What’s the outlook for innovation? That’s a seemingly superficial question. But in many ways, it is the central question for investors, policymakers and citizens. The rate of innovation, which we measure after the fact as “productivity growth,” determines the rate of per capita economic expansion, which is the key factor for personal incomes, government budgets and our overall standard of living. In short, it’s everything.
At just over 1 percent, however, annual productivity growth in the U.S. over the last dozen years has stalled, leaving behind a giant growth gap. If the economy had resumed its historical rate of expansion after the Great Recession, annual U.S. output would be nearly $3 trillion larger today. Instead, it’s taken a decade for the economy to look even halfway normal again. Continued slow growth, moreover, will only intensify the nation’s political anxieties.
Smart people, however, predict wildly divergent futures. Economist Robert Gordon argues that the information age is basically over and that it wasn’t all that powerful to begin with. With most of the important technologies already invented, we should expect 25 years of stagnation, according to Gordon.
Tesla CEO Elon Musk, on the other hand, insists innovation is racing ahead so fast that we must preemptively regulate artificial intelligence (AI), lest AI take all the jobs and even challenge humans for control of the world.
Slow productivity growth over the last decade seems to bolster Gordon’s pessimism. Musk’s worries about a runaway future, meanwhile, gain credence from his own self-landing Space-X rockets, recent AI wins over humans in complex games like "Go," and mind-blowing genetic editing advances, such as CRISPR.
I think both brands of techno-pessimism — tired impotence and dangerous omnipotence — are wrong.
The fact is that technology-intensive sectors of the economy are generating both productivity gains and lots of high-paying jobs. In recent research, Michael Mandel and I found that the digital economy, which accounts for around 30 percent of U.S. output, makes 70 percent of the investments in information technology and has enjoyed robust productivity growth of 2.7 percent over the last 15 years.
It’s the sectors of the economy that aren’t fully leveraging technology that are suffering both low productivity growth and slumping jobs and incomes. Productivity growth in the 70 percent of the economy that is underutilizing technology was just 0.7 percent over the same period. Digital jobs and incomes grew faster, too.
This massive “information gap” helps explain the dissonance between the Gordon and Musk worldviews. Musk only sees the rapidly advancing digital world, while Gordon focuses on the backward-looking data, which show the bulk of the economy struggling.
New research shows that official estimates likely understate the productivity of information technologies and overstate productivity in the rest of the economy. This only bolsters our case that creative use of information technology is central to innovation.
Fortunately, information is about to diffuse through the other 70 percent of the economy. Healthcare, education, transportation, manufacturing, energy, retail and advanced distribution are all in the early stages of historical transformations, based on creative applications of mobile devices, cloud computing, artificial intelligence, 5G wireless networks and the Internet of Things. Far from an end, the information age will likely supply its biggest and broadest benefits in the decades to come.
If Gordon’s outlook of stagnation is wrong, so too is Musk’s prediction of catastrophe. A common joke is that we should start worrying about superhuman AI when your smartphone’s autocorrect starts working, or when websites don’t display new advertisements for the items an individual just purchased.
If we analyze the situation more quantitatively, it’s clear we’re not close to matching or exceeding the brain in silicon and software. The human brain has a million times more synaptic connections than today’s best artificial neural network. At around 20 watts, the brain uses just 1/50,000 the energy of the Go-winning AlphaGo computer, thus giving it something like 50 billion times the general purpose thinking ability per watt. Fifty billion is a large gap, even for computation that advances at something like the pace of Moore’s law.
So Musk’s calls for preemptive regulation aren’t just overkill, they could actually delay the diffusion of these important technologies across the sectors of the economy that desperately need jolts of innovation. Far from imposing new regulations on the economy, to fully realize the promise of technology, we need to free many of the lagging sectors from outdated rules and a stifling tax code.
Massive infrastructure projects, such as the new 5G wireless network, new advanced manufacturing investments and big biomedical research programs will all happen much faster if we substantially reduce the world’s highest corporate tax rate and reform the indefensible code.
If Washington can get its policy act together, the nation’s economic future is actually very bright.
Bret Swanson is president of the technology research firm Entropy Economics LLC, a fellow at the U.S. Chamber of Commerce Foundation and a visiting fellow at the American Enterprise Institute.
The views expressed by contributors are their own and not the views of The Hill.