My Take: Seven Ways to Avoid Losing Your Job to AI

Tyler Cowen wrote Seven Ways to Avoid Losing Your Job to AI for The Free Press. The short version: look for messy jobs that are hard to describe, be wary of work from home, become genuinely good with AI tools, work in biomedicine, run experiments, gather data, and get hands-on with energy. The common thread is that AI eats describable, repeatable, screen-bound tasks first, so your best move is to orient away from those.

There are real merits here. The messy-jobs framing is genuinely useful, and the data and energy points are solid. Those are the actual bottlenecks of the AI era and both are turning into hiring markets. As an economist’s argument it is honest too.

Two things give me pause.

Nothing is predictable at this speed. Capability jumps in the last two years have repeatedly invalidated expert predictions written six months earlier, and nothing suggests that stops. Any list of safe jobs written today has a shelf life of months. Staying adaptable beats picking a supposedly safe lane, because you cannot bet a career on seeing the board that far ahead.

The AI proficiency advice is also less of a hurdle than it sounds. Using these tools well is closer to mastering Excel than to learning to code. Pick up the advanced patterns: the formulas, the macros, the prompts that actually work. Within a month or two you are better than 95% of the population, mostly because nobody bothers. Just like the colleague who summons a pivot table from nowhere, you look like a wizard. A few weekends of deliberate work beats a career change any day, which turns the scariest item on Cowen’s list into the easiest.

The direction is right. Spend a couple of months getting genuinely good with the tools, and stop over-indexing on predicting which jobs survive. Nobody knows, and the goalposts move every quarter.