The 2024 Summer Olympics in Paris, France closed with fireworks, interpretative dance, and a parachuting cameo from none other than Tom Cruise. What might 2028 have in store for us?
There are close to 180 projects identified within the International Olympic Committee (IOC) that could be looked into for AI activations and AI deployment, since the inception of the Olympic AI Agenda earlier this year. Among these ideas, they proposed tools that would use machine learning to provide athletes with instant performance insights, a software platform to automatically generate sports highlights, and an AI system to ensure judges and referees evaluate performances fairly and accurately.
To explore these concepts further, Juan Antonio Samaranch, Vice President of the International Olympic Committee, ventured to MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) for a fireside chat with Daniela Rus, Director of CSAIL and MIT EECS Professor.
“I’m here to try to understand emerging AI technology,” Samaranch said during his visit to the Stata Center. “We care about the Olympics, and it matters to all of us. To keep it alive, I’ve come here today.”
Samaranch highlighted three priorities for potential AI projects: the remits that affect athletes (like in injury prevention, assisting recovery, and training), impact engagement (like with fans, broadcasting, and the in-person viewing experience), and influence efficiency (such as internal legal processes and operational efficiency). These initial ideas ignited an intellectual fire among MIT researchers in attendance, who were carefully brainstorming a technology-rich future for the Olympics.
Chatting about a hot topic in computing
Rus and Samaranch first explored the idea of an “AI anchorman” — essentially, a digital guide that can take viewers from one event to the next amidst many viewing options and incorporate clear personalization according to each user's interests and background. The broadcasting system, as they imagine it, could use artificial intelligence to integrate streams with human presenters and keep more casual viewers engaged.
Having chaired the IOC Coordination Commission for the XXIV Olympic Winter Games Beijing 2022 (2018-2022), Samaranch also noted the operational upside and opportunity of AI for both viewers and organizers. It could help identify when specific athletes are about to perform, advertise where tickets are still available, and assist with planning food and transportation.
AI’s potential Olympic impact spurred the speakers to reflect on the lineage of both the technology and the event. Rus traced AI’s roots back to 1956 at MIT as a burgeoning, but relatively young field of science compared to physics and chemistry. Samaranch noted its game-changing potential, literally: “Our model of togetherness and understanding hasn’t changed since the first modern Olympics in 1896, but artificial intelligence could help us break out of today’s world of silos and share ideas.”
They agreed on AI’s potential to unite people from different backgrounds, helping tell unique stories and connect people who share similar passions across the sporting world. Still, there are two sides to the proverbial AI coin according to Samaranch: it could make competition fairer by giving less privileged athletes access to advanced training methods, information, and technology, or reinforce disadvantages if only the wealthy can use it.
But could artificial intelligence also impact sports in real-time? Samaranch envisions that AI could learn routines to help judges in gymnastics and figure skating score performances, making their verdicts more objective. “I’m not sure we’ll have automatic judges, though,” he cautioned.
Artificial intelligence could even help prevent injuries and accelerate recoveries via predictive analytics. By observing a basketball player’s landing mechanics, for instance, AI could predict safer movements. This assistance could come in the form of a virtual coach who monitors Olympians’ rehabilitation progress.
During the fireside chat, Samaranch also touched upon the new place that e-sports has in the world, advertising the first Olympic Esports Games in 2025. The events will include fully virtual games and others that merge physical activity, like golfing and rowing, with gaming and virtual processes.
Hearing from the audience
Questions during an audience Q&A session illuminated the many grey areas involved in incorporating advanced artificial intelligence into over 800 potential medal events across the Olympics and Paralympics.
For example, one audience member asked if humans and AI could collaborate in future events, much like how humans can play chess with the program AlphaGo. Samaranch teased that it wasn’t off the table, so long as fairness could be guaranteed.
MIT researcher Emelie Eldracher, a rower and silver medalist in the 2024 Summer Paralympics, sat in attendance and mentioned using chatbots to analyze teams’ patterns during their heats. Samaranch highlighted that while coaches can use AI, it’s an ongoing question within the IOC.
Data was another source of curiosity among attendees. The IOC confirmed they have both a library of moving images and a performance insights database. While neither is open-source yet, it would be interesting to explore if these insights could be shared with MIT researchers — and if Olympians embrace sporting wearables during their events, computer scientists will have even more data to study.
Before the fireside chat concluded, Samaranch was asked about having robots or drones compete against each other in races and other contests. While the IOC isn’t currently considering this proposal, they have over 180 projects that will benefit from a little MIT computer wizardry.
