The trick with giving people new skills is a) the skills needed tend to be super specialized - it's hard to retrain a laid off line worker to be a genetic scientist; and b) we are rapidly approaching the time when computers/robots can be trained to do work faster than people. Even if you can only automate 50% of a job, that means half the people already doing that job will be out of work. The growth of new jobs will come in areas of rapid expansion into the unknown -- entrepreneurs and inventors. Once something is understood well enough to be replicated, it will be automated. Other positions that will still need people are ones where there is an underlying level of chaos, so that what worked yesterday won't work tomorrow. Or ones where people want to deal with people, not machines.