Will Your Job Disappear?
Erik Brynjolfsson and Andrew McAfee, both at MIT’s Sloan School of Management, are co-authors of “The Second Machine Age.” James Manyika is head of the McKinsey Global Institute.
As the worldwide slump stretches into its sixth year, we hear ever-louder and more insistent cries that we’ve reached the end of innovation, and that growth is dead.
Don’t believe them.
It’s natural for pessimism to spread during downturns, and the Great Recession was the worst slump in 70 years. But history shows that it’s a serious mistake to equate an economic contraction with a technology crisis. Recent research by Alexander Field makes a compelling case that the 1930s were actually among the most innovative decades in American history. In fact, the 30 years after the Depression actually delivered record productivity growth and an unprecedented surge in living standards.
When we look past the current business cycle, we are convinced that the next 30 years will deliver even greater innovation and economic bounty to America and the world. We’re in the early stages of what we’re calling the Second Machine Age; the first was the Industrial Revolution, which began in the late 1700s and changed the world like nothing before or since.
Is it fair to say that we’re now entering a similarly transformative period? Consider the breakthroughs we’ve seen in just the past decade. For the first time in history, we have technologies that understand our speech, carry out our wishes, and report back to us in a synthesized voice. Other applications use related natural language processing capabilities to write news, sports and business stories and sift through millions of legal documents to find which ones are relevant to a case. For the first time, machines can beat humans not only at chess and other highly structured games, but also answer unstructured queries on a wide range of topics quickly and accurately enough to beat world champions in games like Jeopardy!
And digital technologies are mastering not only knowledge work, but also activities in the physical world. Historically, robots have struggled mightily to map an unfamiliar room and place themselves within it. They’re now accomplishing this with the Microsoft Kinect, a sub-$150 consumer device. For the first time in history, we see cars built by Google and Mercedes drive themselves on streets and highways, in traffic. Leading car manufacturers like Nissan and Daimler Benz have committed to bring this technology to the general public by the end of the decade. Three dimensional printers are not only creating low cost items for consumers’ homes, but also high precision components for jet engines. Last year, the McKinsey Global Institute catalogued a dozen breakthrough technologies like these that could each have over a trillion dollar impact on the global economy in the years to come.
Just 10 years ago, summaries of progress in autonomous vehicles, natural language processing, or robotic localization and mapping made for depressing reading. But then all of these fields took off. Improvement in digital technologies brings to mind Hemingway’s observation of a how person goes broke: “gradually, then suddenly.”
Why is the sudden progress happening now? The Second Machine Age has been sparked by the confluence of three powerful trends. The first is an exponential improvement in computational power, memory and storage capacity, bandwidth, and most other elements of digital gear. Computers get better faster than anything else, ever. For example, in 1996 the $55 million, the ASCI Red Supercomputer, which was the size of tennis court, became the first computer to reach a speed of 1 Teraflop (trillion floating point operations per second). Last year, Sony introduced the PlayStation 4 which has almost twice the computational power and costs more than 100,000 times less. By comparison, it took 70 years for the steam engine to double in efficiency.
The second characteristic of this age is the immense amount of digital information now available. Digitized information has some weird and wonderful properties: it can be transmitted almost instantaneously and copied at virtually zero cost, and copies are identical to the original. The idea of having goods that are free, perfect and instant obviously leads to some very unusual economics compared to traditional goods and services. And these economics aren’t limited to a few small sectors of media or high tech. To use venture capitalist Marc Andreessen’s great image, software is eating the world, meaning that digitization is becoming the core of how smart companies in every industry innovate, improve, and change themselves.
The third characteristic is the combinatorial nature of innovation. Digital innovations can be combined and recombined to create ever-more value. And that’s a very encouraging thing; a larger base of inventions is a larger set of building blocks that can be reassembled to yield the next wave of innovations. This combinatorial view is very different from the traditional perspective that innovations like the internal combustion engine get “used up” over time. Autonomous vehicles show how much more can be wrung out of this technology once it’s combined with some exponentially improving computer hardware and vast amount of digital data.
Any societal transformation that is exponential, digital or combinatorial will have tremendous implications. A transformation that has all three characteristics is a true game changer. While we’ve seen significant economic disruption in the recent past, we’ll see much more as we head deeper into the Second Machine Age. Improved technologies are the ultimate driver of productivity growth, and productivity growth is essential to higher living standards. Today, the net wealth of the United States is at a record level, $77 trillion, as are other metrics like GDP and corporate profits, and they’re all growing. Technological progress is the only free lunch that economists believe in. In the Second Machine Age, a feast awaits.
However, there’s no guarantee that everyone will have a seat at the table. For much of the 20th century, improvements in productivity and wealth were typically widely shared in America and elsewhere. This created an implicit contract that led to optimism and support for technological advancement.
But there’s no economic law that this will automatically continue to happen. It’s possible for a technology to create enormous wealth, while also re-arranging how the economic pie is shared. In particular, new technologies demand new ways of working. If our people and organizations don’t keep up, then many may end up worse off than before. To a troubling extent, that has happened in the past decade, reducing the opportunities and economic mobility of many Americans.
The improved technologies of the Second Machine Age will create enormous bounty, but shared prosperity depends on more than technology. It also depends on our willingness to reinvent our skills, organizations and policies to keep pace with accelerating technical change.