College graduates across the country recently booed commencement speakers discussing AI's labor market disruption, reflecting widespread anxiety about job security. The technology threatens entry-level positions particularly hard, with computer programming shifting from writing code to managing AI agents. M.I.T. economist Daron Acemoglu and Wharton professor Ethan Mollick joined former Salesforce and Meta AI executive Clara Shih in examining how workers should adapt to this uncertain landscape.

Shih, who recently launched a startup, noted that AI agents enabled her company to accomplish in days what previously required dozens of employees and multiple law firms. This displacement is both wonderful and horrible, she said. Acemoglu countered that economics shows excess entry into occupations creates costly distortions, raising concerns about a flood of AI-enabled entrepreneurs.

Dean Ball offered a more measured view, arguing that highly automated professions like coding represent only small portions of total costs in sectors like mining and agriculture. Changes will happen gradually through tiny marginal automations, creating time for workers to adapt. Meanwhile, large corporations like Walmart face different challenges reorganizing around AI agents while maintaining physical operations and organizational structures.

The panel agreed that while startups may boom due to AI supercharging entrepreneurship, corporate transformation will proceed more slowly. This differential pace provides a window for policymakers and educators to prepare workers for hybrid roles that combine human judgment with AI capabilities.