Microsoft executive's AI-generated 25-page report sparks debate: does outsourcing thinking to AI enhance productivity or weaken our minds?

At Microsoft's December shareholder meeting, President Brad Smith demonstrated Copilot's Researcher Agent by using it to generate a 25-page report with 100 citations about a business problem from seven or eight years ago, impressing colleagues who asked to learn his method—raising the question of whether they then used AI to summarize rather than read the full report themselves.

Education scholar Audrey Watters and New Yorker writer Jessica Winter warn that outsourcing thinking to AI may atrophy cognitive skills: the less we practice reading, writing, and deep thinking ourselves, the weaker our ability to think deeply becomes, and relying on AI to do the mental work denies us the consolidation of memory and knowledge that comes from effort.

For students and educators, the concern is concrete—if schools adopt AI tools to summarize readings and generate answers, students may graduate without practicing the critical thinking skills employers expect, while large tech companies gain influence over what and how public schools teach.