The Mechanical Professor

When ChatGPT first launched in late 2022, most people saw it as a novelty or a neat party trick. Ethan Mollick’s early experiments with "The Mechanical Professor" were among the first to treat it as a serious economic and structural shock. By trying to offload his own analytical, creative, and administrative tasks to the AI, Mollick showed that LLMs weren't just autocomplete on steroids; they were a viable, albeit imperfect, surrogate for high-level intellectual labor. It was a fascinating, hands-on look at how quickly a highly specialized, prestigious profession could have its core duties unbundled by an API.

Reading this as a startup founder, it highlights a profound shift in how we think about productivity. Traditionally, scaling an organization meant hiring more people to handle administrative overhead. Now, the cognitive friction of running a business—drafting syllabi, grading, summarizing reports, brainstorming ideas—can be delegated to a machine in seconds. The challenge is no longer about finding administrative help, but about managing the quality and drift of automated output. It makes me realize that we are entering an era where the leverage of a single skilled human has increased by an order of magnitude, completely redefining the economics of small teams.

What stuck with me

• Intellectual labor unbundling: Administrative and creative tasks in highly specialized professions can be instantly offloaded to language models.
• The management shift: The bottleneck of modern work is shifting from executing basic cognitive tasks to editing and filtering machine-generated output.
• Solo operator leverage: Small teams and solo founders can now operate with the administrative capacity of much larger organizations by leveraging LLMs.