The AI Crusher: Bryson DeChambeau on his experiments with AI and the future of sports performance

Like any star athlete or performer, golfer Bryson DeChambeau is constantly looking for ways to improve his game — even when it feels like he’s doing everything right.

Despite years of elite training and world-class coaching, DeChambeau discovered something surprising about human perception when he began experimenting with video-based AI analysis of his strokes.

"I could feel a certain way, right, but that perception doesn't always meet reality," he explained in a recent interview. What seemed like perfect technique was actually costing him strokes. And only AI could see why.

A big part of what makes DeChambeau stand out in the golfing world is his methodical and meticulous approach. He 3D-prints his own club heads, followed data to focus his practice on the driving range, and uses his physics degree to put his equipment to the test in new ways.

This inquisitive and experimental style has even earned DeChambeau the nickname “The Scientist.” Yet it might be equally appropriate to start calling him “The AI Engineer.”

DeChambeau's exploration of AI and budding enthusiasm for the technology began just over a year ago, in the run-up to the U.S. Open golf tournament. Since then, DeChambeau has begun to find so many applications for AI, they might soon outnumber the variety of his shaping shots and the outfits in his closet.

In seeking to build out all these capabilities, DeChambeau has decided to partner with Google Cloud to define what’s possible not only for athletic performance, but also for the business, social, and everyday enhancement that makes up his — and everyone’s — modern life.

“Google is one of the most important technology companies in the world, and I think they are the future of AI,” DeChambeau said. “And as someone very much interested in that future, especially as it intersects with sports, I was very excited to get to work with Google Cloud.”

A victory for technology

DeChambeau began exploring AI-powered video analysis of his swing starting in 2024, driven as much by curiosity as any desire to fix specific issues. What he discovered, when he began experimenting with tools from a startup called Sportsbox AI, wound up changing the arc of his career.

The way platforms like Sportsbox AI (a Google Cloud customer, coincidentally) and the one DeChambeau has been developing with Google Cloud work is, an AI vision model is trained on historic footage of ideal swings from a certain golfer. This helps establish a baseline swing, which generative AI can then interpret and give instant feedback on in direct, concise language.

In DeChambeau’s case, he and his coaches chose his U.S. Open-winning swing from 2020 and a few similar performances as their baseline. When DeChambeau began analyzing training and play sessions last year with AI, the team was able to identify microscopic deviations in his swing mechanics during more recent events.

This training and insights helped lead DeChambeau out of potential slumps and right back to top form — including the leaderboard at the 2024 U.S. Open. It all came down to a dramatic one stroke win over Rory

McIlroy, which solidified DeChambeau’s place at the top of the sport. During a post-tournament interview, DeChambeau even credited part of his win to the help he got from AI, which he had been using in the weeks and even hours leading up to the tournament at Pinehurst Resort in North Carolina.

“I'm actually making interpretations and iterating on my golf swing with the data that's presented to me, minutes before my tee time,” DeChambeau said. “Having that information allows me to play the best golf, because it gives me the sense of clarity of what's actually going on in my golf swing.”

When expert intuition meets advanced measurement

Another defining moment in DeChambeau’s AI handicapping came at Pebble Beach — not the tournament, but an AI event hosted by Google Cloud. The Google events team had developed its version of an AI-powered golfing advisor, and the team invited DeChambeau and his Crusher GC teammates to help tune the model and even provide sample swings for it.

Even for someone regularly pushing the boundaries of his sport, DeChambeau found himself mind-blown. “It’s amazing how personalized, how responsive, and how quick Gemini is to understand you, the individual golfer, your quirks and strengths and weaknesses,” DeChambeau said. “It’s literally a game-changer.”

This uniqueness is central to DeChambeau’s approach to AI, and it’s a helpful reminder that even world-class performers benefit from objective measurement that goes beyond their own experience, impressive as their past feats may be.

In any endeavour, from sports to business to even our personal life, success — especially continued success — is often defined by small, ongoing improvements. DeChambeau, like a growing number of athletes, is keenly aware of the benefit of marginal gains. And he wants to share that with the world, through his Google collaboration, finding ways to bring AI coaching, training, and innovation beyond the pros, to help scratch players and weekend duffers, too.

Just consider the small insights DeChambeau got from using AI, and the difference it made to his game. He discovered that his pelvis moved about an inch too far toward the target on his miss-to-the-right shots, creating side-bend that prevented proper hand release to square the clubface. This micro-movement, happening in milliseconds during his swing, was imperceptible to him, his coach, and traditional video analysis.

But it became indisputable as a hole-in-one when seen through the lens of AI.

Real-time optimization through superhuman vision

DeChambeau is now working with Google on developing a platform that can run on smartphones, so he — and, hopefully soon, any users — can analyze and adjust their swing in near-real-time out on the course. If something is feeling off, he can take a few swings recorded by a Gemini-powered app, and have feedback on his biomechanics in under a minute.

The technology uses computer vision to track key elements of DeChambeau’s swing, measuring subtle variations in his pelvis lift, chest rotation, and body positioning that would be nearly impossible to detect through traditional video analysis. The system employs deep learning algorithms with proprietary 2D and 3D models, tracking more than 30 key points on the body, club, and ball. It tracks the outcomes of a shot, as well, such as top of swing, impact, follow-through, and finish.

“Google, using their metrics, showcases to me exactly what I’m doing,” DeChambeau explained. "Minutes before my golf rounds, I’m using these tools to enhance my performance in real-time,” DeChambeau said. "It acts like a golf coach for me, that's what I’m using it for now, and it's allowed me to make iterations on my golf swing."

It’s become DeChambeau’s pocket coach — and he thinks we could all benefit from one.

“Do I believe that AI can change how golf is fundamentally coached? 100%,” DeChambeau said. “I think it's the future. I don’t think AI is going to take over the role of coaching, but I can see it being a supplement to coaches from here on out. It's a great tool in the toolbox.”

Through refinement, the technical team has been able to build a comprehensive biomechanics measurement system that can process complex movement data in real-time. The technology also integrates multiple data streams — including movement patterns, environmental factors, historical performance — creating comprehensive analysis through accessible hardware.

The system is also being adapted for the putting green, which requires a totally different motion, while still employing the same basic technology of analyzing a baseline against current shots. This methodology is why DeChambeau hopes such a system can bring benefits to other sports, and even support personalized health and medicine. The underlying technology is ultimately so adaptable.

"Google's platforms have been incredibly instrumental in helping me understand all the positions of my golf swing, the metrics and the biomechanics," DeChambeau said. "AI is the future,"

The future of performance visibility

DeChambeau’s real-time approach is a fundamental shift from the reactive coaching of the past, toward a goal of predictive optimization. It’s a paradigm shift we’re also beginning to see everywhere – from factories, to smart cities, to individuals – where edge-connected sensors and devices can detect issues as, or even before, they arise or become critical.

It’s a place where performance improvements happen before opportunities are missed rather than after bad results.

Such gains are naturally at home on the sports field, where AI tools are fundamentally changing how we understand and optimize performance (just look at the success of McClaren Racing, to take one of many examples). Instead of only relying on the assessment of experts and intuitive adjustments, AI provides additional data-driven recommendations.

DeChambeau believes there is a strong market for AI coaching tools already developing, especially as consumers grow more and more familiar with AI assistants.

“Eventually, you’ll be able to talk to your AI assistant 24/7 — any time you are thinking about your swing, any time you’re playing poorly, or playing well, you’ll be able to check in with that personalized assistant,” he said. “The AI will help steer you closer and closer to your optimal performance.”

The question now is the same one DeChambeau has successfully asked and answered: Are you ready to trust AI enough in those big game — or big life — moments? Are you ready to let it help you see what’s missing and confirm those hunches? Victory awaits those who can find the right answers.