It’s a knockout for AI

How one company is building the future of combat sports

Greetings, fellow humans!

Last week, I came across a great use case for AI, and I love it!

Therefore, in today’s edition of the Practical AI newsletter, I will discuss how AI will shape the future of combat sports.

You might even say it’s a knockout for AI!

Jabbr, the AI that never misses a punch

Jabbr is a Danish startup that has set out to change the future of combat sports.

They have developed the world’s first Computer Vision AI specifically designed for combat sports called DeepStrike.

A striking name, I must say…

Many reading this newsletter will not care much about an AI that does “something with boxing.”

But hear me out.

Combat sports are big business!

For example, according to Nielsen Sports, MMA - Mixed Martial Arts - is the third most popular sport behind soccer and basketball.

While DeepStrike AI is being tested and rolled out for boxing, it can also be trained and used with MMA and all other combat sports like Karate or Judo.

And that’s kind of a biggie.

But why is this interesting?

The features of DeepStrike

Let’s first discuss the features of this innovative AI.

Keep in mind that, for now, it is only working with boxing. But this will be applicable to any combat sports in the future.

In short, DeepStrike will give you punch statistics and analytics.

DeepStrike will measure 50 parameters for each boxer, including punch types, quality, footwork, and aggression.

The AI works on any input video in any setting (professional, amateur, sparring, etc.).

It works on any camera or smartphone, handheld or stationary.

And since it works on basically anything, you can easily convert a couple of smartphones into your own virtual video production crew with tracking, round highlights, stats, and more.

It can be used on old recorded videos like “The Fight” (Mohammad Ali vs Joe Frazier), your own training videos, and in real-time games.

Jabbr has a great video playing on their home page showing multiple uses (, or you can see some great shots on their Twitter, like this one:

Or if you care to see a full fight and how it tracks the punches etc, this is a great video:

In short, this tool has enormous potential, from your own training, to live streaming (entertainment), to even use by betting companies.

And if you have ever done any sports at a level beyond the hobby level, you know tracking progress and studying your technique is important for your improvement. Now, you can easily use your smartphone and DeepStrike to analyze your training, track your progress, etc.

I think this is an excellent use of AI, and the application of this is far-reaching!

It is a unanimous decision from me; this is a winner!  

What are the benefits of AI in combat sports?

I have already alluded to one of the benefits this AI can have on combat sports: tracking your training and progress.

One of the greatest athletes of all time, Michael Jordan, is famous for his hard work and perfecting his game. He said, “If you do the work, you get rewarded. There are no shortcuts in life.

While there are no shortcuts to success, utilizing DeepStrike will benefit you more than your trainer alone.

It is hard to hide my enthusiasm as I believe DeepStrike is transformative in combat sports. If I were a boxer, I sure as h*ll would want this!

Would’nt you?

Heck, even if you are not training to become a pro, this AI will improve your fighting skills and overall training results.

What else?

Broadcasters will love this tool. Pairing real-time objective stats with a boxing commentator will heighten the experience for those watching!

Sure, the red and blue boxes will be disturbing, but I suspect those can be removed (or will be a feature broadcasters can access).

And for odds setters, they will be able to adjust their odds throughout the fight with real-time stats.

But what about cheating?

There has been a lot of news coverage on cheating in the last year. From cheating in chess to cheating in fishing competitions by adding metal inside the catch...

But most professional sports are plagued with cheating.

Boxing is no exception.

From putting plaster of paris inside the gloves that harden when the boxer sweats to Mike Tyson famously biting off Evander Holyfield's ear, there is a long list of cheats that have been attempted in boxing.

Having an AI that analyses 50 factors in real-time will make cheating a thing of the past.

If we go back to one of the greatest boxers of all time, Iron Mike, he was notorious for being a “dirty boxer.”

An AI would be able to pick up on his tricks.

Tricks many boxers use as well.

Mike Tyson was famous for using his elbows. He would throw a punch, and if he missed (or missed on purpose), he would follow through and use his elbow “by accident.”

Evander Holyfield was famous not only for losing his ear to Mike Tyson but also for headbutting. Headbutting can cause some real damage to the opponent's face.

Other tricks are low-blows (punches to the groin), Rabbit Punching (punching to the back of the head = extremely dangerous as it can damage the spinal cord), and more.

The point is that a system like DeepStrike can pick up cheaters in real-time, and the judges will be able to act accordingly.

Lastly, with an objective AI counting and analyzing, paying off judges will be impossible.

I assume boxers throwing a game would be caught by the system as well (at least would be able to indicate if it happens).

Final thoughts

There are many use cases for AI in sports, competitive or otherwise.

The case of Jabbr is, in my opinion, an excellent use of technology and AI for good.

It elevates the training and progress for the athletes and enhances the experience for the spectators, and it will help eliminate the bad from sports.

If that is not a good use of AI, then I don’t know what is…

A lot has happened in AI in the last week, from Apple entering the AI arena (one last pun for the road), to ChatGPT becoming less accurate, or the growth of ChatGPT slowing down.

I will not go into those or other AI news, at least not in today’s issue.

However, I want to mention that I have decided to expand this “AI stuff” and will post weekly videos on YouTube.

The videos will be about practical use cases of AI, typically how-to guides.

In last week’s newsletter, I shared a video about the millionaire who could have saved $100,000 if he had access to an AI like Claude 2.

My first YouTube video was basically an updated and edited version of that video but with a real-life example of how the AI caught changes in an example contract.

If you did not watch the previous video (all 893 of you - I see you), here is your chance to see my chubby face once more...

Here is my first YouTube video (a work in progress, I know):

That’s all for today.

I hope you enjoyed this one!

As I said last week, please send me questions you have about AI, and I might be able to answer them in this newsletter (or possibly by email). Just reply to this email!

Have a great weekend!

AI’ll be back!

- Thomas

Why don't boxers have time for small talk?


Because they prefer to get straight to the punch!

(One last punny boxing joke courtesy of ChatGPT / GPT-4)


or to participate.