Colleges aren’t preparing their students to join a labor market being transformed by AI, and the consequences may be psychological as much as economic, a leading economist has said.
Tyler Cowen, a professor at George Mason University, said that universities were “producing a generation of students who will go out on the labor market and be quite unprepared for what they’re expected to do.”
Students are being taught in ways that were “counterproductive, even,” he said, in a conversation with podcaster Azeem Azhar that aired Tuesday.
He said studying at college was “the time to do it,” referring to learning to work with AI.
Cowen didn’t name specific academic disciplines, but his comments come amid a debate about how AI could reshape the type of knowledge work that college students expect to do after graduation. As generative AI tools become embedded across white-collar industries, proficiency with these systems is fast becoming a baseline expectation.
Cowen said he didn’t think AI would destroy jobs, but would reshape hiring practices, career trajectories, and productivity standards.
Cowen said new graduates may struggle to get hired, and warned that the damage may run deeper than lost wages.
“The output will be lower, but I think many of the highest costs will be psychological — people feeling they do not fit into this world,” he said. “And they’ll be somewhat correct.”
Related stories
Business Insider tells the innovative stories you want to know
Business Insider tells the innovative stories you want to know
A disruptive force in education
Cowen’s warning joins a growing chorus of concern that elite institutions and traditional curricula are not changing fast enough to keep up with rapid technological change.
Open AI’s VP of education, Leah Belsky, said on the company’s podcast last week that college graduates need to know “how to use AI in their daily life” and should be taught to use it “in ways that will expand critical thinking and expand creativity.”
Google DeepMind research scientist Stefania Druga told Business Insider in May that if AI can complete a student’s assignment, teachers need to change it.
“If an AI can solve a test, it’s the wrong test,” she said.
Druga, who helps design AI education platforms for children, said young people were using it to shortcut learning entirely, rather than being taught to use it as a tool for co-creation.
Educators are adopting diverging strategies: some are doubling down on analog tools like handwritten essays and oral exams to preserve academic integrity, while others are embracing AI to build more personalized, harder-to-game assessments.
