Development Is Dead

2026-09-22

I started in 2011, freelance until 2014, and I never left. Fifteen years. Long enough to have opinions, short enough that whoever started in 1995 gets to laugh at mine.

I remember Dreamweaver, back when nobody was embarrassed about it yet. I remember Eclipse eating a laptop alive just to open a project. I remember Sublime Text landing like someone had finally understood that a text editor is a text editor. I remember the pattern libraries we all copied from and nobody credited. I remember CodePen, which Chris Coyier co-founded in 2012 with Alex Vazquez and Tim Sabat, and CSS-Tricks, which he ran from 2007 until DigitalOcean bought it in 2022. And I remember him on stage, around 2016, walking through his own workflow and the tools he actually used, beer in hand, the whole thing closer to a guy explaining something to you in a bar than to a keynote performed at you. That was normal then. The web lost it and never got it back.

That was the culture. Small, loud, generous, and built on the assumption that we would keep doing roughly this for a long time.

The Italian footnote

Italy was never part of it. We had one real moment: the Olivetti Programma 101, New York, 1965, arguably the first desktop computer, more than 40,000 units sold, ten of them bought by NASA and used to plan the Apollo 11 landing. Then nothing on that scale. Ever again.

There are exceptions and they are real. Salvatore Sanfilippo released Redis on 26 February 2009, written in Sicily to fix a problem in his own analytics product, and it ended up as infrastructure under a serious chunk of the internet. But one guy is not a scene. The average here has been mediocre for twenty years and it has been kept above water by a handful of people who cared more than anybody was paying them to care. The famous one percent. Anyone who has actually worked in this country knows that number is not a figure of speech.

I dealt with it the only way that works: I ignored the local conversation and followed the people who were actually building, one per domain, directly. First on Twitter, then on whatever Twitter turned into, now on Bluesky for the ones who still want a timeline that belongs to them.

This time is not like the other times

And then this year happened.

We are not doing the job we did for the last twenty years. I do not mean the tools changed again. The tools change every three years, every single time somebody declares the profession dead, and every single time they are wrong. I have sat through four rounds of that.

This is not that round. The work moved. What I do now is specify, constrain, review, decide. The typing, the part that was the job, is the cheapest part of the job. And the rate of change is not linear, it is not flattening, and nobody I respect thinks it is about to.

The split is not between who uses the new tools and who does not. It is between who kept the fundamentals and who never had them. If you actually know what a system does, you can move up and down the altitudes: read the generated code, judge the architecture, smell the wrong abstraction, decide what does not get built at all. If you never had that, you are stuck at one altitude saying "yes, and make no mistakes" to a machine you cannot audit. That is not a career. That is a handoff.

So when somebody asks me where I see this profession in ten years, I skip the comforting answer about new roles emerging and the work merely shifting.

I do not see it. Something else will be there, and it will not be called development.

The one percent, again

There is an idea I keep circling back to, and I want to be exact about what it is and what it is not.

The Transcendental Meditation people have a claim they call the Maharishi Effect: a group the size of the square root of one percent of a population is enough to measurably move that population. The research goes back to the late 1970s. It is also rejected as pseudoscience, the studies have real problems with controls and with correlation dressed up as causation, and I am not asking anybody to believe it.

I am stealing the shape, not the evidence. The shape is this: the fraction of a population that has to move before everything else reorients around it is far smaller than people assume. That part I have watched happen, with zero meditation involved, in every shift I have lived through. A handful of people went somewhere, and two years later the industry was arguing about whether it had ever been anywhere else.

Which is the uncomfortable part. The square root of one percent of this field is already gone. They are not waiting for the rest of us to finish debating whether any of this is real.

So where do we think we are going?