You know that thing people do at the edge of a swimming pool?
One foot in. Straight back out.
“Shit, that’s cold.”
Then they stand there for five minutes, as if the pool might take the feedback on board.
I understand the logic. You want to get in, but you’d prefer to skip the unpleasant bit. So you hover. Maybe lower yourself down another step. Get your knees involved.
Now you’re cold AND having a terrible time.
Eventually, you jump.
For a few seconds, this feels like a serious mistake. Then you start swimming. Someone throws you a ball. You’re chatting, messing around, enjoying yourself.
You’ve forgotten about the temperature.
All that negotiating at the edge, and the bit you were worried about was over in seconds.
I think about this a lot with AI.
There are people standing at the edge, watching what other people are doing, trying to work out whether they should get in.
They’re interested. They can see something is happening. But they’re waiting to feel a bit more qualified.
“I’m not technical.”
“I wouldn’t know where to start.”
“What would I even build?”
Fair questions. Although I suspect some of the answers only become available once you’ve started.
One of my favourite uses of AI began with a parking ticket.
I gave an agent a screenshot of the letter and a voice note of me moaning about it. It helped put together the appeal and supporting evidence. A week later, the ticket was rescinded.
My contribution to this technological breakthrough was being annoyed.
But doing that changed what I noticed afterwards. Other irritating tasks started looking different. Filing company forms. Finding information. Navigating websites that seem personally offended by the idea of being useful.
I started asking: could an agent do this too?
That question became much easier to ask once I’d seen it work on something I actually cared about.
And this is the part I think people miss about experimentation.
You’re allowed to start with a small, slightly stupid thing.
Something for your own use. Something that saves you twenty minutes. Something that makes your partner say, “Oh, that’s handy.”
You might discover that nobody needs it. You might discover that the annoying bit you set out to fix is attached to a much bigger problem.
You might show it to a friend and hear, “Wait. Could you make one for me?”
Now you have something interesting to investigate.
It’s hard to plan that conversation in advance. You had to make the thing for the conversation to happen.
We put so much pressure on the starting point. The idea needs to be good. The timing needs to be right. Ideally, someone should confirm that we have the appropriate personality for entrepreneurship.
Meanwhile, a person with half a plan has made something and is finding out what happens.
I want to be more like that person.
Of course, an experiment can fail. I can waste an evening. I can make something that turns out to be completely useless.
I’ve also spent entire evenings scrolling on my phone, so let’s be honest about the competition.
Pick something you wish existed. Or something you’re tired of doing. Give yourself an evening to see how far you can get.
You can decide what it means afterwards.
For now, you’ve been standing there with one foot in for ages.
Jump in the f*cking pool.
- Garvey