That's where the second half of my comment comes in. I am positing that if LLMs are able to increase the output of developers, the demand for software will increase along with it. The economic term for this is induced demand, and the classic example is roads, when you build more of them the number of cars on the road increases because there were people who used to stay home or take other modes of transportation, and now they drive because there are more and less congested roads. It's not what most regular people believe but it's also not my idea, it's the basis of a ton of tech investor successes.

I think we are in that position with software. Increase the supply of programmer output, virtual or not, and the demand for it goes up, because there's a lot of demand on the sidelines today (like maybe programmer output is too expensive for them now and LLM tooling makes each unit of output cheaper, which in turn makes new customers appear).

Again Marc Andreessen to name just one guy who wrote a famous essay about it has made billions of dollars in large part by betting on this phenomenon which most people don't understand or believe in.

If you look at things from the perspective of the job that you have today, depend upon and which might go away, yeah a job is a fragile thing so it's easy to get scared. But if you look at it from the perspective of there being these guys who keep betting on the idea that software demand will increase and then reality has proven them right and made them rich, this idea of software jobs disappearing is less scary.