Artificial Intelligence (AI) is developing rapidly and countries from around the globe are beginning to articulate national strategies for handling the political ramifications. Powering innovations like driverless cars, autonomous drones, full-sequence genetic analytics and powerful voice assistant technology, the future certainly looks full of potential. However unsettled questions about who will reap these benefits and when they will be achieved leave storm clouds on the political horizon. Amid questions of industrial concentration and economic inequality on one side, and concerns about lagging U.S. productivity and the slow pace of AI diffusion on the other, this paper attempts to lay out a framework that can begin to address these various issues.
The first of these questions could be simplified to ask: what if only Google, Amazon, Facebook and Apple (GAFA) are able to develop the AI system that powers the economy of the future? The second considers the reasons that AI techniques have diffused through the economy at such a slow rate. However, although these appear to be two distinct questions, there is an under-examined overlap that connects these issues to the same set of policies: namely, high barriers to entry due to supply- side constraints.
There are significant barriers to entry in AI development and application, many of which stem from the direct result of government policies. These barriers have inadvertently boosted the market power of incumbent firms and thus in reducing them, we may enable new firms to better compete, while also removing some of the bottlenecks that slow down research and integration of AI systems across the entire economy.
Accordingly, the present study provides an overview of the various inputs to the production function of AI and analyzes the policies that should be reconsidered or implemented to reduce these barriers. Specifically, it will focus on the inputs of skilled AI analysts, high-quality datasets and specialized AI hardware. It will conclude with a short discussion of the relative attractiveness of focusing on entry barriers when compared to the high-risk options presented by traditional antitrust enforcement. While there are certainly other potential policies or AI inputs that are beyond the scope of this paper, the policy framework presented herein will nevertheless provide a useful primer for future analysis.
