What is at stake is whether we let AI do the tasks that make us better thinkers and doers — and who pays for that choice. The essay’s author, who teaches public policy at the Harvard Kennedy School and the Munk School at the University of Toronto, argues that some tasks are “work” to be outsourced and some are “gym” to be practiced.
Daniel Meissler’s “work versus the gym” distinction
The central analytic device comes from AI researcher Daniel Meissler: “it’s the difference between work and the gym.” In the analogy, tasks that must be completed — moving heavy things across a room — should use whatever assistive technology is available, whether a wagon, a forklift or an AI-powered robot. By contrast, at the gym you lift the weights yourself because the point is the exercise, not the moved object. The essay applies that simple distinction to decisions about whether to use AI for a given task.
Why policy memos and assignments are treated as gym work
The author treats classroom writing as gym work. “The writing assignments I give my students are gym tasks, not work tasks,” the piece says, because the act of writing — “thinking and outlining and drafting and editing, making and criticizing and revising arguments” — is how students develop critical thinking skills. Without that repeated practice, the author warns, “those skills will atrophy.” The essay reports that students often hand in AI-produced essays that are “catchy, plausible, grammatically perfect” but “not particularly well-crafted or logically coherent,” and that the author, having developed writing skills over years, can usually tell the difference.
How AI reshapes demand for writers and visual artists
The essay argues AI lets society separate “work” writing from “gym” writing for the first time. Most commissioned writing — instruction manuals, sales presentations, government disclosures, legal briefs — is described as “work”: dry, predictable, and exactly the type of task AI handles well. The author notes that if AI does most work-type writing, “society doesn’t need as many human writers.” The same logic is applied to visual art: many images the world needs — “a corporate mascot, a ‘beware of the dog’ sign, or a packaging label” — are primarily work, not art, and historically those jobs supported some artists. The piece suggests we are now forced to make deliberate choices about how to value art and artists in a world where AI can produce many such images.
What this means for policymakers, employers, and students
Policymakers and teachers: “Policy analysis will definitely involve AI from now on,” the essay states, and educators must help students “reimagine what it means to learn and practice that skill.”
Employers and recruiters: the author adds that “Employers are already noticing” a decline in students’ critical-writing skills when work is outsourced to AI, implying organizations will be watching for those competencies.
Students and job-seekers: the essay documents incentives pushing students toward AI — time pressure, stress, and peer use — and notes a social pressure: students “feel like they’ll look bad in comparison if their peers are all using AI.” The author also calls out the broader incentive problem: “No one pays us to go to the gym; maintaining healthy habits requires discipline.”
A simple, actionable rule for everyday decisions
Beyond prediction or fear, the essay offers a practical decision rule: if a task must be done and no one cares how, let AI help; if the process of doing the task matters to developing skill, don’t outsource it. The argument contains caveats: the AI must be able to perform the job well, its mistakes must be “minimal and correctable,” and it must be secured from cyber-attacks that would influence its results. Given those conditions, the author urges individuals to “wall off our cognitive gym tasks from AI” in the same way one might choose stairs over an elevator.
The author concedes that the boundary between work and gym will shift as humans and AI adapt, and that change will not occur at the breakneck pace claimed by AI companies: “AI is going to fundamentally change the nature of work. Not nearly as fast as the AI companies want you to believe, but eventually it will.” For now, the essay concludes, the distinction is “pretty clear” — and the straightforward injunction is: “Use it on yourself.”
