A businessman turned politician calls for insurance against job losses in a quickly changing America. His own experience in business pits him against an indifferent Washington, D.C., as he becomes the lone voice of a single issue. Eventually, he decides to run for president. Who is this man? Jacob Coxey.
In 1894, this industrialist organized a march on Washington, D.C., to demand employment security for Americans. In a rapidly industrializing country, he argued that sudden unemployment during economic downturns would become a growing social problem. He, and the tens of thousands of people who marched with him from Massillon, Ohio, to the capitol, demanded the creation of unemployment insurance. Unfortunately for Coxey, only 500 finished the march and it ended with his arrest for walking on the grass of the capitol building. But that was the first sign of a movement that 41 years later gave us FDR’s Social Security Act.
When I saw Andrew Yang at last night’s debate, I was reminded of Coxey. Yang spoke about seven times. Nearly every time, he made a direct or indirect plea for his pet issue: universal basic income (UBI.) Yang, like Coxey, warns us of an impending crisis. Over the next twelve years, he says, AI and automation will decimate whole industries like call centers and trucking, leading to mass unemployment. His solution: a monthly “freedom dividend” of $1,000 to help us navigate a future of career disruption.
We’re not yet seeing evidence that AI will decimate jobs, although it’s early. Gartner predicts that the number of AI and ML uses cases will increase tenfold over the next six years. According to a University of Oxford study, 47% of U.S. jobs could someday be automated. AI, like the kind of automation that powered textile mills in Coxey’s day, will create jobs as it replaces them. But will net employment continue to outpace job losses? And how uneven will these gains and losses be across blue- and white-collar workforces?
I believe automation will cause significant job losses in certain sectors and significant disruption in others. But will the speed of this displacement be so abrupt that we need UBI? And does it make sense to start those payments now? Yang argues that it takes time to put such measures in place, hence the need to address it now, before an army of displaced workers marches on Washington again.
It’s true that call centers have been rapidly automating. It’s expected that 15% of all call center jobs will be automated by 2021. But artificial general intelligence (AGI) that could handle any customer with any issue, may literally be decades away. Therefore, some number of nimble, creative-problem-solving human brains will still be needed in call centers for the foreseeable future.
Trucking seems like a no-brainer. We have self-driving cars, so why not self-driving trucks that only need to stop to refuel? I can tell you why this transition may not be imminent: It costs at least $30,000 to retrofit a truck to self-drive. There are approximately 3.46 million heavy duty trucks in the U.S. and the industry employs 3.6 million people. How quickly will the trucking industry be able to afford a $103.8 billion investment in automation? And I haven’t even mentioned the legal and regulatory challenges that must be resolved before the highways can fill with autonomous vehicles.
There will be natural human resistance to overcome as well. Healthcare is an industry where both patients and caregivers have shown receptivity to AI, but according to our research, only as assistants or advisers, not as artificial healthcare providers. Both “physicians and patients have a visceral and emotional reaction to AI impacting the doctor-patient relationship.” How many other industries will have to navigate a similar resistance to AI?
Yang is probably not the next FDR. But, Yang may be another businessman-turned aspiring politician who manages to be ahead of his time before fading into obscurity. I don’t know if Yang will ever be vindicated, but I do suspect he won’t do better than Coxey. Our need for UBI will be directly proportional to how advanced AI becomes and how affordable it is to automate aspects of our economy. Right now, no one can predict either.