She was ten, and she had to do a front dive to pass Red Cross beginners.
It was her grandmother’s camp, a family place. I was teaching swimming that summer. She could swim perfectly well, she had everything else signed off, and it came down to one dive off the dock into water she had been in all week.
She would not do it.
We stayed after her session ended. Then we stayed after the next group left as well, and then it was just the two of us and the lake going flat the way it does late in the afternoon. I did not take her through the technique. She knew the technique. We sat on the end of it and talked about nothing much, and every so often she would get up and go to the edge and stop, and then she would sit back down and we would talk about nothing much again.
I do not remember how long. Hours. Long enough that people started wondering where we were.
Then she dived.
That is the whole story. I did not teach her anything. I outlasted the thing that was stopping her.
A few years later I was in Costa Rica with my best friend, in a place called Montezuma on the Nicoya Peninsula. Up in the forest there are pools and waterfalls, one above the other, and at the top there is a spot people jump from that is high enough to be a bad idea. There was also a small rock you could dive off safely into the pool.
Everyone drifted away and it went quiet, and my friend told me he had never done a front dive. He was a grown man. He had been carrying that around for however long you carry a thing like that.
So we stayed. Eventually he did it. Same dock, different continent.
I have been thinking about those two afternoons a lot this year, because I now spend most of my working life with AI tools, and the thing I keep running into is not a technical problem.
It is the edge of the dock.
People do not say "I don’t understand how the model works." They say "I can’t do that." They say it before they have tried, and they say it with a certainty they would never apply to anything else. Someone who taught themselves a spreadsheet in 1997, who has run projects and hired people and raised children, will look at a terminal and say I can’t do that, and mean it.
That is not ignorance. It is the same thing the ten year old had. It just does not look like it, because adults have learned to dress it up as a preference. I’m not technical. It’s not really for me. I’ll leave that to the young ones.
And the standard response is a tutorial. Here are the eleven steps. Here is a video, here is a course, here is a newsletter. All of which is the equivalent of standing on a dock explaining the mechanics of a front dive to a child who already knows the mechanics of a front dive.
What actually works is much worse for everyone’s diary. You stay. You let it take as long as it takes. You talk about something else. You do not count the attempts and you do not mention that it is getting late.
I have started doing this on purpose. When someone I know wants to try one of these tools I no longer send them a link. I sit with them while they do the first thing, and I make sure the first thing is small, and I say nothing at all when they hesitate. It is slow, it does not work in a room of forty people, and I have no idea how you would put it in a training budget.
But I have watched people who told me flatly that they could not do this go on and do it, and every time, the thing that changed was not information. It was that somebody stayed.
Which leaves me with a problem, because staying is the only method I have, and in forty years it has not got any quicker.
So I will ask instead. When you have brought someone past the point where they had decided they were not capable, at work or anywhere else, what actually did it?