I graduated from college in 1985 and went straight into graphic design, when the work was entirely physical. Paste-up was the job. You cut type with an X-Acto knife, waxed it down on boards, and hoped nothing shifted before it went to film. That was simply how design was made.
If a client wanted a text change, it was not immediate. In most cases, I got in my car and drove to the typesetter. I would drop off the correction and get it back the next day. If you worked at a large company or ad agency, a driver might handle that run for you, but it was still an overnight process.
In the years before I started, it was even slower. Type was often set in places like New York City or at Forestall in New Orleans, then physically shipped by bus or by commercial flight. Words literally traveled across the country before they landed on a page.
That was the pace of the work.
By the end of the 1980s, that pace began to collapse.
What Paste-Up Actually Was
Paste-up imposed natural limits. It was slow. It was expensive. It required coordination. Every change had consequences.
When I took paste-up boards to a commercial offset printer, they were sent to a specialist known as a stripper. In my experience, this was usually a man with many years in the trade. He worked in a darkroom, manually creating negatives that would later be used to generate printing plates. Cigarette smoke hung in the air, mixing with chemicals that everyone hoped were nonflammable.
It was skilled, hands-on work. And almost all of it disappeared within a few years.
Some people adapted. Others moved on. Roles faded out quietly, not because the people lacked talent, but because the processes they supported no longer existed.
That matters because when workflows collapse, jobs tied exclusively to those workflows tend to disappear with them.
When the Macintosh Entered the Studio
When the Macintosh arrived, artists and designers quickly took to it. Not because of one program, but because the machine itself made sense. You worked visually. You moved things around on a screen in a way that resembled how you worked on a paste-up board.
The computer did not replace creativity. It replaced physical actions. Cutting, pasting, waxing, resizing, and repositioning all moved onto the screen. What once required boards, knives, cameras, and chemicals could now be done with a mouse.
The real transformation came when digital files could move directly to the printer. Once printers could output plates from their own equipment, everything changed. Darkrooms disappeared. Prepress departments shrank. Multiple handoffs collapsed into a single digital workflow.
Designers gained speed and control. Printers gained efficiency. Expectations changed just as quickly. Changes that once took a day now took minutes. The pauses that had once enforced restraint were gone.
Early digital work was not better. It was simply faster. But faster has a way of becoming the standard.
Where the Term “Desktop Publishing” Came From
The phrase desktop publishing was coined in 1985 by Paul Brainerd, the founder of Aldus Corporation. He needed a way to describe what was happening as these new tools came together.
At the time, publishing depended on expensive equipment, specialized vendors, and clearly separated roles. Brainerd’s idea was that professional-quality publishing could now be done at a standard desk using a personal computer, rather than in an industrial production environment.
What made that possible was the convergence of a few changes happening at once. The Apple Macintosh made visual, screen-based layout practical. Aldus built PageMaker to take advantage of that environment. At the same time, printers began accepting digital files and output plates directly, rather than relying on film and physical boards.
Desktop publishing was not just about software. It described a collapse of the traditional workflow. Tasks that had once been spread across typesetters, production houses, and prepress departments could now be handled in one place.
At the time, the phrase sounded like a description of print production. In hindsight, it described a much deeper shift in where work happened and who was responsible for it.
Desktop Publishing Did Not Stop at Print
Desktop publishing was not a destination. It was a starting point.
Paste-up gave way to desktop publishing. Then came web design, which was an entirely different kind of work. Early websites required writing HTML by hand, which was difficult if you had never coded before. As the web grew, so did the capabilities of personal computers. Video, audio, animation, and interactive media followed, each expanding what design could mean.
At every stage, you either learned enough to stay useful or you gradually fell behind. This was not about chasing trends. It was about staying connected to how work was actually being produced and delivered.
That is the pattern I recognize in AI.
The Next Shift Has Already Started
Right now, AI feels like a tool. In many ways, it is a fun one. Desktop publishing felt the same way when it first appeared.
I was recently asked, in a genuinely concerned voice, whether AI had changed my graphic design business. My honest answer was no, not really. But as I walked away from that conversation, I realized something. Her profession may change even faster than mine.
She is an attorney. And all I could think about afterward was the many ways AI is likely to affect how law is practiced. Document review that once required teams of junior associates can now be handled by AI in hours. Contract analysis that used to take days of billable time could now take minutes. Legal research that involved manually combing through case law is being automated. The structure of how law firms bill, staff, and train new lawyers is quickly being disrupted by AI.
I often get asked a version of the same question. Can AI do my job now? Not really. One of the biggest reasons is that real client work is iterative by nature. Clients want changes. They want revisions. They want to explore an option and then back up. Right now, AI does not handle that process as smoothly as a designer working directly with a client. And even if you were to create, let’s say, an illustration, the client will probably ask for a minor change that AI will struggle with.
But I have started using it. One of my favorite uses is to bounce ideas off it. This is a luxury that solopreneurs have never had. I use it for SEO, specialized website coding, and more. AI shows me directions I might not have considered, collapsing the time between idea and execution.
I think of AI today the same way I once thought of clip art. Early clip art was simple and one color, maybe two. Stock photography was around, but similar-quality images became cheaper and more accessible. Then, free stock rivaled paid libraries. Now AI generates imagery on demand. Each step in the growth of “clip-art” changed production and possibly eliminated someone’s job. The illustrator who drew generic business people for stock catalogs. The photographer who shot standard office scenes. The designer whose entire role was sourcing and licensing images.
If you fight new tools outright, you lose by default. Not because the tools are perfect, but because the work moves in that direction whether you participate or not. This was true with desktop publishing. It is true now with AI. The people who did well in earlier transitions were not the ones who waited for certainty. They were the ones who stayed close enough to adjust.
The paste-up artists who survived were the ones who learned PageMaker in its first awkward year, not the ones who waited until it was polished. The difference was not talent. It was timing and willingness.
Using AI now creates familiarity. It shows you where it helps and where it breaks. That understanding is what gives you options later. When the tool becomes standard—and it will—you want to already know what it does well and what it does poorly. You want to have judgment about when to use it and when to work around it.
The Long View
After forty years in this field, I am cautious about certainty. The tools always change. The work changes more slowly. Experience tends to matter more after the noise settles.
I do not know exactly what AI will do to graphic design. No one does. What I do know is that resisting change outright has never been a winning strategy. But neither is adopting technology blindly. The goal is not to use every new tool that appears. The goal is to understand what is actually changing in the work itself.
Desktop publishing did not just make paste-up faster. It changed who could publish and where publishing happened. AI is not just making design faster. It is changing which parts of the work require human judgment and which can be automated.
This feels like one of those moments again. Not because everything is about to disappear, but because the direction is clear even if the destination is not.
Engaging now creates options later. Waiting for certainty usually means arriving late.
That has been true every time the work has changed. I do not see a reason to believe this time is different.