Elise Hu opens by invoking the canonical names of sporting greatness — LeBron James, Simone Biles, Lionel Messi, Cristiano Ronaldo, Kylian Mbappé — and noting that with the World Cup approaching, the world's gaze is fixed on those who made it. But the episode's guest, sports scientist Richard Felton-Thomas, is interested in a different question: what about the talent we miss? A brief exchange establishes the central tension — talent exists everywhere, but visibility is the bottleneck. Hu frames what's coming as a meditation on AI, biomechanics, and a question she admits she hadn't considered before: what does AI have to do with sports equity?

Three distinct sponsor segments fill this break. Dell promotes its XPS laptop — powered by Series 3 Intel Core — with student pricing starting at $599 and broader deals at Dell.com/deals. Kohler positions its Veil Smart Toilet as a design-forward, technologically sophisticated bathroom fixture, leaning into 150+ years of brand heritage. Apple Card rounds out the break with its pitch for unlimited daily cashback on a titanium card accepted wherever Mastercard is accepted, directing listeners to apply via the Wallet app on iPhone.

Felton-Thomas begins by asking the audience to visualise sporting greatness — and notes that we inevitably picture athletes from a small subset of countries. But talent, he argues, precedes greatness: LeBron, Messi, and Biles were always talented. What made them famous was visibility and opportunity. Traditional scouting is the primary mechanism for that visibility, but it's structurally broken — each Premier League scout sees around 2,000 players per year while millions play the game. Geography, cost, and access bias who gets seen. And when athletes try to self-scout by uploading social media clips, they've simply replaced a human gatekeeper with a social media algorithm never designed for talent identification.

The founding story of AI Scout begins in Felton-Thomas's biomechanics lab, when Darren Perry brought in his son Reeve for injury analysis. Watching all that sophisticated equipment in action, Perry asked the pivotal question: what if we could put these lab protocols — all the data, all the equipment — into standardised smartphone drills so any kid anywhere in the world could be tested fairly and equitably? For Felton-Thomas, it was a vision that matched a genuine problem. He joined Perry's company, AI.io, which builds AI-based solutions across sport, and the first product they built was AI Scout.

AI Scout's mechanics are elegantly simple on the surface. A child downloads the app for free, records themselves performing standardised drills — sprints, counter-movement jumps, dribbling, passing, shooting — and the footage is uploaded to the cloud. There, computer vision AI analyzes 22 key body segments, extracting data on direction, turning speed, jump height, symmetry, and coordination. The system can even convert standard 2D smartphone footage into inferred 3D movement data, giving it lab-grade analytical depth without any specialist hardware. The heavy lifting happens in the cloud, meaning the phone itself needs no special capabilities — a basic smartphone is sufficient.

To make AI Scout useful rather than just impressive, AI.io went directly to Chelsea FC and Burnley FC scouts and asked a simple question: if we could give you football-specific metrics from a smartphone analysis of any kid in the world, what would you need? The answer was clear: comparable, benchmarkable, reliable data — with full transparency about where it came from, no ambiguity. So the team built predefined drills that mirrored what scouts already assessed. But the harder challenge was scoring. When scouts watched video pairs and chose a preferred player, they couldn't explain why — the art of scouting, it turned out, is largely intuitive. The AI.io team had to sit with scouts, draw out their reasoning, and encode that intuition into a scoring algorithm, then pump thousands of videos through it to build benchmarks across age and gender.

The proof-of-concept moment came early. AI.io recruited 50 college footballers in the UK to test the app — a fairly average group, except for one: Ben, aged 17, who was head and shoulders above everyone else in the drill data. The question hit hard: how had this guy never been scouted? The answer was simply that the system had never looked at him. He wasn't in a remote village — he lived minutes from Chelsea FC's training ground, one of the world's best academies. AI Scout flagged him, he got a trial, scored on his Chelsea under-18 debut, later signed for another Premier League club, and went on to represent his country. It was the moment the team knew what they were building was real.

The India deployment pushed AI Scout into genuinely new territory. Reliance Foundation's existing programme sent scouts across India each year to find the best 11-year-old talent and award 5-year scholarships covering sport and free education. Same structural problem as Chelsea: a few scouts, potentially millions of eligible kids. AI Scout distributed through WhatsApp meant parents and students in hard-to-reach locations could download the app and trial directly. Tens of thousands of children now do this every year, and the highest-scoring performers are invited to in-person selection days where scouts make the final call. The standout story: one player who downloaded the app on a shared community phone, had never played organised sport, and earned a scholarship entirely on the strength of his movement data.

When the IOC flagged that Senegal's national teams didn't have enough athletes to fill Youth Olympics rosters, they turned to AI Scout — but this deployment looked different. Instead of trialing for one sport, school teachers and military leaders simply recorded children in their classes on tablets. AI Scout then analysed each child's movement profile and matched them to the sport where they'd most likely excel — great acceleration and reactive strength might mean rugby sevens or futsal; upper body power and hand-eye coordination might mean baseball or softball. From thousands of kids assessed over just a few days, 40 are now training for the Youth Olympics in wrestling, athletics, and football. It was a live demonstration that the technology works in reverse: the right sport can find you.

Felton-Thomas closes by mapping the road ahead. In the US, MLS Next has rolled out AI Scout to 45,000 youth players who are assessed three times per year — preseason, midseason, and postseason — giving coaches a longitudinal performance dataset delivered via a fully transparent control centre. Globally, multi-language support and multi-cloud architecture are under development so the app can be deployed in any country's preferred infrastructure. The underlying movement primitives — the cut, the deceleration, the jump, the throw — translate well across sports, so movement libraries for American football, basketball, baseball, and cricket are in progress. And the same computer vision engine is being explored for at-home healthcare and medical applications. The final message: talent is universal, brilliance is everywhere, and with a smartphone, we can finally make it visible.

Host Elise Hu closes by noting the talk was delivered by Richard Felton-Thomas at TED Sports Indianapolis in 2025 and originally published in November of that year. She directs curious listeners to ted.com/curationguidelines and credits the production team including Martha Estefanos, Oliver Friedman, Lucy Little, Emma Taubner, and Tansyka Sungmarnival, with additional support from several named contributors. Listeners are directed to podcasts.ted.com, and Hu signs off with her usual promise of a fresh idea tomorrow.