23 October 2018

The Power Structure of Artificial Intelligence

Connecting the three-tiered governance challenges posed by AI

Much of the scholarship to date on the regulatory responses to artificial intelligence focus either upon the near-term immediate challenges which revolve around discrete decision-making processes being delegated to AI applications, Hin Yan Liuor concern more speculative questions concerning the impact of AI upon the future of humanity. Such divergent responses tend to overlook the distortions to values at an intermediate level.

This article seeks to unite these disparate approaches, and does so through the lens of power. What happens if, instead of concentrating upon artificiality and intelligence as the definition of AI, we simply consider AI to be a novel form of power or a new entity possessing and exercising power? Suddenly, this makes lawyerly perspectives relevant to the governance of AI, and foregrounds more familiar responses which might also be effective in stemming some of the problems that AI poses. Such an approach might also demonstrate that the challenges that are perceptible at each level are all part of the ‘whole problem’, thereby eliminating the opportunity costs in pursuing research at each of the various levels.

Read the article