(Bloomberg) -- Don’t give up on the robots just yet.
Artificial intelligence and machine learning are not currently driving major pickups in productivity and output, but that doesn’t mean they won’t, Erik Brynjolfsson at the Massachusetts Institute of Technology and Chad Syverson at the University of Chicago write in a new paper.
That’s the lead item in this week’s economic research roundup, which also summarizes studies on the dollar’s influence on global trade and a look at how the recession scarred trade. Check this column every Tuesday for summaries of new and pertinent studies from around the world.
("A hopeful case for AI-driven outputArtificial Intelligence and the Modern Productivity Paradox: A Clash of Expectations and Statistics," published November 2017, is available on the NBER website).
Productivity growth has declined by half over the past decade and inflation-adjusted incomes are stagnating. The trends are confusing in an era marked by technological advancement and the dawn of artificial intelligence, but Brynjolfsson, Syverson and their co-author Daniel Rock argue that this is partly due to a lag effect: AI and machine learning technology capabilities just haven’t had time to diffuse widely.
New technologies have historically taken a long time to spread in an economically important way, the authors note. E-commerce didn’t become the driving disruptor in the retail industry until years after the dot-com boom, for instance, as infrastructure and customers took time to adjust. Now, leveraging the newest technologies will require investments in human capital and skills and the creation of new processes and business models. “Both the AI investments and the complementary changes are costly, hard to measure, and take time to implement, and this can, at least initially, depress productivity as it is currently measured,” the authors write.