Tuesday, March 24, the College of William and Mary’s Department of Economics hosted Jed Kolko, an economist and researcher, for a talk on the impact of artificial intelligence on the labor market in the Sadler Center’s Tidewater rooms. Kolko discussed labor market patterns and the increasing effect of AI on economic trends.
Kolko, the Department of Economics’ March Tang Scholar in Residence, is currently a senior fellow at the Peterson Institute for International Economics.
Kolko previously served as undersecretary of commerce for economic affairs for the United States under President Biden’ administration. Earlier, he was the chief economist at Indeed and the chief economist and vice president of analytics at Trulia. His work centers around economic analysis of public policy regarding labor markets and technology.
Kolko noted the significance of working across such a wide scope of roles. He explained how he believes these experiences help him better understand the concepts he handles daily.
“It’s always been very important for me to work in different kinds of places and see how different kinds of organizations do research and use data,” he said. “I feel like having worked in a range of different kinds of organizations has helped me see how research and economics gets done from a lot of different angles.”
Early in the talk, Kolko made clear that he could not provide a definitive answer to how AI will affect the labor market—something he said is nearly impossible to determine.
“The truth is, this talk really isn’t about AI in the labor market,” he said. “‘AI in the labor market’ is kind of an excuse to take a step back and talk about something else. What this talk is really about is how we make sense of something we don’t understand.”
However, he noted that the mass implementation of AI has still led to widespread concern.
“In a very deep sense, kind of nobody understands AI,” Kolko said. “We don’t really know how it works. But what is striking is how much of a potential impact AI could have and how much concern, excitement, anxiety and investment AI has spurred for a technology that even its creators fundamentally don’t understand.”
Kolko stressed that comparing claims regarding AI to the evidence helps reveal how narratives are constructed.
“This is a talk about storytelling and various forms of bias, how data gets used, some of the limitations of the data that are used for these sorts of analyses, the data that we wish we had and some ways in which history and past technological innovations are useful,” he said.
Kolko noted that inconclusive labor market statistics and their divergence from historical trends obscure understanding of AI’s impact. “Right now, the U.S. labor market is sending mixed signals: The key measures that we use to understand the state of the labor market are pointing in different directions,” he said. “These mixed signals make it very possible to really choose your own adventure about what you believe about the labor market.”
For students who seek to enter the job market in the near future, Kolko pointed to low hiring as a particularly concerning indicator for their prospects.
“The low hiring rate has a particular implication for you, and that is that low hiring hurts young workers,” Kolko said. “When there’s a low hiring rate, even when the unemployment rate is low, it means people who are trying to get into the labor market for the first time can’t as easily.”
Jenny Wang ’29 found Kolko’s talk especially relevant for college-aged students. She highlighted the importance of understanding exactly how AI may impact these workforce outcomes.
“It’s really important, especially for young people our age, to learn more about AI and how it’s going to affect the labor force,” said Wang. “We’re going to be the next generation that is affected by all of this.”
While some have attributed low hiring to AI, Kolko remains skeptical of this view.
“People have wondered whether AI is responsible for the low hiring rate, and I think the answer is probably not,” he said. “The likeliest answer seems to be that there was a lot of over-hiring during the pandemic, and we’re paying the price of that now.”
Kolko attributes the emphasis on AI in the labor market to an effect he calls “pundit projection.”
“Pundits, or anyone who tells the stories, writes the narratives, does the research, essentially generates the discourse about AI in the labor market,” he said. “In other words, they worry about the things that they think might affect themselves.”
According to Kolko, reports of both labor market outcomes and company performances tend to overstate AI’s impact.
“AI is not driving the trends we’re seeing in the labor market, even though there is an incentive to see everything equivalent to AI,” he said. “CEOs and lots of other people have an incentive to attribute what they’re doing to AI, whether or not it’s true.”
Kolko said that individuals in the workplace should focus on being adaptable and developing new skills to avoid being outpaced by AI. In this respect, he expressed hope for college-aged students early in their careers.
“Who can adapt might or might not have much to do with whether their job is exposed to AI or the automation risk of AI,” Kolko said. “Almost all of you who are not my generation will adapt more easily than I will.”
Kirsten Johnson ’26 MPP ’27, a public policy student who will soon be graduating and looking for work, found that Kolko’s perspective on potential AI replacement in the workforce eased some of her concerns.
“I have a stake in the labor market, because obviously we’re all going into that pretty soon,” she said. “It gave me a lot more hope than I was expecting. I was expecting to come out of this more depressed about AI and its labor market, but when he pointed out how students that go to a pretty elite university and are U.S. citizens really have amazing opportunities and do have adaptability, [it] made me a lot more secure.”
Despite what Kolko views as the overattribution of economic consequences to AI, he said it is still necessary to work toward increased awareness of AI’s impacts on the labor market. While this depends on clear, reliable data, he noted that acquiring this information has increased in difficulty.
“There’s been a lot of news about government data being under threat,” Kolko said. “These risks will make it harder for us to see what is happening today, compare it to the past and imagine what the future might look like.”
Kolko ended the talk by assuaging students’ worries about AI’s accelerating role in the workplace.
“Be the person using AI,” he said. “Figure out what the boundaries are of what AI can do. Think about how it is helpful to solve problems, to understand other people.”
