But are our fears justified?
To find out I asked a friend who recently finished a PhD in robotics (and who preferred not to be named to keep a low profile at a new job) for his insider's take on the subject. He started by explaining how the technology is progressing and what our smart machines are -- and are not -- capable of:
It's important to note that what Watson did is sort of low-hanging fruit in the "AI" world. Yes, Watson is able to amass facts and make some connections between them, but it's still mostly raw information. It's important to note what Watson did NOT do: see or hear. Watson is a data machine -- a glorified computer. It doesn't really interact with the world.So what does that mean for work? Paul Krugman recently penned a much discussed op-ed arguing that the divide between what robots will and will not be able to accomplish soon ignores the human divide between white and blue collar jobs. Instead, Krugman claims routine is the hallmark of a soon to be extinct profession:
People find it hard to understand that what is so easy for us is so hard for computers -- seeing, hearing, spatial reasoning, language, etc. This is because the brain devotes the vast majority of energy to these tasks. Evolution has developed them over eons to be really fast, accurate, and unconscious. If they were not, we would not survive. By contrast, building this form of intelligence into machines is very hard. This is where the automation revolution is really going -- interaction with the real world.
Any routine task -- a category that includes many white-collar, nonmanual jobs -- is in the firing line. Conversely, jobs that can't be carried out by following explicit rules -- a category that includes many kinds of manual labor, from truck drivers to janitors -- will tend to grow even in the face of technological progress.When it came to dividing the doomed professions from the saved, does my friend agree with Krugman? While there was plenty of overlap in their opinions, my friend's description of which professions are likely to be taken over by robots in the foreseeable future took a technologist's-eye view, which focused on the nuts and bolts of building robots, rather than an economist's theoretical lens, with interesting results:
The jobs that will go first are the least connected to the physical world. Lawyers and accountants are good examples. Certainly, this will not put all lawyers and accountants out of business any time soon. But there will be decreased demand for their services as computers do more and more of the easy work, and eventually more of the hard work.I suppose we can all breathe easier knowing our lives will not soon resemble an episode of Battlestar Galactica, but on the other hand, my friend's final verdict on the changes soon to be wrought by robotics wasn't exactly soothing either. "It'll be huge," he said.
We are fast approaching, however, the time when machines can interact reasonably well in the real world. They will only get better and more efficient. This will be true for tasks like: driving, flying, manufacturing, mining, stocking, shipping, construction. Jobs that will be "safe" will fall into creative or human service categories: writers, engineers, artists, waiters, actors, nannies, teachers, etc.
Krugman agrees with that assessment at least and is pushing stronger labor unions and healthcare reform as bulwarks for the middle class to defend its prosperity against the rise of robots. What adjustments do you think society needs to make to ensure wealth is shared broadly in an age of increasing automation?