As the AI boom continues to grow, economists are exploring the potential effects its abrupt disruption could have on the economic geography of the U.S. According to a recent study, Baton Rouge and New Orleans are among the list of large metropolitan areas with low employment exposure to large language models, or LLMs. In the study by Scott Abrahams, an LSU economics assistant professor; and Frank Levy, an economist, Duke University faculty affiliate, and MIT emeritus; the economists looked back at the downstream effects of the decline of U.S. manufacturing employment in the 1980s to predict how another major shock to national production, such as LLMs or AI, could impact the U.S. without policy intervention. “There are certain parts of the country where it is more likely for AI to replace or assist people,” Abrahams says. “Baton Rouge’s top industry is petrochemical, the port and related things. The top occupations are a lot of health care, nursing, sales. … If you look at all the types of occupations and industries in Baton Rouge and you see what share of things people do in those jobs could be done by AI, it tends to be a bit lower than in other places like Austin or San Francisco.” The key determinant of how disruptive AI will be is based on how much AI improves and how fast people start implementing it, Abrahams says. “Baton Rouge is not going to become Silicon Valley,” Abrahams says. “Maybe that’s what you’d hope for to get investments, but usually it’s pretty persistent over time. The general mix of industries and occupations in place is a lot slower to change than people. We think people can adjust faster, which is why we’re predicting they might relocate.” Abrahams says Baton Rouge has the potential to be an attractive city for professionals to move from cities most vulnerable to having an abrupt AI disruption such as Washington, D.C., San Francisco, New York City and Austin. Baton Rouge has high education levels, affordable housing costs, and its leading industries are less likely to be replaced by LLMs, according to the study. He recommends policymakers should moderate the speed of adoption of LLMs. He says the more time people have to adjust to the technology, the fewer issues cities will face.