GoKawiilSign up to see the future, today Can’t-miss innovations from the bleeding edge of science and tech Email address Sign Up Thank you!
While military brass and corporate bigwigs race to hand every decision to an AI chatbot, one major American institution is drawing a bright white chalkline in the dirt.
According to reporting by the New York Times’ Athletic, the use of AI in major league dugouts will no longer be tolerated, after concerns were raised that the tools were giving some teams an unfair advantage.
Speaking to the publication, anonymous baseball insiders estimated that roughly a third of all major league teams had already been using AI chatbots installed on dugout iPads to give advice on in-game strategy.
Per a memo sent out from the MLB commissioner’s office and viewed by the Athletic, baseball teams were using iPads to access AI chatbots, in a way that went “beyond their originally intended purpose to include recommendations regarding substitutions, pitch calling, and other in-game decisions traditionally made by players and coaches.”
It’s a major policy adjustment, as the MLB rarely issues sweeping rule changes mid-season, preferring instead to wait until the offseason. Arguably the most recent analogy came in early summer of 2021, when the league mandated in-game glove checks as part of its crackdown on foreign substances on the pitcher’s mound.
All told, the league is probably tilting at windmills. In lab settings, which lack the chaos and pressure of a real-world batting lineup, AI chatbots have flunked pretty much every test of sports analytics and prediction.
As one major study found in June, though AI can do basic tasks with some degree of accuracy — identifying which basketball players are on the court at a given time, for example — the tech is horrible at prediction and analytics tasks, like anticipating what the final score will be or accurately explaining why certain plays happened the way they did.
That being the case, it’s unclear exactly how much of an advantage AI gave ball clubs that used it. While ChatGPT could maybe predict the first pitch or two thrown during your next at-bat, or recommend which reliever to sub in when the fatigued left-handed starter loads the bases, it surely isn’t predicting full pitch sequences or making personnel suggestions with any meaningful accuracy.
Still, it’s probably just as well to see the tech leave the dug-out for good — a welcome sign that, in the MLB at least, human ingenuity is still the name of the game.
... continue reading