Who can doubt that technology is supplanting human beings as an arbiter of our endeavors. Whether replacing baseball umpires calling balls and strikes, HR personnel ranking job candidates, or university professors grading students’ work, such developments are proliferating along with our embrace of artificial intelligence (AI).
Could the day be far off when AI supersedes corporeal judges? After all, an artificial judge would be incredibly fast, tireless, consistent and, it is said, free of human emotions and biases. Current US Supreme Court Chief Justice John Roberts has likened a good judge to an umpire—a neutral figure who does not let personal penchants shape decisions[1]—and some may contend that AI can (or soon will) fit that requirement better than humans.
Indeed, countries around the world are already experimenting with AI to help judges, and AI-assisted arbitration platforms are gaining prominence. Emboldened by this, let’s try an experiment that enlists Google’s NotebookLM to scrutinize the briefs in Wisconsin Supreme Court cases and appraise the persuasiveness of the parties’ arguments. Not only can NotebookLM read a set of briefs in just a few seconds, it then explains in considerable detail which party was most convincing.[2] Although these “verdicts” are based on the parties’ briefs alone (not oral arguments)—and bearing in mind the familiar cautions associated with AI output of any sort—it should be interesting to compare NotebookLM’s conclusions with the positions taken by each of the seven justices.[3][Continue Reading…]