Ten years ago learning to drive a truck meant a long-term stable career, even if it's not the most glamourous. Today, we're discussing the short-term viability of automating that shipping system. Five years ago in Alberta, getting into a trade that was in demand in the oil industry meant a lifetime career with a pretty great salary. Today, the industry is on shaky footing and many will have to retrain.

In software, while it might not be near or certain, there are plenty of discussions about automating most projects since so many of them conform to common patterns and publicly accessible, free code already existing. We don't know any better how large much of the software development industry will be in ten years time. Many of those skills might be as obsolete as some IG "celebrity".

I think OP is at least correct in that you sort of have to weigh these things yourself. Sometimes that life experience in other facets will inform your future more than picking up a skill that is just enough to put you into someone else's employ. It's pretty difficult to know. You can apply any kind of probability measure you want, but the gaps in between any sound reason are just attempting to know the future before it happens.