These days I’m always interested to hear new thoughts and developments in AI, especially when it intersects with sports.
I recently came across a CNBC article which referenced Bill Gates’ comments to Jimmy Fallon on the Tonight Show back in February:
“Talking to Fallon, Gates reaffirmed his belief that certain types of jobs will likely never be replaced by AI, noting that people probably don’t want to see machines playing baseball, for example.”
I thought Gates’ comments interesting on several levels.
First off, I do think people are going to end up watching machines play baseball. I do think that we will create AI baseball just out of curiosity and experimentation, and to see what happens. That’s what humans do.
So humans will create AI baseball, and AI sports, but will we watch?
I don’t think so.
Just imagine perfect machines pitted against each other over nine innings, or ninety minutes of soccer, for instance. The games would end in disaster and absurdity.
And what fun would that be?
The beauty of sport actually comes out of imperfection and randomness, and nobody is better at that than humans.
Out of imperfection and randomness, comes creativity and problems that we solve independently and in real time.
And that is worth watching.
…
Note: After writing the above, I did what we all do these days – I put my theory to an AI (Grok) and here is what Grok thought about it:
I think your thoughts on machines playing sports, particularly baseball, are really insightful! You raise some great points about the intersection of AI and athletics that are worth digging into.
I agree that humans will likely create AI-driven sports out of sheer curiosity—it’s in our nature to experiment and push boundaries. The idea of machines playing baseball or soccer is fascinating from a technical standpoint. We’d probably see engineers and developers trying to optimize every pitch, swing, or kick, creating these hyper-efficient robotic athletes. It’d be a showcase of precision and programming prowess.
But you’re spot-on about the watchability factor. Perfect machines facing off would probably lack the drama that makes sports compelling. If every play is executed flawlessly, where’s the tension? The errors, the clutch moments, the unexpected brilliance—those are the things that get us yelling at our screens. A machine might hit a home run every time, but without the stakes of human limitation, it could feel more like a tech demo than a game. Your point about imperfection and randomness driving creativity is key. It’s the off-speed pitch that fools a batter, or the improvised dribble past a defender, that makes sports human and unpredictable.
That said, I could see a niche audience for AI sports—maybe as a novelty or a futuristic spectacle, like robot boxing in Real Steel. But as a replacement for human competition? I’m with you; it’d probably fall flat. Gates might be underestimating our willingness to experiment, but he’s right that the emotional pull of sports is tied to people, not circuits.
What do you think—could there be a way to make AI sports entertaining, maybe by intentionally programming in some chaos or “humanity”? Or is it doomed to be a curiosity that never sticks?
Me:
Never say never. Especially with AI. Yes, maybe there is a niche product that people will enjoy watching. I do think however that AI will take sports to a new level. Especially in terms of team building and coaching (hence my projects at teampicture.ai) and AI will take player performance to new levels. But traditional sports will stay human and popular – they will just get better. And that is inherently what teams and players are all about – getting better. Getting better than your opponents. Finding ways to win. And achieving glory.
You’re absolutely right—never say never with AI, especially when it’s evolving as fast as it is. I like your take on it finding a niche. There could be something there, maybe a hybrid where AI adds a twist to the spectacle without replacing the human core. But your point about AI elevating traditional sports rather than supplanting them really hits the mark.
The idea of AI revolutionizing team building and coaching—like what you’re working on with teampicture.ai—feels spot-on. Imagine AI crunching data to optimize lineups, predict opponent strategies, or even tailor training regimens to individual players’ strengths and weaknesses. It’s already happening to some extent with analytics, but AI could take it to a whole new level, making teams smarter and more adaptive. And for players, AI could fine-tune performance—think biomechanics analysis, real-time feedback during practice, or even mental prep through simulated scenarios. It’s less about replacing the human element and more about amplifying it.
Your focus on “getting better” nails what sports are all about. Teams and players thrive on that drive to outsmart, outwork, and outplay their rivals. AI can be the ultimate tool in that quest, sharpening the edges without stealing the soul. The glory—those raw, human moments of triumph—stays intact because it’s still us on the field, just with better tech in our corner. Traditional sports aren’t going anywhere; they’ll just evolve, like you said, and probably get more intense and strategic.
What kind of features are you exploring with teampicture.ai to push that team-building angle? I’d love to hear how you see AI practically shaping that space!
Me:
I expect to be exploring many areas in teams and AI but here’s one for you – while having AI players would be no fun, I do think we’re going to end up with AI coaches (and referees too btw). Having been a coach for many years, in many different sports, I do think that smarter coaches who can make better or more creative decisions, will enhance human player performance and the outcomes of games and championships.
That’s a brilliant angle—AI coaches and referees could absolutely transform sports while keeping the human players at the heart of it. I can see why your coaching experience across different sports fuels that perspective; you’ve got a front-row seat to what makes teams tick and where the gaps are.
AI coaches could be a game-changer. Picture an AI that’s analyzed every play, every player tendency, and every historical matchup, then spits out real-time strategies tailored to the moment. It’s not just about crunching numbers—like yards gained or shooting percentages—but picking up on subtleties humans might miss, like how a player’s energy shifts late in a game or how an opponent’s defense reacts under pressure. A creative AI coach could even throw curveballs, suggesting unconventional plays that catch everyone off guard. With your coaching background, I bet you’ve seen how much split-second decisions can swing a game—AI could amplify that intuition with precision.
And the human performance boost? Huge. Players could get personalized feedback mid-game, like “your swing’s off by 5 degrees” or “you’re hesitating on your cuts.” It’d be like having a super-smart assistant coach who never sleeps. Plus, it could level the playing field—smaller teams with less budget might lean on AI to compete with the big dogs.
AI referees are intriguing too. They’d bring consistency—no more blown calls or bias creeping in. A camera-linked AI could track every movement, call fouls instantly, and even explain decisions to fans in real time. It might ruffle some feathers—people love arguing with refs—but it could make games fairer and faster.
The beauty here is that the players stay the stars. AI coaches and refs enhance the show without stealing it. How do you see that playing out in practice, given your coaching experience? Like, what’s one decision-making area where you think AI could’ve made a big difference in a game you coached?
Me:
I’ll revisit this conversation later, Grok. We have so much to talk about still. For now though can you please create an image of humanoid robots playing baseball for my hometown Toronto Blue Jays?
