Programmers develop new technology and paradigms to solve problems caused by existing technology and paradigms. As those new technologies are adopted, they are perceived to give an edge. And so, this creates a demand for experience in this new technology.
As time moves on, the impact of the new paradigm or new tech is getting better understood and the initial optimistic bias aligns with reality. Meanwhile, programmers develop new solutions to solve problems caused by... and the cycle repeats.
The problem here is that the problems that are being solved aren't necessarily new. It's just this foundational "how do you build a website or web interface efficiently" problem that underpins everything.
In a way, it's not much different from re-inventing the wheel. There are only so many specific ways of doing that before the entire exercise turns silly.
The bigger issue here is that - unlike tire manufacturing - the subsequent cycles of innovation are too short for a majority part of the workforce to keep up with changes. Why? First, because the complexity of those solutions creates a prohibitive cost for their employers to easily migrate on short notice and keep up with the cycle. Second, because the vast majority of new tech are just small optimizations in the real world with a limited impact on business making migration a risk and not an opportunity.
Hence why, once a tech has been chosen, companies stick with it for the long haul at the expense of individual developer experience keeping up with the innovation cycle.
That's what turns this into a self-fulfilling prophecy. The faster the cycle goes, the more tight the labor market becomes. That's why you'd see those crazy "Must have 15 years of React experience" requirements.
Finally, what we see is developers are being goaded into whimsical HR hiring processes that shake them down. In reality, it's not developers being "unemployable", it's the hiring process that's essentially broken as it is driven by individual business needs that aren't really that sustainable in the long haul.
When businesses lament that they can't find good talent, but at the same time reject the vast majority of candidates because they can't possibly be proficient in the latest cycle of hype technology (unless you're a core contributor to a particular framework or live a monastic life foregoing any and all other aspects that make life worthwhile living, that is), well, that's a form cognitive dissonance I would say.
As a developer, I feel it's important to keep an eye out of how things evolve, but there's little to no point in trying to keep up with the nitty-gritty at all costs without knowing if you're ever going to need it.