Economist Allays Fears of Artificial Intelligence in Workplace
Robotic technology is more likely to improve medical treatments than replace doctors, according to Robert Atkinson. | David Stillman/Flickr CC BY-NC 2.0
Fears that advances in artificial intelligence, machine learning and robots will gut the labor force are overblown, according to the leader of a science and technology policy think tank.
Robert Atkinson, economist and founder of the Information Technology & Innovation Foundation, spoke at AAAS headquarters on May 2 about the future of AI and other technologies and their effects on the labor force. The presentation was co-hosted by the AAAS Colloquium Series, a monthly program that explores timely topics relevant to science and society, and the AAAS STEM Volunteer Program, which deploys scientists and engineers as volunteers in K-12 classrooms in the Washington, D.C., area.
Atkinson started by reassuring the audience and dismissing famous doomsayers. “You can ignore Elon Musk,” said Atkinson. “He certainly doesn’t understand AI when he says that we’re ‘summoning a demon,’” quoting the Tesla CEO’s 2014 remarks.
Even the predictions of academics – one widely cited 2013 study from Oxford University researchers estimated that 47% of jobs in the U.S. are at risk of being automated in the next 20 years – are wrong, Atkinson said.
Why are such studies off-base? Atkinson said their formulas take into account occupations with characteristics that might make automation promising without considering other practical factors. Take, for instance, an occupation that some studies predict may soon be obsolete: school bus driver. Despite the rise of self-driving vehicle technology, Atkinson argued that studies ignore a key factor. “Any of you who work with kids think that you would let your students or your kid ride on a school bus with no adults?” Atkinson asked, getting laughs from the audience.
Robert Atkinson | Andrea Korte/AAAS
Many occupations are hard to automate in an effective – and cost-efficient – manner, he said. Even a reasonably priced robotic grip hand, for example, is still more than 15 years away, he said. “The idea that fully intelligent machines are going to be here is like saying that just because we invented the internal combustion engine, warp drive is around the corner,” Atkinson said, paraphrasing roboticist Rodney Brooks.
We have been panicking about the effects of automation on people and the labor force for decades now, but technology-driven job change trends are nothing new, Atkinson said, citing the effects of technology on locomotive engineers – the fastest growing occupation in the United States in the 1850s, he noted – and auto mechanics. The effect of AI on the labor force is “not going to be as disruptive as some people think,” he said.
Atkinson predicts that the most important AI will have greater effects on quality of life than on the labor market. For instance, a doctor, rather than being replaced by a robot, might use AI tools for diagnostics or treatment to improve medical care, he said.
Atkinson’s predictions were not completely optimistic, however. Jobs will be eliminated, but “you won’t have a problem if you’re a lawyer,” he said. Instead, people in low-wage, low-skill jobs will be hardest hit. And even though AI technology will create new jobs, the numbers will be insufficient to replace all eliminated positions.
Some leaders in tech – including Elon Musk and Bill Gates – have suggested implementing “robot taxes” to stem the potential displacement of workers by automation, but Atkinson argued such taxes would simply slow the rate of innovation. Instead, we should have a “pro-technology” outlook that will enable us to improve productivity. “We’re going to need every ounce of productivity growth we can get out of the economy,” he said.
Atkinson offered several educational options for ensuring that workers are well-equipped to participate in a labor force that is enhanced, not replaced, by AI, including better on-the-job training for workers and exposing students early on to fields like statistics and computer science.
Reviewing the employment landscape, Atkinson noted that this era of innovation “will provide all sorts of new opportunities as we go forward.”