Programming has been one of the best-paid white-collar jobs over the last three decades, but it could soon be just an oddball curiosity.
That is the striking claim from Peter Steinberger, creator of OpenClaw, who made the remarks in what reads as a candid, almost elegiac reflection on the future of software development. What makes the quote remarkable is not just its boldness, but its tone — Steinberger is not a detractor of the craft. He is someone who clearly loved it deeply, and that makes his verdict all the more sobering.
“The actual art of programming will stay there,” he says, “but it’s going to be like knitting — people do that because they like it, not because it makes any sense.”
The comparison to knitting is pointed. Knitting was once an essential economic activity; now it is a leisure pursuit, practised by enthusiasts who value the process rather than the output. Steinberger suggests programming is heading in the same direction, and he is not entirely unsentimental about it. “I read an article this morning about how it’s okay to mourn our craft, and part of me very strongly resonates with that,” he says. “In my past, I spent a lot of time just being really deep in the flow, cranking out code, finding really beautiful solutions — and it’s okay to mourn it.”
But mourning, he argues, is not the same as resisting. “It’s not something we can fight. The world, for a long time, had a lack of intelligence — of people building things — and that’s why salaries of software developers reached stupidly high amounts. And they will go away.”
The implications of that last line are significant. Steinberger is not describing a gradual softening of the job market. He is describing a structural unwinding — one already visible in the data. In 2024 and into 2025, major tech companies including Google, Microsoft, and Salesforce announced waves of engineering layoffs even as their revenues grew, with AI-assisted development cited as a key productivity multiplier. Platforms like OpenAI’s Codex, Cursor, and Anthropic’s Claude Code are now capable of producing production-ready code from natural language prompts, compressing what once took a developer days into a matter of minutes. Andreessen Horowitz and other venture firms have begun funding startups built almost entirely by AI agents, with human engineers serving more as supervisors than builders. If Steinberger is right, the transition will not be a crisis for technology — it will be a crisis for the people who built it, and who, for a brief few decades, were paid extraordinarily well for doing so.