By Richard Freeman, Herbert Ascherman Chair in Economics at Harvard University.
This week Google’s alphaGo artificial intelligence program defeated human champion Lee Se-dol in the ancient game of Go. This triumph of artificial intelligence follows computers beating humans in Jeopardy, Chess, and enough other activities to suggest the world of work is entering a new era. In 2015 Stephen Hawkings warned of the dangers of artificial intelligence and robotics. Elon Musk, and many others, raised concerns about machines that could outperform humans in diverse ways. Every day the media reports new robots able to do human work – as surgeons, nurses, receptionists, accountants, and so on.
Economists tend to view the fears of robots taking our jobs as science fiction that ignores economic history and logic. Past fears of automation have proven false as economic growth has turned recession joblessness into booms. Comparative advantage tells us that as long as machines and humans have different skill set, the economy will allot the work that machines do relatively better than humans to the machines and the work that humans do better to humans, even if machines have an absolute advantage in all work tasks. To overcome global problems from climate change to poverty for billions of people requires lots of work from both humans and robots.
And yet… the constant flow of reports of better computer programs and robots able to do our work is not science fiction but reality … today we follow the GPS program that tells us where to drive our car but tomorrow we will sit in driver-less cars. What will the Go learning algorithm do next?
Today’s technology differs from the mechanization and automation of years past, by being increasingly based on learning algorithms that continually make the computer better at its task, and on soft robotics in which the machine mimics human behavior. Developments in quantum computing that increase the speed of machines millions of times faster than current computers will likely create ever stronger machine competitors for the work humans do in the next decade or today.
Today’s economy is also different from the economy of years past. In nearly all countries capital’s share of income has risen. Inequality in labor earnings has increased as the gains of technological advance accrue to those at the very top, whose pay is linked to capital through stock grants, options, and bonuses. The labor unions whose task is to raise workers pay and giving them a share of increased output are weaker than they have been in decades.
My analysis of the future of technology and jobs follows from the three laws of robo-economics:
With the current distribution of capital ownership, the beneficiaries from improved robotic technology will be the small number of people with great wealth. To assure that the workers who will face increased robotic competition benefit from this technology ownership of capital must spread to more people. The million dollar prize that went to AlphaGo for winning the Human/Artificial Intelligence contest is just the tip of the rewards that will go to robots able to do our jobs better than we can. The solution lies in public and private policies to spread ownership.
About the Author
Richard Freeman holds the Herbert Ascherman Chair in Economics at Harvard University. He is currently serving as faculty co-director of the Labor and Worklife Program at the Harvard Law School, and is senior research fellow in Labor Markets at the London School of Economics’ Centre for Economic Performance. He directs the National Bureau of Economic Research / Sloan Science Engineering Workforce Projects, and is co-director of the Harvard Center for Green Buildings and Cities. Freeman is a fellow of the American Academy of Arts and Science and the AAAS. He is currently serving on the AAAS Initiative for Science and Technology. Freeman has served on 11 Panels and Boards of the U.S. National Academy of Science. He has published extensively on a variety of labor markets issues. His current research activities include role of firms and institutions in inequality and unions and workplace organization.