Slate published my short story When We Were Patched on sports and artificial intelligence. In the piece, I explore how A.I. might impact officiating in sports — specifically a future version of tennis with room-temperature superconductivity and augmented reality assistants. It’s tempting to think that code can solve the imperfections of refereing through instant replay and virtual analysis. Officials sometimes make glaring, infuriating mistakes. But there are many times when they’re managing human behavior as much as enforcing the rules. And while algorithms may be executing code, that code was written by human beings using data sets that they often curate themselves. Which means A.I. can have inherent flaws.
I’ve also long been fascinated by the people drawn to officiating, which at the highest levels of sport tend to be a thankless job with meager pay compared to star athletes. This week Serena Williams lost in the U.S. Open tennis final to Naomi Osaka and accused the match official Carlos Ramos of bias. This had nothing to do with whether a ball landed inside or outside the court, which would be assisted by a Hawkeye assistant. It had everything to do with Williams’ behavior and the officiant’s interpretation of that behavior. Williams called the decisions sexist, since she believes male players get away with similar behavior with fewer penalties.
But it’s also worth considering the factors Carlos Ramos was mulling over: how the rules of tennis applied to Williams’ behavior; what her behavior meant for the “spirit of the game”; and whether her behavior was fair to her opponent, Osaka (whose childhood hero was ironically Williams herself). And then there’s the fact that this was probably an important match for Ramos and his entire family was likely watching, so he was both trying to save face and distinguish himself before his peers at the same time.
How do you program such considerations into code? It will be hard.
My short story was edited by Joey Eschrich at the Center for Science and Imagination, and owing to their unique format, there is an excellent “response” essay on Future Tense by algorthmic bias expert Jeanna Matthews, with Kinjal Dave.