AI Is the new MVP.
Is generative AI ushering in Moneyball 2.0?
I lead an AI research group at Deloitte called the AI Braintrust (nod to Ed Catmull, founder of Pixar).
The Braintrust meets every two weeks and has a candid conversation on the power and pitfalls of AI. We share papers, demos, tools, experiences, and hypothesis centered around AI. We’ve gathered over 90 members and our numbers continue to climb.
A few weeks back, Mark Osis took the lead on our meeting and gave an AWESOME presentation on AI in Sports. Here were some of the most interesting takeaways:
Sports industry quick stats
Sports industry worth $500B
In 2020, the market value of AI in sports was $1.4B (Allied Market Research predicts this number jumps to $19.2B by 2030)
Whoop worth $3.6B
FitBit worth $2.1B
Oura worth $2.5B
Global Sports Coaching Platforms worth $1B
Highlighted use cases
Film Splicing: LLMs can automatically create highlight reels from sports events by identifying key moments. (Demo)
Game Planning/Strategy: Transformer-based AI can help coaches devise strategies based on historical data and opponent analysis.
Player Analysis: Generative AI can analyze player data, including video footage, to provide insights on performance. (Demo)
Scouting: LLMs can assist scouts by analyzing player statistics and reports. (Demo)
Coaching: Generative AI can simulate game scenarios and offer real-time feedback during practice.
Fitness Tracking: Transformer-based AI can monitor athletes' physical conditions using wearable data.
Diet Optimization: Generative AI can create personalized nutrition plans for athletes.
Structured Data from Unstructured Footage: LLMs can tag plays and events in unstructured game footage, creating structured data for analysis. (Demo)
AI-Powered Playbook Generation: Using Generative AI and LLMs to create dynamic, adaptive playbooks for sports teams that evolve based on real-time data and opponent analysis.
In-Game Realtime Strategy Adjustments: Real-time analysis of game data by AI algorithms to suggest optimal strategy adjustments to coaches during a match or game.
Dynamic Game Simulation: Utilizing Generative AI and LLMs to simulate various game scenarios and predict outcomes based on different strategies, player lineups, and conditions.
Real time commentary translation: Translate game commentary to any language in near real time (Demo)
Injury recovery: AI physical therapist / AI physical therapy copilot (Demo)
More to think about
Are we entering another Moneyball era with the advent of transformers and Generative AI?
What ethical considerations should be addressed as AI becomes more integrated into sports, including issues related to privacy, data usage, and fair competition?
Have you seen any interesting applications of AI in Sports?
Which sports are leading the AI race? How are they using it?
Bonus
In the beginning of September, Founders Inc. held a sports-focused AI hackathon. The projects that came out of the event provide a glimpse into the future of AI in Sports. Click here for links to the demo videos.
-Bennie iii