> flip-flop every 6 months between thinking that everybody can do this and being absolutely floored that I myself can keep up with this.
I agree that it takes a certain way of thinking to be able to reason about strict logic and symbolic relationships. Just ask someone what (X or Y) equals when X and Y are both true. I knew people in my CS classes right up until the end that couldn't tell you correctly (because people in regular English think of "or" as "exclusive or" and its hard for some people to change their understanding).
I disagree with the concept about the rapidly changing tech jargon. I think in SF tech startup scene it may be that crazy, but once I start a job, I use their stack, and it stays that way for years. Different companies different stacks though. Big-Tech is even easier since they have their own custom stacks that are pretty stable.
> I'm not a doctor/surgeon, but I don't think that the kind of work they do changes year to year at this pace?
On the contrary. It changes all the time. They always have to be learning about new medication, new surgical techniques, new diseases. Ask a doctor in 2019 to diagnose the Sars-Cov-2019 Novel Coronavirus. Then ask them about ivermectin.
> I guess you could say that none of these concepts are really that difficult conceptually, but...what is...? Algebraic Topology? Chaos Theory?
I would say that real theory is actually pretty hard in CS. Math is famously unapproachable to people who don't "get it" and I think CS theory shares similar logical rigor - honestly similar to law and how it is a "different" logic than most people are ready for.
I know rocket scientists, and their classes were actually-intense math - but conceptually/logically easy IFF you could do the math. The basic biology of how a viral infection works is pretty straightforward (and taught to 13yo's), but again, thats not the hard part of the job of a doctor.
But most CRUD apps are not. And most people write CRUD apps. At my BigTech jobs, the company always had custom-in-house build tools that were great at doing deterministic reproducible builds and they were always based on strong theory and rigor and had white papers. But how many people actually have to do that at those companies?