IT OWES YOU NOTHING

Coaching, instinct, and why the most important things still can't be downloaded

I was in Stockholm this weekend. Not racing, just watching.

Standing at the side of the road as thousands of people moved through the city on marathon day, I found myself not necessarily watching the race but watching the people. The ones flying through at kilometre twenty, all efficiency and quiet confidence. The ones starting to hurt at thirty, jaw tightening, form beginning to unravel at the edges. The ones at thirty-five who have clearly had a conversation with themselves that nobody else was party to, and are still moving forward anyway.

My coach, Rob Foster, was beside me. As we walked along to hunt down a good spot to cheer our friends on, he said to me, "The marathon distance can't be taken for granted. It owes you nothing."

He was right. And watching those thousands of people work through something that no training plan, no data set, no algorithm can fully prepare you for, I kept thinking about a conversation that's been running through both coaching circles and our personal circles all week, how AI can be used and integrated into both coaching and performing.

I'll be honest, I use it quite a lot. It can be extremely useful. For session planning, research, working through ideas, and understanding training science, I might not have dug into deeply enough on my own. It can fill knowledge gaps that might otherwise take weeks of reading to close. Applied thoughtfully, it can make a good coach better informed.

But the thought often crosses my mind, what is it helping, and what is it quietly hurting?

Because the information was never the hard part.

What's hard is reading the person in front of you at eight o'clock on a Tuesday morning when they say they're fine and everything in their body language says they're not. Knowing when to push and when to ease off. When to say something and when to say nothing. When the plan needs to change not because the data says so, but because you can see it in their eyes and hear it in their voice.

Over-reliance on AI could erode a coach's observational skills and intuition over time. If you outsource enough of your thinking, you can slowly lose the instinct that took years to build. The feel for a session. The read of a room. The sense that something is off before anyone has said a word.

The best coaching moment I can think of is usually the simplest. A last-minute adjustment to a marathon plan, the right word at the right time, a message that just puts someone's mind at ease, sometimes nothing to do with training at all, but everything to the athlete on the other end of it.

This is where AI will struggle. Because it doesn't know your athlete, not really, not the way you do after months and years of showing up alongside them.

Standing in Stockholm, watching people move through something genuinely hard, I was struck by how human the whole thing was. Every one of those athletes had a reason they were there. Some were running toward something. Some were finding out what they were made of. And the coaches and supporters on the side of the road were watching, reading and feeling the moment. Trying to give the right boost, the right comment, the right cheer at exactly the moment it was needed.

That's not something you can automate.

So where does that leave AI in coaching and performing? I think the answer is genuinely positive….. If the foundation is right!

For coaches, use it to learn more. To challenge your thinking, pressure-test a plan, find research you might have missed. Use it as a thinking partner, not a replacement for thought. The coaches who will benefit most are those who already care deeply, observe closely, and understand that the athlete in front of them is a person first. For those coaches, AI adds a layer, but it doesn't replace what's underneath.

For athletes, the same applies. Use it to better understand your training, to ask questions you might feel awkward asking, and to explore the science behind what you're doing. There's something genuinely empowering about an athlete who takes ownership of their own development, who shows up having thought about it, having engaged. AI can be a tool for that. A way to become more informed, more curious, more invested in the process. But for both coach and athlete, if the foundation isn't there, no tool fills the gap.

A flashy programme means nothing to an athlete who doesn't feel seen. A perfectly periodised training block lands differently when the person delivering it has actually noticed you've had a hard week. And all the self-coaching in the world doesn't replace the moment a good coach says exactly the right thing at exactly the right time, not because an algorithm told them to, but because they were paying attention.

AI can give you better information. It can't give you better judgement. That still has to be developed, earned, and applied by the person in the room.

Rob's line stayed with me all weekend. Marathon distance owes you nothing. And I think both coaching and competing are the same. Trust isn't owed, and effort isn't guaranteed. Both are earned slowly, through presence, time, and showing up consistently enough that the other person knows you mean it.

The basics haven't changed. Care about the person, not just the performance and be there when it matters. Adapt when you need to, and use every tool available (AI included) in service of all of that.

The technology is a means. The relationship is the point.