer workers than in the past, AI-based technologies will enable the creation of more, often highly-specialized, factories in the U.S. than we have seen in decades, producing a wider range of goods in safer conditions, with fewer pressures to move offshore in search of cheaper labor. This prediction may seem counterintuitive at first, but smarter robots and manufacturing employment can (indeed, must) coexist in an advanced economy such as America’s. Among the dramatic early examples of this are the bizarrely maligned data centers springing up across America. Factories in their own right (data is a highly valuable “good”), data centers and their associated investment in information processing equipment were responsible for 92% of the country’s economic growth in the first half of 2025, according to Harvard University economist Jason Furman. In other words, without data generation and the related investments and jobs, America’s growth rate right now would be closer to 0.1%, rather than the 4.3% we logged in the third quarter. Whose job security would have been helped by that situation?   An adaptive economy  To be sure, AI will not render people outside of manufacturing jobless either. AI is a tool, immensely powerful but still subject to its users’ direction. For this reason, workers will still be needed to oversee AI’s task execution, discover new uses for AI and its spin-off technologies, and marry them to developments in other fields.  Perhaps the profession most challenged by AI, unfortunately, is my own recent one: politician. Nurturing a thriving labor market in a wider AI-driven economy will require leaders at all levels of government to do things that do not come easily to us, such as challenging the status quo, encouraging policy experiments, deferring final decisions to the people closest to problems, and counseling patient optimism rather than assigning blame for short-term challenges. Retooling American education systems for the age of AI, for example, will test all of those leadership requirements. From primary school curricula, to the construction of high schools, to the funding of research at elite universities, education and politics are awkwardly bound up in the United States—and therefore resistant to transformation or even evolution without considerable drama. But we cannot let that stop us from adapting.   For instance, skills training in the basic components and operation of AI and ever-more immersive digital technologies is essential. In the manner of “shop class” from generations ago, it needs to begin in secondary schools at the latest. For reasons that go far beyond getting a job, young people must understand the dangers and limits of what lies at their fingertips.  In the economy of tomorrow, community colleges and vocational-technical schools—optimized for true lifelong learning with apprenticeship and mentorship programs—will likely be the backbones of public and private education alike. Experimentation will be critical to getting this right. And we must simultaneously support liberal arts programs, since their essential gift of “learning how to learn” has never been more important as we become the discerning masters of AI. Heterodoxy and rigor must persist in thriving liberal-arts institutions, since their alternative, indoctrination, leaves people vulnerable rather than wise to the chatbots. And in the higher education sector we must ensure that engineering, math, and science programs are aligned with hands-on laboratories and are the best in the world.  No technological race in human history will be more consequential than the one before us. AI is here, but it is mainly up to Americans to keep it humane, open, and harnessed to individual flourishing rather than centralized power.  Immigrants who understand the alternatives should be welcomed with open arms, via an efficient reformed legal process, which is clearly a political test in its own right.  The importance of energy Setting energy policy on a new course will be as important as retooling education.  Small towns and rural areas—with their land and underutilized talent—are the natural beneficiaries of America’s AI-driven manufacturing renaissance. However, as we have learned the hard way at times in Indiana, neglecting power and water infrastructure concerns can create tension and hold back growth. Fortunately, there are many options to address energy demands, including small, safe nuclear reactors alongside more efficient utilization of solar, wind, and traditional energy sources—and soon, fusion-driven power.  And energy will be especially important in the AI era. Perhaps the most profound national security challenge of our time, is to win the AI race so that we are able to define and demonstrate how these epochal technologies can help people to flourish in freedom.   Across workforce development, education, power development, and other policy challenges posed by AI, we cannot decree either a standstill or a utopia, and should not pretend otherwise. But there are undeniable opportunities for AI to drive an American manufacturing renaissance. And if we fail to foster it, we risk being forced to play by someone else’s rules entirely.