AI and sports – get ready for the rise of the centaurs

Are you ready to be a centaur?

Centaurs are, of course, mythical half-horse, half human creatures from Greek mythology. But they are also chess players who combine machine power and human creativity to play at perhaps the highest levels we’ve ever seen, often defeating the best machines and humans alike.

And they might just be the future of sports.

Embracing AI in the Top 10 Consumer Trends
In our Top 10 Consumer Trends 2017, it became apparently that many people are not only enamored with the idea of an AI assistant but with an AI boss, with many even embracing the idea of upgrading their minds with AI intelligence:

More people want an AI advisor at work (35 percent) than those who do not (24 percent). And even though more than two in five are against it, almost one in four would even like an AI as a leader of a company.

Still, if you can’t beat them, join them. A third would rather technologically enhance their own intelligence than rely on an AI assistant, and almost as many want to upload their minds to the internet and become AIs themselves.

AI and sports
To be honest, I have trouble imagining an AI boss (much less enhancing my brain with computer power) and I’m also proving very resistant to the idea of any kind of AI assistant. But then I think about sports, and suddenly, it all becomes a bit clearer and a lot less scary.

Think about how much data you could collect during a football match. Or a car race. Or a tennis match. And then what you could do with it.

There is a video of Jason Hoffman, the Head of Product Area Cloud Systems at Ericsson, in which he talks about how much data it would be possible for GE to collect from jet engines – it’s on the order of the entire enterprise storage market every 14 days.

I think about sports data in the same way. I think about watching 22 players in an American football game, in a blizzard of scripted yet chaotic and uncontrollable interaction. I think about playing the game – how I had to read about what five other players were doing on the other team while keeping track of my teammates as well. I think about the level of complexity, not to mention the geometry of some sports and the inherent drama and the violence or beauty or simplicity or complexity of others.  All this data will be collected. It will all get fed into AI that is like electricity and hosted in the cloud.

It’s like the analytics wave that has transformed the sports world already, but on steroids, so to speak. And as long as it leaves some room for chaos and the unexpected, I think it’s great.

Centaurs take over chess
Right now, if you search for sports and AI you get two main themes. The first is AI-powered play by play. The second is the use of AI, by athletes and coaches themselves, to maximize training and develop strategies and tell coaches exactly what player to put in exactly what situation.

This all happened to chess years ago, with its astronomical complexity but boundaries of pieces on a board. And right now you could argue that the best chess “players” in the world are often not machines or humans but the two working in conjunction. These centaurs use the brute force of the AI but leave open the creativity, the vision, the insight of humanity.

Former world chess champion Gary Kasparov wrote an article back in 2010 about the role of computers and chess. He was the first champion to lose to a computer and also the founder of “advanced chess,” which is what centaurs play. You might think Kasparov would resent the brute force of the computers. But he doesn’t. From the article:

The computer could project the consequences of each move we considered, pointing out possible outcomes and countermoves we might otherwise have missed. With that taken care of for us, we could concentrate on strategic planning instead of spending so much time on calculations. Human creativity was even more paramount under these conditions.

Now we’re looking at a world where more and more athletes will be hybrids of this kind – with AI intimately involved but the humans still with the final say. I can imagine this, especially in racing of all kinds and solo sports. It’s hard to imagine an AI fast enough to tell, say, a tennis player to hit a backhand in the heat of action but maybe if AI can be directly uploaded to our brains (as so many in the Top 10 Consumer Trends survey wanted), anything is possible.

AI as a service
To make this a reality, AI will truly have to be a utility to be plugged into. AI will have to be ubiquitous. Wired co-founder Kevin Kelly said in a 2014 article that AI will have to be “cheap, reliable, industrial-grade digital smartness running behind everything, and almost invisible except when it blinks off. This common utility will serve you as much IQ as you want but no more than you need.”

And how exactly do you make that happen? How will AI be embedded in our daily lives like our lamps, televisions and washing machines today?

Well, with the cloud.

This is from a blog post by our CTO, Ulf Ewaldsson:

When connected to the cloud, smart machines will be able to use the powerful computational, storage and communication resources of state-of-the-art data centers. Today’s intelligent software robotics systems can support repetitive administrative tasks with current development pushing toward advisory tasks. Cloudification shifts the capabilities of these systems into a new sphere that includes complex problem-solving and decision-making on a mass-market scale.

This will be part of a loop where machine learning – a subset of artificial intelligence – will drive advances in how hyperscale datacenters perform, which will naturally improve the performance of the cloud and AI. And, working with the utilities analogy, this kind of cloud and AI power will start to move away from the Super 7 cloud giants that design and build their own hardware and software and toward, first, broader enterprises – where we need to answer serious questions about security, governance, and scaling of data – and then to us all.

And when that happens, I hopefully won’t be all that old. Maybe I can tap into some AI, maximize my training, get some spot-on coaching, keep my knees from blowing out, and relive flashes of my glory days on the playing fields.

Yeah, maybe this whole AI thing isn’t so bad after all ...

Related:

• Live sports experience
• Ericsson ConsumerLab