Three ways AI is changing the 2024 Olympics for athletes and fans
From training to broadcasting, artificial intelligence will have an imprint on this year’s event for the first time
The Paris Olympics will feature several elements that have been influenced by AI, including an app to guide athletes.Credit: Chesnot/Getty
When more than 10,000 athletes from around 200 countries gather in Paris for the 2024 Summer Olympic Games this week, they will have an all-new friendly — but faceless — voice to greet and guide them.
How do I reach my sporting venue? Can I livestream the opening ceremony? Will there be a computer refereeing my games? Where can I get freebies from sponsors?
These are just some of the questions that athletes will be able to ask AthleteGPT, an artificial intelligence (AI) chatbot designed for them, accessible through the Athlete365 mobile app. It’ll be able to scour through “thousands of information pages very quickly, and be available 24/7 to answer questions”, says Todd Harple, the Olympics AI Innovation programme lead at Intel Labs in Hillsboro, Oregon, who is involved with the effort.
The chatbot — a large language model (LLM) built using an AI developed by Paris-based company Mistral AI and Intel’s Gaudi processors — is just one way in which AI is leaving its imprint on this year’s Olympics, which begin on 26 July. Few would have been familiar with LLMs or heard of ChatGPT during the last summer games in Tokyo in 2021. But sprinters in Paris can only hope to match the strides that AI technologies have made since.
The International Olympics Committee (IOC) is embracing the technology. In April, it rolled out its AI agenda — an effort to streamline the explosive growth of AI research in sport and to strategize its use in the Olympics. “We have to be leaders of change, and not the object of change,” said Thomas Bach, the president of the IOC in Lausanne, Switzerland, at a press event in London, which showcased the capabilities of various AI sports tools.
Nature explores three ways in which AI is changing how athletes and spectators will experience the Olympic Games.
Athlete performance and training
As early as 1900, when Paris first hosted the Olympic Games, French scientist Étienne-Jules Marey was pioneering the use of technology to study athletes in motion. His high-speed chronophotography — which involved rigging a camera like a machine gun, feeding it photographic plates like ammunition to rapidly capture frames — snapped sprinters and long-jumpers. He analysed the body’s biomechanics to “discover the secret of superiority of certain athletes”, a Nature editorial commented in 1901.
Today, it is possible to do much more just by recording with a smartphone. Intel’s 3D athlete tracking (3DAT) technology uses AI to track 21 points across the human body to render its precise physical movement, providing “all the biomechanical insights that coaches look for” in elite athletes, Harple says. He thinks that such technologies will lead to closer competition and new records.
The ways in which AI is being used to enhance athletes’ performance range from designing custom-built athletic shoes and clothing to determining optimum nutrition and training schedules. “It may even accelerate our discovery of new strategies of playing sports,” he says. A historic example of such a fundamental change is the Fosbury flop — now the dominant style of high jump pioneered by US athlete Dick Fosbury at the 1968 Olympics.
Capturing photos and videos of movements and analysing the data is one way that athletes can enhance their performance with AI.Credit: Hector Vivas/Getty
The ease of collecting individual data, combined with AI analysis, could also help coaches to identify talent, making sports more equitable. In March, the IOC piloted a scouting programme that used 3DAT to identify more than 40 children in Senegal who showed promise in becoming Olympic athletes, by analysing simplerills such as running and jumping.
But sports and nations with big professional leagues will retain a big advantage, because they have the resources to gather high-quality data, and to train algorithms with them. “The problem with some of the Olympic sports is that there is not a big footprint of data,” says Patrick Lucey, chief scientist at the sports-technology company Stats Perform in Chicago, Illinois. This extends how the technology can be used in other aspects of these games, such as judging and officiating.
Refereeing and real-time data
Olympics water polo referee Frank Ohme is no stranger to AI. His day job as an astrophysicist at the Max Planck Institute for Gravitational Physics in Heidelberg, Germany, involves hunting for signals of colliding black holes — sometimes with AI’s help — in noisy gravitational-wave data. But when he dons the all-white referee uniform in Paris, he’ll need to peer through splashing water waves to decide whether the ball has crossed the line into the goal.
AI is already informing such decisions in sports such as football, using information recorded by an array of cameras around the stadium and chips implanted in the ball. But other sports are yet to catch on — and AI will probably be slower to pervade areas such as refereeing, which require real-time data analysis.
Another hurdle is funding, and the specific needs of each sport — there will be 32 events at the Paris games. Despite water polo being the oldest Olympic team sport, there’s not nearly as much money in it as there is in basketball or football, says Ohme. Using AI in water polo would also present different challenges, such as training algorithms on images taken underwater and in chaotic scenarios, he adds.
Precise and open communication is key whenever AI assistance is used to make calls in real-time. “The easiest way to convince teams and spectators is to give them all the information through an image or visualization where they can determine [the call] themselves,” Ohme says.
It’s also hard to remove ambiguity in actions such as fouls in contact sports. These are split-second decisions that not even all humans can agree on. “I wouldn’t even know how to start putting those into numbers,” says Ohme, who thinks that detecting black holes is an easier task for AI in comparison.
Enhancing viewer experience
The torrents of data collected during the games will not only feed AI algorithms, but also television viewers hungry for statistics. “Sport is its own language. It crosses barriers to help everyone communicate,” Lucey says. Statistics and numbers enrich these conversations by providing extra benchmarks for comparison. “Of course people want that,” he adds.
Broadcasters are rushing to find ways to augment this new wealth of information and put it onto television screens. Viewers were enthralled when the virtual world-record line was superimposed onto the screen for television viewers during the Sydney 2000 games. In 2024, broadcasters have the ability to display much more, such as acceleration, top speeds and stride lengths, says Harple.
What excites Harple most is the prospect of personalized highlights made available to viewers through Intel’s Geti computer-vision AI platform, which could be a feature of future broadcasts. With so much sporting action being recorded simultaneously, Harple says that the ability of AI to pick out exactly what viewers want to see will be a game-changer. This could be particularly beneficial for coaches and broadcasters from nations that have more limited access to production resources. “If someone wants every three-point shot made by the Nigerian men’s basketball team, AI can go through all the footage and automatically put them together,” he says.
doi: https://doi.org/10.1038/d41586-024-02427-0
This story originally appeared on: Nature - Author:Sumeet Kulkarni
