AI Democratized Software Engineering
Software engineering has been democratized — rapidly, and perhaps irreversibly. Anyone can now generate working software from a description. One consequence is already visible: fewer engineering jobs than in previous years, and the shift arrived suddenly rather than gradually. When everyone can build, building stops being scarce.
Whenever a skill gets cheap, I think it’s worth asking what happens to the economy built around it. Consider an analogy from agriculture: China periodically produces such an oversupply of certain crops that farmers dump or burn their surplus, because selling it would crash prices and ruin them. Abundance, arriving too fast into a scarcity-designed system, destroys value instead of creating it. Is our economic system, built on scarcity assumptions, simply obsolete?
Push the thought further. If AI keeps driving costs toward zero across more domains, we approach a state of genuine abundance — where money matters less and less. Would that world need something like universal basic income? And what would it mean to be someone who spends their life hoarding money in a world where hoarding no longer confers security?
I ask these questions knowing most people flinch at them. Many of my friends react to AI-generated art and code like modern Luddites — with rage at anyone who prefers machine assistance to human craft. Bo Burnham said “art is dead,” and I love his songs precisely because they’re full of contradictions, but I can’t join the rage. Why are we angry at something that helps people create? The opposition often strikes me as insecurity wearing the costume of principle: skeptics who feel liberated when they’re proven right and stay silent when they’re wrong.
The deeper bias underneath, I think, is scarcity thinking. We are hypnotized by a corrupt economic system into seeing everything as limited — attention, work, meaning itself. But the little earth we have is extravagantly abundant; even this small dot of a planet overflows. Why bitterness toward anything that aims to make people more capable? If technology could let every disabled person see, hear, and create as fully as anyone else, would that make the world better or worse? It seems obvious to me.
There’s also a philosophical current beneath all this, one I keep circling. Everything decays — entropy only increases, and every tradition that contemplates mortality agrees that everything alive is heading toward its end. There is time in light, time in sound, time in attention itself. If you take time seriously as the architect of everything, then obsolescence isn’t a tragedy unique to software engineers; it’s the universal condition. Fighting the obsolescence of your particular skill is fighting the tide.
So I land here: the democratization of software engineering isn’t the death of the profession, but the death of its scarcity premium. What survives — judgment about what to build, taste, responsibility for consequences — was always the hard part anyway. The tools just removed the excuse for confusing typing with thinking.
Drafted with AI assistance from my personal journal (2025-05-11), then edited by me. See On AI Assistance.