Examples would be appreciated. But this seems to be a case of trying to time the market and the usual objection applies; if you can time the market to within a year you can make huge piles of money. One of the contributors on HN, lsc of prgrmr.com talks about how he was calling the property bubble in the Bay area for years before it popped, and how if he had just got in at the frothy height of the dotcom bubble like everyone else, he'd still be ahead now on property, very far ahead.

In such case, I guess you are more likely to be correct about this than me. Only the "cheap to learn" part feels wrong to me. I mean, the financial costs of learning programming are already literally zero in the recent years, and somehow still most people don't learn one of the highest paying professions. Why? If they didn't do it during the recent five years, why should they do it during the next twenty? Maybe an ability is the problem, not the financial costs of learning. I suspect that those other industries either employed less people than IT, or that they were easier to learn. On the other hand, IT has its own specific risk -- a possibility to work remotely, so it is easier to outsource.