It may not be obvious, but the U.S. presidential election offers critical lessons about how policy and technology leaders should think about the future of artificial intelligence. In fact, just days before Donald Trump was sworn into office, these lessons were a focus of the Davos meeting of the global elite.
Technology executives expressed concern over a growing fear throughout the world that robots destroy jobs and discussed the possibility of a backlash against innovation. It was this same fear of job loss that has contributed to the recent backlash against trade agreements.
After all, proponents of trade agreements won every argument except one: that trade increases employment. That made killing the Trans-Pacific Partnership (TPP) a central message of candidate Trump’s campaign. It was one of the first things President Trump did after taking office.
The lesson is clear. When it comes to artificial intelligence, the industry might win every argument about innovation, progress, and new goods and services, but lose the jobs argument. If that happens, technology companies could face new limits on digital commerce, reduced investments in research and development, burdensome tax treatment, and more.
The opportunity of automation is enormous. Consider that, as autonomous vehicles become the primary means of transportation, accidents will decline by 90 percent, saving lives and billions of dollars. Furthermore, automation will actually return jobs to the United States. One-quarter of the decline in U.S. manufacturing jobs is due to competition from China, driven largely by lower labor costs. But this offshoring is a station on the way to the new globally-competitive automated U.S. factories that are creating good paying jobs for skilled workers.
Of course, computer technology does affect the nature of work. It has eliminated some tasks and lowered demand for some workers. A recent study by McKinsey & Company estimates that almost half of all current tasks are subject to automation, providing fodder for arguments that widespread technological unemployment is near. But the story is more complex. Computers can eliminate all job-required tasks in only 5 percentof occupations, and there will still be plenty of tasks to perform in existing occupations, while many new tasks will be created.
We’ve already seen the way automation creates efficiencies that lower production costs, thereby stimulating demand and creating more jobs. Recent history is filled with examples of lowering operating costs. ATM machines led to increased bank teller employment, and cost savings created by robots have actually increased human employment in warehouses. In the overall economy, automation has led to a greater need for non-routine, high-skill work that pays high wages and for low-skill work that pays lower wages.
While all this may be true, the reality is that the world is focused on bridging income divides and spreading economic opportunity. We have a responsibility to make certain that the bounty of automation can benefit everyone.
An important step is to match computers with human skills. On the computer side, this means creating programs that augment human skills. As described by IBM data scientists, humans and machines will “need to collaborate to produce better results, each bringing their own superior skills to the partnership.”
On the human side, people need to be trained for tasks computers cannot perform. This means prioritizing science, technology, engineering and math (STEM) education. But that’s not the only solution. Our computer-intensive work environment is creating high-paying jobs for those with credentialed skills from quality technical schools or training programs. Reauthorizing the career and technical education program with adequate funding will jump-start the programs that can match human skills with the new workplace, which has many unfilled jobs waiting for skilled workers.
Even with these efforts, some workers will not be able to gain the skills needed to flourish. A late-career truck driver without a college education can’t be expected to become a coder. For many of these workers, a social safety net is essential, and that net can be supported by the wealth that technology generates. Policy and technology leaders must work together on programs that support the collective good.
Ultimately, technology can continue to create more jobs than it displaces, while driving U.S. economic gains. But the only way to achieve the full measure of this opportunity is to ensure that the benefits are clearly realized by those who see technology as more of a foe than a friend.
Mark M. MacCarthy is senior vice president of public policy at the Software & Information Industry Association. He has been a consultant on technology policy issues for the Organization for Economic Cooperation and Development and the Aspen Institute. He is an adjunct professor of communication and technology at Georgetown University, where he teaches courses on artificial intelligence and the future of work.