The way I think you do an end run around that is to not think in terms of tools but in terms of skills. If your skill today is building CRUD forms, that's what you'll do in 30 years, only you'll do it in that decade's tools. I think you're then supposed to look at the supply and demand of skills, and try to gain a skillset that will be in high demand but with a limited supply. For example, building CRUD forms has probably reached peaked demand, and the influx of programmers in low wage markets is going to put downward pressure on wages for decades to come, so no matter how well you keep up with tools, you don't want to be in that market for the next few decades. On the other hand, if you specialise in data science and assorted algorithms, you're probably set for life, because demand for those skills will keep rising as software gets smarter, and the difficulty of gaining those skills will limit supply.