The tools of today, even fucking Eclipse and gah Xcode from the early 2000s make programmers far far more than 10x more productive than machine code on punch cards. By your implied logic there should be a tiny fraction of the programmers there were in the 60s.

Anyway I think this 10x more productive claim is still a bunch of BS with the current crop of AI demos.

Why hire new people, because companies want to do things they couldn't do before. This has been happening for decades already as tooling has evolved. It used to be very hard for average companies to write cross-platform commerce software, now it's a standard thing in the form of a website + payment processor.

There might be some companies or departments that don't need to evolve anymore, and they cut people. That happens already; the product gets finished, so they lay people off. But those people find new jobs.

My guess is that the goals for existing software projects have always been limited by how much talent could be purchased, not by vision. So unless we run out of vision, nobody is actually going to be put out of a job by this technology. We'll just shoot for bigger goals.

This will create more demand for our field, not less. Increasing the power of programmers unlocks so much more potential.

Companies always have ideas on the backlog. I myself have a huge backlog of dream projects that I don't have time to implement, and I'm just one person.

Tech like this helps us make a bigger impact faster.

The new people will still work for a little less.