Are robots outsmarting us?

BY Erik Brynjolfsson and Andrew McAfee  February 13, 2014 at 2:22 PM EDT
Robots are leading the second machine age, automating not just manual labor but cognitive function. Photo by Flickr user Gregory Lee.

Robots are leading the second machine age, automating not just manual labor but cognitive function. Photo by Flickr user Gregory Lee.

Editor’s Note: The robots are coming? The robots are coming!

Actually, it turns out they’re finally here: humiliating us at Jeopardy, driving us from the Redwood Forest to the Gulfstream waters, running our lives online as software “bots.” Famous as a someday threat to factory jobs, robots (from the Czech “robota” — drudgery or serf labor) are about to give almost all of us human workers a run for our money. As a result, we live in the “The Second Machine Age,” according to authors Erik Brynjolfsson and Andrew McAfee.

As far back as 1893, Ambrose Bierce, not exactly known as a science fiction writer, imagined a mechanical man so frustrated, it beat its inventor to death. And machines replacing humans is a theme we’ve long explored on Making Sense, as in our 2011 report “Man vs. Machine.” The fear, of course, is that technological advancement, of the kind we’ve featured from Singularity University, will at last decisively benefit the highly skilled few at the expense of the workaday rest.

Digital advancements, like iPhones and Twitter, are only creating a handful of well-paying jobs, techno-pessimist Tyler Cowen told us in 2011. His take: an innovation drought is now stymying economic growth. He argued, in a tour of his kitchen and major appliances, that no contemporary invention has significantly changed our lives as much as the refrigerator or gas stove of many decades ago.

But Brynjolfsson and McAfee, director and associate director, respectively, of the MIT Center for Digital Business, know of no drought; instead, they see the increasing automation of cognitive tasks as the second great era of technological change.

In this extended conversation, they explain how this latest industrial revolution, with potentially negative repercussions for the labor market that they fully acknowledge, will have any even bigger effect than the first.

In a second post, Brynjolfsson and McAfee explain what kind of educational and infrastructural investment humans need to make in order to stay a step ahead of robots. Their tour of Boston’s robots with Paul Solman aired on the NewsHour Thursday. Watch that segment below.

Simone Pathe, Making Sense Editor



Andrew McAfee: Are we really in some new era? Well throughout all of human history, if you wanted to move a vehicle from point A to point B, you needed to involve a person in that work. Not anymore. If you wanted a decent understanding of something written in another language, you had to involve a person in that work. Not anymore. If you wanted to listen to what a person was saying, figure out what they wanted, get an answer back to them using human speech, you had to involve a human being in that work. Not anymore.

MORE FROM MAKING SENSE:

Should We Fear the End of Work?

So we’re seeing these longstanding challenges, capabilities, opportunities where we have always needed human beings, and suddenly, we have digital alternatives to all of them that are already pretty good, and getting better very, very quickly.

Paul Solman: If you think of a timeline and technological progress, there have been other points at which the rate of change has changed, has increased. You’re saying that that’s happening again, now?

Erik Brynjolfsson: We are at an inflection point. The first big inflection point in human history was about 200 years ago, when the steam engine started the industrial revolution. That was a period that saw a whole set of new machines come along that could automate muscle power, physical work.

In recent years, we are seeing a wave of technologies that can augment, automate all sorts of cognitive tasks, and we think, ultimately, those will have as big, or an even bigger effect on humanity as the first industrial revolution.

Andrew McAfee: If you look at a graph of human history, it’s actually incredibly boring because you can go back thousands of years and graph things like how many people there were on the planet or what the state of advancement of civilization is. We’ve got decent measures for that…

Paul Solman: And by boring, you mean?

Andrew McAfee: By boring I mean, if you draw the graph, nothing at all happens for thousands of years. It stays very close to zero, there are a couple little bumps, but really, it is a flat horizontal line, almost close to zero, until at one point in time, the graph goes like this [points up], and that point in time was the industrial revolution, and we overcame the limitations of our muscle power.

Paul Solman: And that’s when longevity shoots up, productivity shoots up, after staying flat basically forever?

Andrew McAfee: Staying flat forever. So human life spans went up, wages paid to the average worker went up, population exploded, the air over London actually started to get cleaner after a little while instead of more dirty, so it was a huge leap forward in human progress, by far the biggest.

Andrew Brynjolfsson: In terms of economic progress, GDP grew by less than two-tenths of one percent per person for thousands of years, but starting at the industrial revolution, it grew more than 10 times faster, about 2 percent per year.

Paul Solman: But there’s the argument made by economists Tyler Cowen, Robert Gordon, Jim Grant on Wall Street, that this just doesn’t compare with, say, air conditioning, or something that really makes a difference in our daily lives.

Andrew McAfee: Yeah, so, Bob Gordon is a wonderful economist, and we’ve talked with him a lot. He doesn’t think the computer revolution measures up to some of the great innovations of the past, like the internal combustion engine, indoor plumbing or electrification, you know. He’s looking at a century’s worth of progress, and we think he’s not paying enough due to the century of progress to come, because of digital progress.

You mentioned Tyler Cowen. It’s really interesting; we disagree with Tyler a bit about how much has been going on in the past decades. He’s on the same page as us about what’s going to happen in the future, again thanks to this digital revolution.

A lot of the technologists that we talk to are so enthusiastic about what they’re creating that they think that it’s going to solve basically every problem, and that any kind of worrying about the future is really misplaced. They just have this great faith in technological progress, for a lot of good reasons. A lot of the economists and policy makers that we talk to are really concerned about stagnant standards of living, the hollowing out of the middle class, and they look to the future as kind of a grimmer and unhappy place.

What Erik and I came to realize is that really, these are two sides of the same coin, and the coin is essentially that technology is racing ahead, that’s the good news, but as it’s doing so, it’s leaving a lot of people behind, and that’s the challenge.

Paul Solman: Yes, this is terrifying to lots of people because if computers can think as well as lift, what are the rest of us going to do? And I’ve been reporting on that for a very long time, but it seems like it’s getting to be closer and closer to a dangerous reality.

Erik Brynjolfsson: There’s an economic fact that not a lot of people feel comfortable recognizing, which is that technology can grow the pie, but it’s possible for some people, even a majority of people, to be made worse off. Possible, but not necessary, and it’s going to depend on the choices we make in terms of our skills and upgrading our organizations, our institutions, to see if we can have them keep up with the technology as it races ahead.

Paul Solman:So what do you do as individuals, people who are understandably and, from your point of view, reasonably scared about the future?

Erik Brynjolfsson: Technology’s always been destroying jobs, but it’s always been creating jobs, and one of the ways in the past we’ve had technology create more jobs is through a constant upgrading of the scales of the workforce. One of the best ideas that America had was wide-spread public education, and that put us in a race between technology and education, where we did pretty well for over a century…

Andrew McAfee: Let’s not forget that we’ve been here before. As the factories in America were electrifying about a century ago, they suddenly didn’t need workers who could just pick up heavy things and move them around, they needed workers with very different skills, people who could think and be literate and numerate and work in teams.

The workforce was not trained with those skills though, so among other things, we started universal primary education, which someone has called the best idea America ever had, and that was in part in response to these technology-brought changes to the workforce. It succeeded brilliantly.

And as we’re rethinking education, let’s keep in mind there’s still a lot of things that computers and artificial intelligence and robots are still lousy at compared to us. They’re very ineffective negotiators, they can’t lead a team of people, they can’t communicate subtleties the same way we can. Robots are still lousy at moving through the physical world and picking up objects, and having a kind of dexterity…

Paul Solman:You mean so we’ve got a little time, still?

Andrew McAfee: We’ve got at least a little time. One of the things that nobody knows the answer to, I think, is exactly how quickly machines are going to continue to encroach into human territory and demonstrate these skills or abilities; it’s not going to happen next year.

Paul Solman: So I don’t have to worry about Max Headroom replacing my job until I’m finally ready to retire?

Andrew McAfee: We think you’re in pretty good shape.