Five Months Later

In March I wrote a post on this site about trying an AI desktop tool. I said something in it that I'd known about myself for years: "I would build something, then abandon it after a few days because it felt too complicated." That was the honest part. I'd done it with budgeting apps, with training plans, with three different note systems. I'd build the thing, feel clever for a week, and then quietly stop opening it. That post is five months old today. I want to write down what happened after it, partly because I'd like to remember it, and partly because it went differently than usual.

WHAT I HAD IN MARCH

A weekly schedule that matched my actual shifts, and a morning dashboard — stock prices, the weather, the basketball scores, that week's to-do list. One screen instead of six apps.

WHAT I HAVE NOW

That dashboard is still running. It never got abandoned. It just kept growing: the schedule, a habit tracker, a weight chart, two trading bots that watch the market while I'm on shift, a push-up counter, a fasting timer. And it isn't only software any more. In April I weighed 98.7 kilos. This morning I weighed 84.0, and the only reason that number went up this week is that I ran nine kilometres on Monday and my body is holding water. Sixteen kilos, four months, no diet with a name. I started running in April. I could not run for two minutes. Yesterday I ran 9.19 km along the coast at Tenby with an average heart rate of 137. Between June and August I published six books on Amazon. Two of them I wrote — a memoir about what running did to me, and a dessert book with restaurant recipes scaled down for a home kitchen. The other four I made for other people to write in: a running log, a habit journal, and two recipe journals for the kitchen, which is where I actually live. And this weekend I built an app. Push-Up Pal — it turns your screen time into push-ups. On Saturday it was 166 lines of code. By Sunday night it was a working app. Today it is signed, installed on my phone, and on its way to the Play Store.

THE PART THAT ACTUALLY MATTERS

None of that is the point of this post. This is: I didn't abandon it. Five months, and the thing I built in March is still the thing I use every morning. That has never happened before. So I've thought about why, and I think it comes down to something I wrote in that same March post without understanding it properly: "Trust in a workflow comes from consistency and ease, not complexity or perfection." I meant it about software. It turned out to be true about everything. The weight came off because I weigh myself once, in the morning, and write down one number. The running stuck because the rule is "go out", not "follow the plan". The push-ups stuck because logging one is typing a single digit into a chat. The books got written in the hours around split shifts, because those were the hours I had, and pretending otherwise would have meant writing nothing.

Everything that survived is the thing that cost almost nothing to do. Everything I abandoned was the thing that asked me to become a different person first.

THE HONEST BIT

I'm not going to write a post that only has good numbers in it. Six books have been on Amazon for months. I have sold one. One copy, to a stranger, on the 20th of August — I got up at half five in the morning and looked at the sales page like it was a lottery ticket. Two pounds in royalties. I'd tell you it wasn't about the money, and it wasn't, but I still checked three times. So the honest scoreboard reads: the making works, the being-found doesn't. That's not a quality problem, it's a shop-window problem, and I know the difference now. That's what this website is for. It's why I spent this evening rewriting it — until today it still described someone who streamed games in 2020.

THE OTHER THING THAT CHANGED

The March post was called "challenged my trust issues", which at the time meant I didn't trust a piece of software to hold my week. Today I handed over an access key to my developer account and let an AI publish an application under my name, while I was at work. I'm not sure what to do with that except write it down.

IF YOU'RE READING THIS AT THE START OF SOMETHING

Start with the version that costs you nothing to keep.Not the good system — the one you'll still do on the day you feel like rubbish, after a nine-hour shift, when nobody would notice if you stopped. That's the one that's still there in five months.The rest builds itself on top.

Lehel Kemenes — chef, Wales.