Engines that learn: Digital general-purpose technologies and the concentration paradox

Abstract

Digital general-purpose technologies represent a new engine of economic growth, with the potential to drive the next wave of innovation and productivity gains. However, the same mechanisms through which these technologies enhance productivity can also contribute to market concentration. The very large dataset that these technologies generate and could enable digital technologies to learn, improve and innovate themselves. They can also make it harder for competitors to enter the market. This creates a policy dilemma as existing competition and intellectual property (IP) frameworks may no longer be sufficient to address the challenges emerging in digital markets.