What does it mean to stay human in a world increasingly run by machines? Written just before the AI wave broke, Kevin Roose's Futureproof offers something rarer than prediction — a practical and quietly urgent philosophy for working in the age of automation.

Kevin Roose wrote Futureproof before ChatGPT existed, before generative AI became a dinner table conversation, before the anxiety it describes became the ambient condition of working life. And yet it reads less like a prediction than like a diagnosis written just ahead of the symptoms appearing.

Roose's argument is directed at AI optimists, those who assume that automation will naturally improve our lives, create better jobs, and free us for more meaningful work. He systematically challenges their four central claims with concrete evidence, and with the kind of grounded reporting that makes uncomfortable conclusions hard to dismiss.

What makes the book worth returning to is less its nine rules — though they are useful — and more the underlying idea that animates them: that the response to automation shouldn't be to race machines, but to become more distinctly, irreducibly human. Roose calls this leaving your handprint. Not optimizing, not competing on speed or scale, but doing the things that carry the mark of a specific person's judgment, sensibility, and care.

Some of his warnings have only grown more urgent. When he describes devices that function less like tools and more like managers — systems that nudge, measure, and redirect human behavior in the service of engagement algorithms — he is describing something most of us now navigate daily. His case for "demoting" those devices, for resisting what he calls machine drift, requires a kind of deliberate friction that our habits and our workflows have been designed to eliminate.

The structural solutions he advocates — large-scale policy protections paired with local, human-scale networks, what he calls "big nets and small webs" — remain, if anything, more necessary than when he wrote them.

To read Roose in conversation rather than in argument, his podcast Hard Fork, co-hosted with Casey Newton, is a natural companion. Where the book reasons carefully toward conclusions, the podcast catches technology in motion — provisional, sometimes contradictory, always consequential. The two formats together give a fuller picture of a writer genuinely trying to think clearly about things that keep changing shape.