Fixing stuff

I've been working in IT for the last 20 years and to my contemporaries and I it sometimes seems that problem solving skills are not what they were. I've been wondering why this might be, when it is alleged that modern educational methods emphasise teaching children to reason and think, rather than learn by rote. I got a clue when I was helping my son set up the Scalextric , and as usual the track didn't quite make a good electrical contact, and the cars were full of fluff so they didn't run so well. I remember being really interested in how things worked, and my friends and I would disassemble all sorts of things, Scalextric cars included, in the search for knowledge. Some of them we even managed to re-assemble, occasionally so that they worked. Most of the technology we used was relatively simple and understandable - lift the lid on a 60's car and you'd see things that a village blacksmith could fix. An SU carburettor was basic, an Amal even more so, ignition systems were mechanical and very easy to comprehend. Even looking at a radio, you could imagine the electrons flowing from anode to cathode. There were also plenty of books that described these technological wonders and their workings. I think that this was useful in shaping our world view and teaching basic problem solving and diagnostic skills. Our mental abilities were further honed by having to cope with imperial measures and pre-decimal currency, without the benefit of electronic calculators!
Today, the technology we use is vastly superior, but more importantly, far less transparent. As Arthur C. Clarke said, 'Any sufficiently advanced technology is indistinguishable from magic', and my suggestion is that to the vast majority of people the things they use every day might as well be magic. This gives rise to a new irrationalism which hinders rather than encourages the development of engineering skills. Despite the huge increase in the number of people going on to higher education, the number studying mathematics and physical sciences is actually declining. Those of us who must work in fields that require these skills find ourselves increasingly isolated. Perhaps it is a race to perfect Artificial Intelligence before the skills we need to do it die out.