Why Humans Want AI to Make the Call

I'm old enough to remember a time when professional tennis players spent much of the match complaining about the umpires. Hearing John McEnroe scream, "You Can Not Be Serious!" at yet another line call was as much a part of the match as the handshake at the net.

Now, of course, this pretty much does not happen. What changed was the introduction of electronic line-calling systems. Starting in 2022, for instance, the U.S. Open eliminated all human line judges in their matches; line calls are all done electronically. And while eliminating humans from the loop in a human game in theory seems off, in practice it has been....great? Players rarely get into arguments over line calls anymore (they find other things to be unhappy about, to be sure), and the general sense is that, in this element anyway, the matches are fairer. Why? Because the electronic line call eliminates the possibility of error from limited human perception and unconscious bias.

Traditionally, tennis players yelled at umpires because they understood that it could work--complaining about a particular call can increase the chances that you'll get the next one in your favor. This is true in many sports. In basketball, for example, players and coaches know that keeping up a constant dialogue with the referees—complaining about a call against you, complementing a favorable decision—gets results. "Working the refs" is a strategy that works because you're appealing to human emotions to help get a better outcome the next time. Electronic systems that feature precise and consistent measurements circumvent this, leaving us with outcomes that seem more fair.

With Rover, we see the opportunity to apply this same standard to elements of human conversation. Having a moderator that is consistent and precise in its actions can lower the temperature in contentious conversations. When all sides feel heard and fairly engaged, better conversations can take place.

It's a difficult problem, for sure. People bring all sorts of variables to a conversation that make it more complex than simply determining whether a tennis ball landed on one side of a line or the other. But there are basic elements of fairness, of courtesy, and of good faith that are clear and measurable that can be used to help us all learn more from and be good to each other, and having an impartial, fair judge direct this can help. In this way, AI can help us all be better humans.