Nobody takes you aside anymore
Print taught a generation of designers when to stop. The constraints that did the teaching are gone, and nobody has priced what they were worth.
Somewhere past your second requested move at a press check, the pressmen would start getting antsy. Nothing was said. The room just changed temperature. Ask for a fourth move and the sales rep would take you aside, friendly, hand on your shoulder, and explain nothing at all while making everything clear: the schedule behind your job had been sold to the next customer, and the next one after that. “You can do whatever you want here,” everyone told you. The building disagreed.
That was only the top layer of the constraint stack. Underneath it sat the economics: every look at your own work cost fifteen minutes of makeready and fifty sheets, of which you kept one. Underneath that, the physics: ink channels run one direction, so the extra red that fixes your flesh tones also runs through the type below the photo, and now you have a different problem. And at the bottom, the plates. You cannot fix a bad design on press. Once the plates are made, your universe gets a lot smaller, and everything after is negotiation inside someone else’s geometry.
I keep thinking about what that stack produced. It produced designers. The good ones could print. They did their own separations. They thought about trapping while they were still designing, since fixing it in pre-press was someone else’s expensive guess. By the time their hand touched a blank page, the press was already running in their heads, and their upstream choices quietly contained all the downstream physics. I was never a designer myself; I was the one with the loupe and the schedule, watching this formation happen project after project. The crucible made them. Nobody enrolled in it voluntarily, and nobody came out of it unchanged.
What the crucible taught was not technique. Technique you could get from a book. It taught that you cannot design forever, and you cannot check forever. It taught that blue and green are genuinely hard to print, that a great printed piece was kind of a miracle, and that you walk into the plant knowing which two moves matter because two is what you’re going to get. Above all it taught you to be OK with “ok.” Sign the sheet. Let it run.
Then the constraints left. All of them, every layer, inside a decade. The 50th sheet became free, the makeready became instant, the plates became a file you can re-upload after publication, and nobody takes you aside, ever, about anything. And I watch people fiddle forever, in design tools and in documents and in code, fifty revisions deep with no sign of landing, and a thought arrives unbidden in the voice of an old plant manager: you wouldn’t have gotten away with that in my shop.
I want to be careful here, because the curmudgeon path is too easy. I don’t miss the waste. I spent this spring iterating a book layout in ways that would have cost a fortune in 1999, and the freedom is real, and I shipped the book. Free iteration is a gift. But it’s a gift to people who already know when to stop, and the thing that taught my generation when to stop was the very apparatus the gift replaced. Discipline used to be supplied by the building. Now it has to be supplied by you, and self-discipline is far more expensive than the imposed kind, because nothing fails loudly when you skip it. You just quietly never finish.
This is where it stops being a story about print. The machines arriving in every field right now are doing to knowledge work what digital did to the pressroom: absorbing the constraints. The grunt reps, the consequences, the slow expensive confrontations with a world that pushes back, the junior years that felt like waste and were actually the curriculum. We celebrate every piece of absorbed toil, the way nobody mourned the burned sheets. I’ve written before about the expertise walking out the door faster than it’s being replaced. The harder version of that problem is this one: the crucible was doing double duty. It made the work, and it made the people. We automated the first function and have no plan for the second.
I don’t think the answer is nostalgia, and you can’t rebuild a Heidelberg in your basement. The answer is that the crucible has to become deliberate. Reps you impose on yourself because the world stopped imposing them. A deadline that means something because you said it out loud to someone whose opinion you fear. A budget of two moves, adopted voluntarily, with the third one forfeited on principle. A plate-burning moment you schedule, where the design is done because you declared it done, knowing the declaration is the only plate left. The book I just sent to the printer is, among other things, seventy-seven of those reps, collected over twenty-five years from rooms where the constraints were still real.
So, a question, and I mean it practically: in your work, right now, what takes you aside? What is the hand on your shoulder, and when did you last feel it? If the answer is nothing and never, you haven’t escaped the constraint. You’ve inherited its job.
You don’t have forever. You need to be able to decide. Be OK with “ok.”
The rep isn’t coming. The hand on your shoulder has to be your own.
The agent-shaped org chart
Every real org has the same topology: principal, role-holder, specialists. Staff AI maps onto it, node for node, and the cost collapse shows up in the deliverables that were always just human-handoff overhead.
AI as staff, not software
Two frames for what AI is doing to work. The tool frame makes tools smarter. The staff frame makes roles unnecessary. Those aren't the same product, the same company, or the same industry.
Knowledge work was never work
Knowledge work was always coordination between humans who couldn't share state directly. The artifacts were never the work. They were the overhead — and AI just made the overhead optional.
The work of being available now
A book on AI, judgment, and staying human at work.
The practice of work in progress
Practical essays on how work actually gets done.
How I made Tailscale and NordVPN coexist: the Linux exit node pattern
A practical pattern for running Tailscale and NordVPN together, plus the larger question it exposed: private networks need explicit exit profiles, not one vague VPN toggle.
The 2 a.m. press check
Twenty-five years ago I stood under color-true lamps at 2 a.m., loupe in hand, waiting on the 50th spread. This month an open-source compiler handed me the thing Pantone only ever promised.
How do I get my dev team to adopt AI?
A stub on helping mixed-interest development teams find their own useful ways into AI.
Your best people were always better than you knew
For thirty years firms outsourced capability because their teams couldn't produce. AI collapses the production gap. What's revealed underneath is what was there all along.
Knowledge work was never work
Knowledge work was always coordination between humans who couldn't share state directly. The artifacts were never the work. They were the overhead — and AI just made the overhead optional.
The costume just got cheap
If 80 percent of what you thought was judgment turns out to be pattern recognition, what does that say about you? Not about your job — about you.