In the beginning there was print design, and when asked what my future goals were, they were that.
In the job market however, there were enough print designers. What drew them to me was my knowledge of the web, so I focused on that. I hacked up Photoshop files and hand-coded HTML. I taught myself this new thing called CSS, and learned all about browser versions.
Then came the CMS, then came the WordPress. Users didn’t need you to update a static file with their content, they could just enter it directly into a text box. Control was lost. We were no longer creating pages, but systems, and the systems had to be able to handle the chaos of inexperienced users.
Then came the templates, then came the Squarespace. Users didn’t need you to setup a system, they found one. They could drag and drop. They got stuck on the domain part, or maybe the email part, but after that they were in control.
At this point I had pivoted more towards application development, but still had the occasional website setup. It’s important to remember that even though the bottom of the market drops out, there’s still clients that don’t want to bother with doing somebody else’s job.
Now comes the AI, now comes the Claude Code. Users want to build the application themselves, they want to contribute the code directly. The question is, is the codebase resilient enough to handle it? Can we expect the agent to always understand the full context of the problem? I’m guessing they’ll still get stuck on the deployment part, or the infrastructure part, but maybe that will eventually change too.
Gradually the industry keeps shifting me out of one role into something new. Now I’m wondering, where else is there left to go? The computer is becoming a means to an end, with less and less human guidance required. The end state isn’t today, but it looks pretty inevitable. I can likely be sustained by those clients who aren’t interested in doing somebody else’s job, but how many of them remain is unknown.