Tech careers continue to be remarkably short: 10 years or less if I remember correctly, which is crazy given how much fun programming inherently is. But people aren't stupid and often see the writing on the wall: if you want the big bucks, outside of either being a founder or a Very Special Person at a few Very Special Companies, you have to start managing and playing the political game.

I blame this for the inherent cyclical nature of technology: young kids with no experience come in, get whipped into a frenzy over some technology we've been over before (see Node.js/cooperative multitasking) and then slowly learn the shortcomings of that approach, just in time for the next snake oil to hit.

The irony is that I think that as technology continues to knock down the amount of developers necessary for a given project, older developers who know how to get from point A to point B with the minimal amount of effort will become more valuable in relative terms, so long as they can avoid getting ground down and stay productive. Couple that with another controversial observation of mine: that theoretical technological progress has slowed dramatically in the last decade (implementations are still catching up, but are getting there), and I think we will see small, elite teams of experienced developers become a preferred (or at least common) way to build software.

That's self serving, of course (I'm 35) but I do think it is true.