The Execution Gap: the distance between knowing what to do and being able to do it

In my initial military training, I learned to slowly manoeuvre down a rope instead of sliding. Except I didn’t.

I could say word for word what I had to do, but couldn’t act on the knowledge. I got to the edge of the platform, took a deep breath, and grabbed the rope.

As I wrapped my one leg around the rope and raised my other leg from the platform, I knew instantly that I was going to fall. I slid, feeling the intensity of the rope burn on the palms of my hands as I tried to slow my descent.

But I knew what to do - why was there a gap between my knowledge and acting on it?

In the space between knowing and doing, your body produces a mix of physical sensations, primary emotions, and hormones that rush to the brain and combine into memory-based thoughts.

This combination had me predisposed to a crash - I had never done this before, I was afraid of looking a fool, and I was convinced I couldn’t do it.

Eventually, I learned how to do it, and my hands started to recover.

So what has this got to do with humans and working with AI?

Let me take the metaphor of friction. The rope burns helped me recognise I wasn’t getting this right, but not what adjustments I had to make. Corrective advice didn’t make a difference; only my repeated attempts did. I recognised what it felt like when I had the right grip, when my feet were in the right place; I began to recognise it instantly when it felt right.

When we are learning something, we can go straight to Generative AI and ask for the answer. However, struggling with a problem, receiving uncomfortable feedback and staying with uncertainty are experiences through which we develop judgement, resilience and expertise.

The leadership challenge is not simply to remove friction. It is to decide which friction is pointless and which is preparing people for the moments when they will need to think and act for themselves.

This is one of the themes I tackle in my book