Many people in the field now aren't nerds anymore either but people who are just incentivised by financial prospects.
I don't think I'd be currently doing software if YouTube and games were accessible and good on phones back then.
Sure the kids get "exposure" to Scratch but they're not really teaching it. It's not like the teachers are programmers. They just let the kids make little scrapbooks. It took me to show my kids triggers, messages, and loops.
And there are going to be lots of new companies that suddenly want to build a load of software.
The profession should hang on for a good while at least.
The same way indie developers have always outmaneuvered larger companies. A number of my favorite apps are built by a single developer (Overcast, Callsheet, Carrot, Widgetsmith, Octal) or small teams focused on building a good product. Caring. That might seem simplistic and it's not to say no one working at a larger company cares, that's clearly not true, but there is an attention to detail and craftsmanship, not to mention soul, I find lacking in most software produced by larger companies.
I have a product that competes (or rather they compete with me? I was "first to market" for all that matters) with a product produced by a company (no clue the size, but I'm 1 person, they seem to employ multiple people). I think my product is better because I "Sweat the small stuff" and spend more time working on improving the experience than they do.
Perhaps I'm delusional. Maybe people think their product is better but the people who have switched don't seem to think so. I don't solicit information about them from their previous clients but I've heard enough "This is so much easier/clearer" or "That's it? The old tool required XYZ!" to believe it.
And we have countless examples of that working. Small company invents a neat thing, big company sees that, allocates 10x the resources on it, and small company is lost.
But we also see that that titanic companies are difficult to maneuver, and a little speedboat team can end-around and run for the goal line. Think Google back then.
I think all these things coexist.
Movie theater numbers are collapsing, in large parts due to the number of entertainment available at home those days.
I do love straining an analogy. :)
I'll grant you that your future is possible. If "searching" via LLM for a tool just offers to build it for you, I'm not sure what that does to the software market. But I have a feeling that even with more people "building" instead of "buying" there will always exist a majority of people who don't want to bother with "building". Heck, there are people in this industry that I would never call a "builder" and there are plenty of people across all industries that simply don't have the drive/interest/etc in building, even when it's easier. I bet if you grabbed someone from the dawn of computing and plopped them down just pre-LLM they would be amazed and flabbergasted "Why isn't everyone writing software custom for their needs?? It's SO much easier than it was when you had to learn punchcards! The web (how very interesting by the way!) offers free courses to learn programming that you can do on the device you carry in your pocket!".
It seems that no matter how easy we make programming, there will always be people willing to pay others to "manage" that process. See also: YouTube has videos for anything you might need to do home-maintenance-wise, and yet the trade jobs are in incredibly high demand [0]. Perhaps with LLM + cheap robotics + unlimited/cheap energy this changes but we are still 2 parts away from that and even today's LLMs are just starting to show fruit w.r.t. controlling robots.
[0] PS: If you are leaving tech, go into this. Respond to people's calls, even if you can't get there _today_, respond and put them on a schedule. Use your tech skills to help automate/manage this. I've considered this every time I've needed just about anything done and gone through the miserable process of getting ghosted by multiple people before finally getting someone out.
Bro, good engineering IS boring and it's supposed to be.
The whole reason that Software Engineering is not actually engineering is because the field is too young for there to be established, rigorous engineering codes that must be adhered to. Can you imagine if every bridge you went over or every elevator you rode in had been built with arbitrarily chosen technologies by uncertified "engineers" using the latest fad in engineering?
I think one of the outcomes of this great shift where most programming output is being produced by LLMs is that we're going to start seeing much more consistency across software implementations. It will be boring. And honestly that is probably going to be a good thing.
There are a number of open source tools that have the same, on paper, features as a paid product. And yet the paid product still exists. Is it because Pixelmator Pro produces a better "blur" than GIMP? No, it's because Pixelmator Pro is more enjoyable to use, has a better user interface.
So yes, your auth/RPC/whatever should be incredibly boring, and now that LLMs can handle that aspect reliably we can focus (if we choose) on improving other aspects of the software and that will be the differentiating aspect.
And that's literally one single profession.