How to Make Sure Robots Help Us, Not Replace Us
An attendee, left, tries out e-eyewear ‘TouchFocus’. (Photographer: Noriko Hayashi/Bloomberg)

How to Make Sure Robots Help Us, Not Replace Us

Bookmark

(Bloomberg Businessweek) -- The world needs robots that make life better, not just ones that put people out of work. But business attitudes, government policy, and scientific priorities are geared toward replacing workers rather than complementing and enhancing their skills. That’s the bottom line of a report by a task force at MIT that was released today.

“It’s super easy to make a business case for reducing head count. You can always light up a boardroom” by promising to replace people with robots, says David Autor, an MIT economist and co-chair of the task force, who gave an interview about the report. It’s harder to get buy-in for robots and other forms of automation that complement human work while improving quality, consistency, and scalability, Autor says.

The tax code makes matters worse because it subsidizes investment in robots, software, and other forms of automation, but is less generous in support of upgrading the skills of human beings, the report says.  (There are tax credits for certain forms of education and training but they don’t measure up to the breaks for capital spending.)

On top of that, the artificial intelligence community suffers from what might be called Turing-test envy—a phrase that does not appear in the report. It’s the idea that the ultimate achievement of AI is to make a robot that can trick a person into thinking it’s human. “The problems that people choose to work on are very much involved in their presuppositions about what is important,” Autor says, adding that the National Science Foundation and Defense Advanced Research Projects Agency add to the problem with their funding priorities.

The irony is that focusing on replacement of workers isn’t just bad for the workers who lose their jobs. It’s also bad for profits. The Wall Street Journal carried a story early this year about the Henn na Hotel in Nagasaki, Japan, which loaded up on robots but ended up getting rid of about half of them because they weren’t as effective as humans. One type tended to wake up guests who were heavy snorers, asking them to repeat their “requests.” 

Caring for the elderly is an example of a field in which human beings will remain vital, with robots assisting in strenuous work such as getting patients in and out of bed, says Elisabeth Reynolds, an MIT principal research scientist who served as the task force’s executive director.

Even Tesla Inc. co-founder and Chief Executive Officer Elon Musk, a huge believer in a high-tech future, has admitted that he went too far with robots at the electric car maker. “Excessive automation at Tesla was a mistake. To be precise, my mistake. Humans are underrated,” he said last year.

The MIT report is in the spirit of work by Stanford University’s Stanford Institute for Human-Centered Artificial Intelligence. “Most jobs are more complex than [many people] realize,” Hal Varian, Google’s chief economist, said during a conference this year at the Stanford institute where Autor also spoke.

A self-checkout kiosk is an example of a “so-so” technology—one that displaces a worker without providing other significant benefits. The “so-so” phrase used in the task force report is a coinage of MIT economist Daron Acemoglu and Pascual Restrepo of Boston University. More valuable, according to the report, are technologies that enhance workers’ abilities, such as the computer-aided design software used by architects and engineers, or “medical imaging tools that boost the speed and accuracy with which medical experts diagnose patients.”

James Manyika, director of the McKinsey Global Institute, told the website Axios this year, “If you were trying to solve this as an economic problem, you’d want to develop AI algorithms or machines that are as different from humans as possible.”

The MIT report says the U.S. has done worse than many other countries in protecting workers from the downsides of automation. Ideally, the U.S. can learn from others how to do things better. Another potential plus for American workers is that low birth rates and restrictions on immigration are likely to produce labor shortages, not surpluses.

The report strikes a note of cautious optimism, concluding that “countries that make well-targeted, forward-looking investments in education and skills training should be able to deliver jobs with favorable earnings and employment security to the vast majority of their workers—and not exclusively to those with elite educations.”

To contact the editor responsible for this story: Cristina Lindblad at mlindblad1@bloomberg.net

©2019 Bloomberg L.P.

BQ Install

Bloomberg Quint

Add BloombergQuint App to Home screen.