About a year ago I had an idea for a darts app. I put it in a drawer and I didn't really expect to open it again.
I was a new player at the time. Darts is one of the cheapest games going. You need a board, a set of darts and a wall you're allowed to make a hole in. Getting started costs very little. The barrier I found was actually taking darts outside the house, how suddenly slow score subtractions, figuring out a checkout route and falling foul of darting etiquette felt very real, very fast. You could feel the impatience like it was burning a hole into your back.
The sensible answer round my way would have been a club or an academy, except nearly all of them are set up for younger players (not a criticism, just a reality). Which is fair enough, that's where the sport comes from. It's just not a door that opens easily for someone with a job and a family, on a Tuesday, for an hour, assuming the hour survives the day.
So darts had to fit around my life. The apps I found did fit around my life, they just didn't make me any better. Most of them score for you. You throw, you tap, it does the subtraction. That's genuinely useful and I used it plenty. But an auto-scorer is a cheat code for the exact skill I was trying to build. It takes the arithmetic away instead of teaching it, and it says nothing at all about the bit that actually decides games. Which treble. Which setup dart. Which double you leave yourself, and where you end up when you miss it, which you will.
That was the idea. Something that teaches you the decision instead of doing it for you.
Why it went in the drawer.
Not because I doubted it. Because I had no clue where to start.
I've always kept half an eye on what's happening in tech. Keeping half an eye on it doesn't mean you can ship an Android app. The maths went like this. Learn to build it properly, which is a year of evenings I didn't have. Or pay somebody, which is thousands of pounds for a version of an idea nobody has ever tested. Both roads led to the same place, which was the drawer.
I don't think that's an unusual sum to do. I think it's the most common one there is, and it's why almost every idea you've ever had is still just an idea.
What changed.
Over the following months I watched that distance shrink to almost nothing.
Claude and ChatGPT stopped being clever autocomplete and started finishing actual pieces of work. You could set a proper challenge and get back working code, tests written against it, and something you could hold the same evening. Not a plan. Not a diagram. A thing that runs.
So I opened the drawer.
What I worked out pretty quickly is the part that gets lost in all the noise. AI is brilliant at making things. It has no opinion about whether they're worth making. It'll build what you asked for, fast, to a decent standard, and it'll do it just as cheerfully when what you asked for is wrong. It doesn't know the coaching tone is off. It doesn't know a feature is technically fine and completely joyless. It doesn't know when to stop.
That's the job it left me, and it turned out to be the whole job. I ended up as the director and the producer. AI was the dev team, the testers and the QA, a delivery team that never gets tired and never gets precious about its own work. I set the direction, read everything that came back, binned a lot of it, and changed course whenever what was in front of me wasn't what was in my head.
The failures didn't stop. If anything there were more of them. They just got cheap, and they turned up the same day I asked for them rather than three months later.
About the fear.
I get why people are worried about jobs and I'm not going to be glib about it. People's livelihoods aren't a talking point.
But there's another side to this and it's the side I've actually lived. For every job that changes, there's a very long list of things that were never going to exist at all. My app wasn't taking work off a development team. There was no development team. No budget, no timeline, no project. Just someone who plays darts badly and had a thought he'd given up on.
I think that's what's really happening here. It isn't mostly moving existing work around. It's dragging a huge pile of work into existence that was never viable before. All the small, specific, deeply cared about things that were never worth a business case but were always worth doing.
That last one is the one nobody warns you about. Once you can make anything quickly, the only thing standing between you and a pile of fast, plausible, mediocre software is your own taste and your willingness to throw things away. AI got rid of my excuses. It didn't get rid of my standards, and it won't hand you any.
Where it's got to.
The idea from the drawer is Dartistry and it's real. It teaches you the decision instead of doing your counting for you. Which treble, which setup dart, which double to leave yourself, and what a likely miss leaves you on. Android first, one-time purchase, currently in development for its first release.
Burnt & Toast Studios is the UK company I set up to put it out properly. The name isn't an accident. Anything made well usually gets made more than once, and the good version is the one that came out of the pile of burnt ones.
If you've got something in a drawer, that's the bit I'd want you to take away. Go and have another look at it. Whatever made you close it a year ago probably isn't true any more.
Dartistry, a strategy coach for 501 darts.
Not a scorer and not a throw tracker. It teaches the decision underneath the throw, and what it tells a new player and what it tells a good one are different answers, on purpose.