The Exponential Paradox
It is an exciting time to be alive in Silicon Valley. Digital technologies, blockchain, AI, dancing robots, private space companies and autonomous vehicles abound around here. Technologies are advancing at an increasing rate, which some call “exponential technology”. This augurs well for our future, because an accelerating rate of advance promises to usher in an era of abundance. Imagine our world in twenty years time, according to IBM’s Jim Spohrer, when commonly available AI technologies will give each of us access to the equivalent of 100 human experts, personally trained by you to provide all the information and advice you need in living your daily life.
Yet a moment’s reflection reveals that this abundant future is missing some essential details. Our rate of economic productivity growth, for example, is going down, not up. The rate of productivity growth in the US has been just 1.1%, far below the 3.7% rate of growth we achieved from 1947 to 2007, according to the US Bureau of Labor Statistics. This is not what exponential technology has promised us. Indeed, I consider it an Exponential Paradox: technological advance is accelerating, while economic productivity growth is declining.
To explain this paradox, It helps to consider a few key points. One is a possible measurement problem, where the new technologies require new measures to capture their true value to society. This is a comforting thought, because it suggests that the rewards are coming, we just need to be patient. But it’s not at all clear why productivity growth would be declining, compared to its historic average. Another possible cause is a growing gap between the Best and the Rest, where the Best organizations are indeed accelerating their growth and productivity, but the Rest are falling farther and farther behind. This is a more concerning thought, because it implies that the economic inequality we already observe may intensify further, precisely due to the (uneven) spread of exponential discoveries.
A third possibility, and a more hopeful one in my view, is that as a society we are living with the failure to renew our investments in innovation infrastructure. Our innovation infrastructure consists of the hard and soft assets in the society to generate, disseminate, and absorb new innovative knowledge. This requires investments in hard assets, like 5G connectivity or up-to-date airports, roads and train stations, as well as investments in soft assets, like training, skills, universities and other forms of human capital development. The infrastructure starts with public investment, which in turn attracts a larger amount of private investment, which culminates in greater innovation capability for the whole society.
We used to do this well. After the second World War, we passed the Serviceman’s Readjustment Act of 1944, known as the GI Bill, and sent millions of returning government soldiers off to college, which greatly enhanced the skills of millions of workers and boosted the productivity of American businesses. In the 1950s, we built the Interstate Highway system, which greatly improved road transportation, and supported myriad new businesses across the land. In the 1960s, we launched DARPA and the Apollo moon program, and developed a wide variety of computing, semiconductor, communications, and materials technologies that culminated in the Silicon Valley of today. These were huge investments that resulted in strong economic productivity growth for decades. We led the world in making these investments at that time. In many ways, we are still coasting on the innovation infrastructure that we built 60 or more years ago. At a time when interest rates are at historically very low levels, the cost to finance a renewed boost to our innovation infrastructure is surprisingly affordable.
Perhaps the best way to resolve the Exponential Paradox is to take the position that the exponential technology folks have it exactly half-right: accelerating technological advances CAN generate an era of abundance, but only if we as a society also make the necessary investments in the innovation infrastructure required to disseminate and absorb that new knowledge in new platforms, systems, business models, and worker skills.