Books

03.14.14

Relax, Robots Won’t Take Every Job: ‘The Second Machine Age’

The co-author of ‘The Second Machine Age’ talks about the possibility that machines will put us all of work and why that is just one of several scenarios of the future.

What is technology doing to us? Between the digital skeptics and the wide-eyed utopians sit the authors of The Second Machine Age, two MIT scholars with an interest in consensus and moderation. Erik Brynjolfsson and Andrew McAfee, both at the university’s Center for Digital Business, realized that things were moving way faster than even they would have predicted. Just as the Industrial Revolution saw machines replacing human brawn, the second machine age sees them replacing our cognitive faculties. These two, whose previous book Race Against the Machine took a cautionary look at the way computers and artificial intelligence were making human beings obsolete, now describe themselves as cautiously optimistic. We spoke to McAfee (who calls William Gibson his favorite living science-fiction author) about the new human condition.

The question that runs through your book is seems to be, What do human being do that machines can’t? As recently as a decade ago, the smartest people pursuing that question thought it came down to pattern recognition and “complex communication.” How has it changed since? Are we realizing that it’s not that simple?

That’s absolutely the case. Winning at Jeopardy, driving a car in traffic, understanding what I am saying now and what to give back to me – these are all exercises in pattern recognition and complex communication. And technology has demonstrated that it can do all of these things pretty well right now, and getting better quickly over time. So our old frameworks of what people were sustainably better at, they don’t hold up anymore.

You knew the process was moving quickly, but it’s now a lot faster than even a few years ago.

There’s a great quote from Hemingway about how a man goes broke, gradually, he says, and then suddenly. We’re at the suddenly part now.

Let’s talk about what this will mean for people’s working lives. There’s a long history of people being put out of work by machines, going back to the Industrial Revolution, and then in the 20th century, secretaries, bank tellers, and so on. What’s happening now, and what can we expect over the next decade.

Let me give you three scenarios for what’s happening now. The most conservative one is that this wave of technological displacement is like the one that took people from farm to factory, or from factory to office building. Economists agree that these things have gone on, and they’ve led to a wrenching displacement of a lot of people. What they have not done is led to permanent technological unemployment. So the most conservative scenario is that we’re seeing that move again – it’s a difficult transition, the Great Recession makes it more painful, but we’ll get through this and will end up with a new equilibrium with people and machines working together. And something like full employment.

The second scenario is like the first, but there is a series of shocks instead of one. There’s one in robotics, one in natural language programming, one in automated vehicles, one in 3-D printing – the shocks keep coming at us, keep barraging the workforce. And therefore the transitions – which are all individually tough – when they come at us so thick and so fast, it makes the recovery, the reemployment, the reskilling, the retraining – it makes the challenge much, much harder.

Scenario number three is like scenario number one, with one big transition, but we don’t go back to full employment. We go into a science fiction, I, Robot, Player Piano, Star Trek, Elysium kind of economy, where there is just not a lot of work to go around. Potential employers say, “There’s an incredibly cheap robot over there, an incredibly smart AI over there. I want to start a company – I don’t need a lot of people to do it.”

In contrast to your last book, Race Against the Machine, you’re more optimistic this time around.

In The Second Machine Age, we’re interested in stressing the good news out there. Because the narrative “the robot ate my job so therefore the robot must be unplugged”… I really don’t like that narrative. There’s a very broad and deep unease about progress – it’s appropriate in some way. But to say we need to unplug … stopping progress is a bad, bad, bad idea.

Has it ever happened historically?

Sure. The Japanese were at one time the world’s best gunsmiths, but for a lot of reasons they outlawed the manufacture of guns, and they fell way behind on their military power. The Chinese gave up on ocean-going vessels, which they used to be very, very good at.

You’re documenting seismic changes in your book – some of which have arrived, some of which are still coming. What are your suggestions for individuals and for U.S. policy, as to how we can move into this period in a humane way?

My main policy recommendation is, Let’s not get ahead of ourselves. We’re recovering from a very nasty recession; we don’t know which scenario we’re in, but we know that the economy still needs work and still needs people. So let’s grow the economy – let’s focus on infrastructure, immigration reform, and entrepreneurship. If we grow the economy, we will grow jobs.

Our long-term recommendation, if we are moving into scenario two or three, it might be a basic income guarantee or a negative income tax. But it’s my sense that work is a good thing for a community, for an individual – when work goes away, a lot of what happens is distressing.

Much of what we’re talking about – the Internet, the driverless car – blossomed out of a period of public investment. We had, from World War II through the ’70s, comparatively high taxes, federal funding, state research laboratories, that led to these things. Are we still investing that way, or have we eaten the seed corn?

The thing I didn’t mention, as to what government should be doing, is very basic research. Much of the theory and evidence says that left to its own devices, private industry doesn’t do enough of that stuff. Things like Google X and IBM’s research labs are fantastic. But in some areas, basic research is dropping.

I ask this with the knowledge that much of Silicon Valley is more or less libertarian, in love with the free market: Without significant government investment in technology, paid for by tax revenues, would the Internet have arrived even vaguely when it did?

No! When you look at most other big-deal technology, including the iPhone I’m talking on right now, and the family tree of that innovation, you don’t have to go back too many generations before you find government programs.

Near the end of your book, you and your co-writer say that you are not technological determinists. Why not? We’ve seen technology shape human history – and it seems to be true increasingly.

We’re just not determinists – if that were not the case, every country, in the rich world, at least, would look exactly the same since we’ve all had the same technology since the Industrial Revolution. Instead we have this incredible variation because of choices people make, through their political systems. And these differences matter, for the overall health of the economies and the quality of people’s lives.

So what Erik and I are very careful to do is never to throw up our hands and say, “There’s nothing to be done here.” The astonishing acceleration we’re seeing is going to continue for some time. But that does not mean that technology has dealt the hand to us and all we can do is play the card.

You say near the end of the book that what we’re seeing now – the driverless car, the 3-D printer, all these things unimaginable a decade ago – are not the culmination of the second machine age, but the warm-up for something that’s still coming.

Oh, yeah. We ain’t seen nothing yet.