I started building websites in 2012, part-time, for small businesses around me. Since then I have been told, with great confidence, that my profession was about to end — four separate times. It's worth listing them, because the pattern is the argument.
Four funerals, no body
WordPress themes were going to end web design. Why hire a designer when a ₹3,000 theme exists? What actually happened: the floor rose. Millions of small businesses that could never afford custom work got decent websites, the bar for "custom" went up, and designers moved up the stack — from drawing pages to designing brands, systems and products. I know this one personally; WordPress sites were my own first practice, and they funded the career that outgrew them.
Bootstrap was going to end UI design. Every site would look the same! Briefly, every site did look the same — and clients learned to tell the difference, and started paying for the difference. Sameness created the market for distinctiveness.
Figma was going to devalue designers because now anyone could open the file and edit. Instead it pulled PMs and engineers into the design process, multiplied design's influence inside companies, and created more design jobs than it touched.
No-code was going to end front-end development. It ended a certain kind of brochure-site project, yes. Meanwhile the demand for real product engineering kept growing right past it.
Four predictions, one shape: a tool arrives that produces the median version of our work for nearly free; the panic assumes demand for the work is fixed; demand instead expands to absorb the new capacity, and the craft migrates upward to whatever the tool can't do.
Why AI fits the pattern (and where it doesn't)
AI is the strongest version of this event we've seen — I won't pretend otherwise. It produces the median version of much more of the work, much faster, and it improves monthly. If your entire value is producing median output, this is a genuine problem, and the honest response is to stop producing median output. That was always the trajectory of a craft anyway; AI just compressed the timeline.
But notice what AI does to demand. The cost of trying an idea is collapsing toward zero. When trying gets cheap, people try more — more products, more experiments, more weird personal tools, more software for markets too small to justify software before. Every one of those things needs judgment about what to build, for whom, and whether the result is any good. That judgment is the part of design that was never about production speed.
Tools lower the floor. They have never once lowered the ceiling. The ceiling is made of taste, and taste doesn't automate — it accumulates.
The genuinely new and good part
Here's where I go past "we'll survive" into actual optimism.
The unglamorous work finally gets done. Accessibility audits, empty states, error copy, localization checks — the work that always lost the prioritization fight now costs almost nothing to do well. The median user experience of software is about to improve dramatically, not because companies got virtuous but because excellence got cheap.
The solo builder is back. The web I fell in love with was made by individuals with strange ideas and a text editor. The industrialization of software pushed them to the margins. AI hands the individual a team: one person with taste can now ship what previously needed five specialists. Expect an explosion of small, odd, personal software — and personal websites like this one, built by hand, on principle, with help.
A billion people get a designer. Most humans will never hire one of us. They have community groups, tiny shops, school events — things that deserve good design and have never had access to any. The same force that worries my industry is, from their side of the screen, pure gain. It's hard for me to look at that honestly and conclude the world is worse.
Calm, not complacent
Optimism isn't a forecast that nothing will be lost — transitions have casualties, and pretending otherwise is marketing, not analysis. It's a judgment about the direction of the sum. Fourteen years of funerals for my profession have taught me to check the casket. It keeps turning out to be full of toil — and the craft keeps walking out the side door, carrying better tools.
I intend to keep walking with it.