Last week, I asked, Should We Work? and discussed the issues of job elimination over the next 50 years. Here’s a terrific article from Technology Review about how the process works. And, here are my fearless predictions of just some of the jobs that will disappear over the next 50 years. I’ve also included my predictions for the types of jobs that we will still need. It’s not a long list.
Drivers – autonomous vehicles will replace cab drivers, truck drivers, and bus drivers. Tell the vehicle where you want it go and it will take you there.
Surgeons – if a robot can make micro-welds in the factory, it can also replace your heart valve. A robot’s hands don’t shake.
Computer programmers – computers can program computers faster than people can.
Doctors – computers can diagnose what’s wrong with you better than humans can. Humans might have a better bedside manner, but avatars accompanied by a cute puppy are close… and they’re always on call.
Pilots – drones can do a better job, especially in fighter jets. The weak link in military aircraft is the human. Without a human, we can build faster, more maneuverable, and much, much cheaper warplanes.
College professors – MOOCs will rule.
Room service staff – why order in when you can order out?
Accountants – accountants interpret rules and enter data. Computers can do that.
Stockbrokers – computers already account for the bulk of stock trading. In the future, you won’t invest in stocks; you’ll invest in the algorithms that you think can pick the best stocks.
Engineers – most engineers solve structured problems. So do computers.
Politicians – computers can find optimal solutions to problems better than perpetually outraged people.
Many of our white-collar jobs today require people to manipulate symbols and process information. For instance a doctor who is trying to diagnose what ails you needs to interpret lab results, recall symptoms of many possible diseases, fight off fatigue, and evade logic traps. Well-trained computers can do this better.
So what kinds of jobs will be left? I can think of two general categories:
Persuasion – I’m not sure that we can train computers to be persuasive. Being persuasive requires an emotional connection and a degree of trust. Can you trust a computer? Perhaps. Still, I think people will be more persuasive, though maybe only marginally so.
Imagination – Can we teach computers how to imagine? Perhaps. After all, innovation typically results from mashing up existing ideas. A computer could mash up ideas. But it would be fairly random; I don’t think a computer would really understand the possibilities. So, humans should retain a competitive edge in tasks that require imagination.
I’m guessing that the ability to manipulate symbols and process information won’t be enough to get you a job in 2063. You’ll need to be imaginative and persuasive. Is that what we’re teaching in schools today?