Unless you’ve been off holidaying on the surface of the sun for the last few months, you’ll probably have noticed that the hot topic right now seems to be artificial intelligence (AI). We’re either all going to be out of a job shortly or obliterated by machines that really don’t see the value in a load of spandex-clad humanoids. As we hide behind our sofas terrified at the prospect of Arnold Schwarzenegger knocking on our doors asking if we’re Sarah Connor, it’s pretty clear that AI is going to change everything, including how we approach sport...

AI as a training tool

First, we need to understand what AI actually is and what it can do. The Oxford Dictionary defines AI as:

The capacity of computers or other machines to exhibit or simulate intelligent behaviour.

Artificial intelligence is generally cited to have emerged as the ultimate destination of computer science back in the 1950’s, but the problem was that the technology wasn’t there to support the concept. However, a few decades of technological progression and the machine finally got the better of the man in 1997 when chess world champion Gary Kasparov was beaten by the ‘Deep Blue’ AI.

Broadly speaking, there’s generative AI or predictive AI. The latter can derive a response by processing large amounts of complex data or (and probably the thing that worries people the most) it can also self-learn. Generative AI is essentially what schoolkids may well be using to shortcut their homework, thanks to packages like Chat GPT. You can ask the AI basic questions and it generates a response based upon information it can access.

As an experiment, I asked ChatGPT to design some training programmes and it actually did a decent job of it. It incorporated the specificity and the progression you’d expect to see in generic training plans, and it did it for me in around five seconds. You need to be wary though as it has a curious tendency to make up fictitious references that don’t actually exist. However, for the recreational athlete looking for answers, this may well be your most powerful tool to help your performance since the invention of the internet.

Image Credit: Chat GPT 4.0... 🙄 ©

Then we have AI’s ability to handle what is known as ‘big data’ which can be used for generative or predictive purposes. I’m reasonably well-educated so I feel that, given a suitably large enough packet of chocolate Hobnobs, I can analyse relatively complex scientific data. The only problem is that it’ll take me weeks and a considerable volume of swear words, whereas AI does this in seconds and without a biscuit in sight.

In terms of performance analysis, AI will be increasingly able to handle the ability to track and analyse the movements of athletes in team sports or in races using video, GPS, and wearable sensors. It has already been used this way in sport. For example, this will help staff in the team cars of the Tour de France to influence their tactics or in your case, maybe change your approach when racing. It could analyse your power, heart rate, calorie consumption, hydration status, the weather, the course, and tell you when to push on or maybe when to hold back.

AI for analysis

Alternatively, AI could be used to analyse an athlete's biomechanics to identify areas for improvement in technique or reduce their risk of injury. This is important because I know from my own experience of working with athletes at the Paralympic Games that much of what is prescribed in terms of equipment or physical rehabilitation is often seen as a ‘black art’ and is based upon the medical practitioner’s unique intuition, which isn’t quantifiable.

The funny thing is that I wrote an article on Big Data in sport 10 years ago and it’s embarrassing  now that I didn’t even mention AI at all. This only shows how fast things are moving and how blindsided we’ve been by its use in sport.

AI could also be used for athlete recruitment by analysing a huge raft of metrics quickly to determine which athletes are not only physically superior, but could include multifaceted behaviour such as social media presence and overall value to sponsors. It would take what the Oakland A’s did in Michael Lewis’ book ‘Moneyball’, but perform a full analysis of the annual draft in seconds.

What’s going to be interesting though is when teams and managers start weaponising it so that you’ve got AI systems competing against each other. That’s going to be like watching the final part of the 80’s classic film ‘War Games’, whereby we avoid nuclear Armageddon by forcing an AI system to play Tic Tac Toe against itself at an insane speed.

Ethical concerns

However, the threat with emerging technologies is that the ethical concerns don’t get discussed until long after that technology has been introduced. Consider this, we’re already engaging with basic communicative interfaces such as Amazon’s Alexa or Apple’s Siri. If you couple that with AI’s ability to generate and process detailed information, would you be willing to be coached by an AI entity? One that you could talk to or converse with, one that’s perfectly attuned to what you need in terms of personality, response and appearance.

There are wellbeing and ethical issues here that sport hasn’t even begun to address. Interestingly too, this isn’t science fiction, its science fact. AI-based coaching systems, like Athletica, Dairus or SportsAI, have already begun to emerge now.

But what does this all mean for you? The take home message here is that to get the best from consumer AI, you need to avoid ‘GIGO’ (garbage in, garbage out). In other words, most current AI systems are only as good as the data that is fed into them or accessed. This is the same whether it generates a response from data or is self-learning.

Keep your devices well serviced, calibrated and crucially, use such technologies as a guide to your own decision-making, but don’t let them be a rod up your own back or a rabbit hole you go down. Whether its sport or military defence, the consistent message seems to be that human oversight should always be in place. AI is a tool.

Ultimately, you’re going to be faced with a choice – to embrace the sheer volume of data you collect when you exercise and hopefully profit from, or to reject it, unplug and embrace nature.

Just like the article I wrote 10 years ago, I’ve probably got little understanding of what systems and approaches will govern our own performances 10 years from now, when I probably write about this again. In fact, I’ll probably get AI to write it for me whilst I go for a swim. Oh, and by the way, rest assured that my constant use of pop culture references here was to reassure you that I didn’t have Chat GPT write all of this for me whilst I went off for a lovely ride out in the woods... as tempting as that was.

Further reading