Wharton School

Ethan Mollick teaches at the Wharton School, and the fiches that reference the institution are all traces of his output: a Substack newsletter, "One Useful Thing", a book, "Co-Intelligence", and a stream of posts analyzed here between April 2024 and March 2026. The school itself is never the subject. It appears as the affiliation that lends weight to a body of empirical work on how organizations actually absorb AI.

That work supplies the numbers the fiches keep returning to. A study with Boston Consulting Group covering 758 consultants recorded 12.2% more tasks completed, 25.1% faster work, and 40% higher quality, while workers using AI on tasks outside the "jagged frontier" scored 60-70% accuracy against 84% for unassisted humans. Survey work on knowledge workers found roughly 70% had tried AI tools and about 20% became regular users. Bottom-up exploration reached around 40% sustained adoption versus around 15% for top-down mandates.

The recurring argument across the fiches is that the blocker is managerial, not technical. In March 2026 Mollick described two firms in the same regulated industry, one 18 months into productive use, the other still convening a committee to approve each use case and still fearing that enterprise models train on company data. He attributes the split to whether one executive will absorb the risk.

Mollick states he accepts no funding from AI companies. What the affiliation contributes is the research apparatus behind the claims.