Are Judges More Transparent Than Black Boxes? A Scheme to Improve Judicial Decision-Making by Establishing a Relationship with Mathematical Function Maximization

This article proposes a way to improve judicial decision-making which entails that judges communicate all legally valid solutions to the litigants and, more broadly, to the outside world. One of the major benefits of this proposed way of judging cases would be the reduction of the human black box character that is currently present in many, if not most, judicial decision-making settings around the world. For example, judges typically produce a single outcome that resolves the case at hand, although they probably thought of other possible solutions. Even though these other solutions might also be legally defensible, judges restrict their judgment to that single outcome. Reasons for not mentioning other outcomes might include the fear that some alternative outcomes are socially not acceptable, the urge to maintain the appearance that the law is unambiguous, or the lack of willingness to thoroughly ascertain all legally valid solutions and compare them to each other. In contrast, black boxes from the domain of artificial intelligence are easily able to produce a set of alternative solutions to a given problem and produce all reasonable predictions in a forecast study. In the debate on the applicability of artificial intelligence tools in legal practice, it is frequently argued that because most of these tools are black boxes, they are inferior to human judges. The given considerations show, however, that reality is much more nuanced, as artificial black boxes communicate a wide range of plausible outcomes, in contrast to most human judges. The central theme of this Article is as follows: If judges are expected to be at least as transparent as artificial black boxes, they should communicate all legally valid solutions to a given dispute.

Publication year:2021

Accessibility:Open