Robots Won’t Take Your Job

I’ve frequently scoffed at the idea that in the foreseeable future many jobs presently performed by human beings will be done by robots but you don’t need to take my word for it. At MIT Technology Review artificial intelligence researcher Vincent Conitzer tries to cast more light than heat on the subject:

Overall, when we try to have AI do existing jobs, we often find it failing in ways a person never would. The history of AI research is littered with examples where researchers create systems that perform surprisingly well at a well-defined task, only to find that it is still hard to replace the people who perform similar tasks in the messier real world.

Perhaps the more typical case will be that jobs are partially eliminated because part of the job can be performed by AI. Technological advances may also further facilitate outsourcing jobs to people around the world. At the same time, many jobs will remain immune, at least for the foreseeable future, because they fundamentally require skills that are hard to replicate in AI. Consider, for example, therapists, coaches, or kindergarten teachers: these jobs require a general understanding of the world, including human psychology and social reasoning, ability to deal with unusual circumstances, and so on. AI may even bring some people back into the workforce. For example, progress in robotics could make it easier for people with disabilities to hold some jobs, and progress in language processing may do the same for people who have difficulty using current computer interfaces.

Here’s his peroration:

The idea that recent progress in AI will prevent most people from meaningfully contributing to society is nonsense. We may have to make some changes in the way society works, including making it easier for displaced workers to retrain, and perhaps at times increasing public spending on (say) carefully selected infrastructure projects to counterbalance job losses in the private sector. We should also be mindful that advances in AI may come unexpectedly, and do our best to prepare and make society resilient to such shocks. But the idea that we are about to enter a techno-utopia with almost no need for human labor is not supported by the current state of AI research.

My experience has been that the less technologically knowledgeable one is, the more likely one is to believe that robots are about to replace human workers. The reason for slow job growth isn’t robots. It’s other human beings.

8 comments… add one
  • michael reynolds Link

    Tell you what. I’ll make you a gentleman’s wager to be paid out on November 2, 2026. Ten years. The bet is this: that the consensus of economists will concede that machines are rendering more humans permanently unemployable. A bottle of good single malt? (Assuming we both last another decade.)

  • michael reynolds Link

    Incidentally, he does not understand how creativity works if he’s so easily dismissing Google Deep Dream. I gave a speech in Edinburgh on this and used Deep Dream as a pretty fair example of exactly how creativity works. If you don’t believe GDD is creating, take a long look at some of the examples. You’ll find that in addition to distorting the initial photo, GDD creates new ‘characters.’ I don’t know what mystical notion of creativity this STEMie has, but that is creativity and the result is art.

    GDD can start with a picture of a pizza and after a few iterations there’s suddenly a dog standing next to the pizza. In other words GDD is riffing, using its database, and the object at hand – the pizza picture – and in the process creating an image that did not previously exist. Not just a distortion, but a whole new object.

    What do people imagine goes on inside a writer’s head when he’s writing? That is exactly why I get paid to make stuff up: because of that object which did not previously exist but which can be clearly shown to have been ‘inspired’ by my data set.

    We don’t want to see it as creativity let alone art because we have romantic notions around inspiration and motivation and the rest. But in my experience every ‘inspiration’ story, from writers at least, is bullshit. (I’ve peddled some of that bullshit because the public likes it, but it’s nonsense.) We have our data set, our education writ large – just like GDD. We have a view of reality, the pizza picture. Using our data set and the object at hand we riff and we play out story possibilities. We extrapolate. We invent dogs that weren’t there to begin with.

    GDD has already created some powerful visual art, art which has a mind-expanding, disturbing effect. And it is taking baby steps toward being able to replace me. I doubt it will take a decade for the first publishable, computer-written book.

  • michael reynolds Link

    Oops. Code error. Sorry.

  • I don’t see how you could lose that bet. All it would take is one person losing his or her job. That’s practically a done deal.

    Let me propose an alternative wager. As of November 2, 2026 if the U. S. unemployment rate as published by the Bureau of Labor statistics is over 10%, you win. Otherwise, you lose.

    The issue isn’t whether any jobs will be automated out of existence. They probably will but that’s irrelevant. The question is whether it will produce persistent mass unemployment. I’m giving you the benefit of the doubt by setting the unemployment rate in the terms of the wager so low. 30% or 50% would be a better gauge of mass unemployment.

  • Guarneri Link

    So there you have it. Technology morphing pizza into pizza eating dogs. And greybeards stroking their chins calling it all art. Perfect. Clearly we are doomed.

    We have been automating for as long as one can imagine. Ever hear of that job killer, the lever?

  • steve Link

    Limiting it to “robots” is silly. Jobs are clearly being automated away by technology. The assumption has always existed that when there is creative destruction, new jobs will come along to replace those lost. Maybe that doesn’t hold true forever. Maybe there won’t be as many new jobs and/or the new jobs won’t pay nearly as well.

    Steve

  • Guarneri Link

    Dateline November 3, 2030

    Yesterday the Rockford Robots expansion team won the World Series for the 10th year in a row. Once again pitching a perfect game was Mike “The Flamethrowing Machine” Mechanical. The Robots score tally was unknown as every pitch made to them was a home run……….. The Robots have already been installed as 1,000,000 – 1 favorites for next year’s Series.

  • michael reynolds Link

    160 books, 13 different series, humor, sci fi, fantasy, horror, romance, alternate history and I’m moving into adult mystery. Six million words, give or take. Yeah, you’re right Drew, what do I know about creativity?

Leave a Comment