Incentivizing positive applications of AI via measuring and ranking deployed systems: Once a government has begun to measure and monitor a salient policy trait, they can synthesize this information and publish indexes that rank systems and research via these traits. There are techniques that researchers use to assess the fairness (or lack of fairness) of deployed AI systems which governments could use rank the ‘most fair’ systems. The positive effects. First, it would give technology consumers and buyers better information about products; second, it would incentivize companies to compete with each other to field systems that could be at the top of such rankings; and third, it would be a disincentive for companies to deploy harmful systems.
