Rise of the Robots: Technology and the Threat of a Jobless Future (11 page)

It would be a mistake, however, to apply that same reasoning to the impact of advancing technology. Up until the moment the first aircraft achieved sustained powered flight at Kitty Hawk, North Carolina, it was an incontrovertible fact—supported by data stretching back to the beginning of time—that human beings, strapped into
heavier-than-air contraptions,
do not fly.
Just as that reality shifted in an instant, a similar phenomenon plays out continuously in nearly every sphere of technology. This time is always different where technology is concerned: that, after all, is the entire point of innovation. Ultimately, the question of whether smart machines will someday eclipse the capability of average people to perform much of the work demanded by the economy will be answered by the nature of the technology that arrives in the future—not by lessons gleaned from economic history.

I
N THE NEXT CHAPTER
, we’ll examine the nature of information technology and its relentless acceleration, the characteristics that set it apart, and the ways in which it is already transforming important spheres of the economy.

*
The Committee on the Triple Revolution did not advocate the immediate implementation of a guaranteed income. Instead, it proposed a list of nine transitional policies. Many of these were quite conventional, and included things such as greatly increased investment in education, public works projects to create jobs, and the construction of low-cost housing. The report also argued for a greatly expanded role for unions and suggested that organized labor should become an advocate for the unemployed as well as those who held jobs.

*
ENIAC (Electronic Numerical Integrator and Computer) was built at the University of Pennsylvania in 1946. A true programmable computer, it was financed by the US Army and intended primarily for calculating firing tables used to aim artillery.

**
Due to a miscommunication, Wiener’s article was never published in 1949. A draft copy was discovered by a researcher working with documents in the MIT library archives in 2012, and substantial excerpts were finally published in a May 2013 article by
New York Times
science reporter John Markoff.

*
Labor productivity measures the value of the output (either goods or services) produced by workers per hour. It is a critically important gauge of the general efficiency of an economy; to a significant extent it determines the wealth of a nation. Advanced, industrialized countries have high productivity because their workers have access to more and better technology, enjoy better nutrition as well as safer and more healthful environments, and are generally better educated and trained. Poor countries lack these things and are, therefore, less productive; their people must work longer and harder to produce the same level of output.

*
There is also a technical issue that comes into play when discussing the gap between wage growth and productivity growth. Both the wage (or, more broadly, compensation) and productivity numbers must be adjusted for inflation. The standard way to do this, and the method used by the US Bureau of Labor Statistics (BLS), is to use two different measures of inflation. Wages are adjusted using the Consumer Price Index (CPI) because this reflects the prices of products and services that workers actually spend their money on. The productivity figures are adjusted using the GDP deflator (or implicit price deflator), which is a broader measure of inflation in the entire economy. In other words, the GDP deflator incorporates prices for a lot of things that consumers don’t actually purchase. One especially important difference is that computers and information technology—which have seen substantial price deflation due to Moore’s Law—are much more important in the GDP deflator than in the CPI (computers are not a big component of most household budgets, but are purchased in volume by businesses). Some economists—particularly those who are more conservative—argue that the GDP deflator should be used for
both
wages and productivity. When this method is used, the gap between wage growth and productivity growth narrows significantly. However, this approach almost certainly understates the level of inflation that impacts wage earners.

*
This is true regardless of political party. In one study conducted by Dan Ariely of Duke University, over 90 percent of Republicans and 93 percent of Democrats preferred an income distribution similar to that of Sweden over that of the United States.

*
SBTC and the college wage premium offer a partial explanation for increasing income inequality. However, since nearly a third of the adult US population has a college degree, if this were the only thing going on, it would imply a much tamer form of inequality than actually exists. The real action is at the very top—and things become more extreme the higher you go. The outsized fortunes of the top 1 (or .01) percent cannot reasonably be attributed to better education or training.

Chapter 3

INFORMATION TECHNOLOGY: AN UNPRECEDENTED FORCE FOR DISRUPTION

Imagine depositing a penny in a bank account. Now, double the account balance every day. On day three you would go from 2 cents to 4 cents. The fifth day would take your balance from 8 to 16 cents. After less than a month, you would have more than a million dollars. If we had deposited that initial penny in 1949, just as Norbert Wiener was writing his essay about the future of computing, and then let Moore’s Law run its course—doubling the amount roughly every two years—by 2015, our technological account would contain nearly $86 million. And as things move forward from this point, that balance will continue to double. Future innovations will be able to leverage that enormous accumulated balance, and as a result the rate of progress in the coming years and decades is likely to far exceed what we have become accustomed to in the past.

Moore’s Law is the best-known measure of advancing computer power, but information technology is, in fact, accelerating on many different fronts. For example, computer memory capacity and the amount of digital information that can be carried on fiber-optic
lines have both experienced consistent exponential increases. Nor is the acceleration confined to computer hardware; the efficiency of some software algorithms has soared at a rate far in excess of what Moore’s Law alone would predict.

While exponential acceleration offers valuable insight into the advance of information technology over relatively long periods, the short-term reality is more complex. Progress is generally not always smooth and consistent; instead, it often lurches forward and then pauses while new capabilities are assimilated into organizations and the foundation for the next period of rapid advance is established. There are also intricate interdependencies and feedback loops between different realms of technology. Progress in one area may drive a sudden burst of innovation in another. As information technology marches forward, its tentacles reach ever deeper into organizations and the overall economy, often transforming the way people work in ways that can further its own advance. Consider, for example, how the rise of the Internet and sophisticated collaboration software has enabled the offshoring of software development; this has made a vastly expanded population of skilled programmers available, and all that new talent is helping to drive still more progress.

Acceleration Versus Stagnation

As information and communications technologies have advanced in their decades-long exponential march, innovation in other areas has been largely incremental. Examples include the basic design of cars, homes, aircraft, kitchen appliances, and our overall transportation and energy infrastructures, none of which, for the most part, have changed significantly since the middle of the twentieth century. PayPal co-founder Peter Thiel’s famous comment—“We were promised flying cars, and instead what we got was 140 characters”—captures the sentiment of a generation that expected the future to be way cooler than this.

This lack of broad-based progress stands in stark contrast to what a person who lived through the final decades of the nineteenth century and the first half of the twentieth would have experienced. Indoor plumbing, automobiles, airplanes, electricity, home appliances, and public sanitation and utility systems all came into widespread use during this period. In industrialized countries, at least, people at all levels of society received an astonishing upgrade in the quality of their lives, even as the overall wealth of society was propelled to dizzying new heights.

Some economists have taken note of this plodding rate of advance in most spheres of technology and have tied it to the economic trends we looked at in the previous chapter, and in particular to the stagnation of incomes for most ordinary Americans. One of the foundational principles of modern economics is that such technological change is essential to long-term economic growth. Robert Solow, the economist who formalized this idea, received the Nobel Prize for his work in 1987. If innovation is the primary driver of prosperity, then perhaps stagnant incomes imply that the problem is the rate at which new inventions and ideas are being generated, rather than the impact of technology on the working and middle classes. Maybe computers aren’t really all that important, and the slow rate of progress on a broader front is what matters most.

Several economists have made this case. Tyler Cowen, an economist at George Mason University, proposed in his 2011 book
The Great Stagnation
that the US economy has run into a temporary plateau after consuming all the low-hanging fruit of accessible innovation, free land, and underutilized human talent. Robert J. Gordon of Northwestern University is even more pessimistic, arguing in a 2012 paper that economic growth in the United States, hampered by a slow pace of innovation and a number of “headwinds”—including excessive debt, an aging population, and shortfalls in our educational system—may essentially be over.
¹

In order to gain some insight into the factors that influence the pace of innovation, we may find it useful to think in terms of the historical path that nearly all technologies follow. Airplanes are a good example. The first controlled, powered flight occurred in December 1903 and lasted about twelve seconds. Progress accelerated from that humble start, but the primitive initial level of the technology meant it would take years before a practical airplane would emerge. By 1905, Wilbur Wright was able to stay aloft for nearly forty minutes while traveling about twenty-four miles. Within a few years, however, things started to really come together; aircraft technology had progressed along its exponential curve, and the rate of absolute progress picked up dramatically. By World War I, airplanes were engaging in high-speed aerial dog fights. Progress continued its acceleration over the next two decades, ultimately producing high-performance fighter aircraft like the Spitfire, the Zero, and the P-51. Sometime around World War II, however, the rate of advance slowed significantly. Aircraft powered by internal combustion engines driving propellers were now very close to their ultimate technical potential, and design improvements beyond that point would be incremental.

This S-shaped path in which accelerating—or exponential—advance ultimately matures into a plateau effectively illustrates the life story of virtually all specific technologies. Of course, we know that as World War II came to a close, an entirely new aircraft technology appeared on the scene. Jet aircraft would soon offer a level of performance far beyond what was possible for any propeller-driven plane. Jets were a disruptive technology: they had an S-curve of their own.
Figure 3.1
shows what this might look like.

If we want to dramatically speed up the pace of innovation in aircraft design, we need to find yet another S-curve, and that curve has to represent a technology that is not only superior in terms of performance but also economically viable.
*
The problem, of course,
is that so far, that new curve is nowhere to be found. Assuming we can’t discover this disruptive new technology simply by hopping the fence at Area 51, it’s going to take a giant leap to get to that new S-curve—and this presumes, of course, that the curve even exists.

Figure 3.1. Aircraft Technology S-Curves