Your phone is a casino, your notes app doesn't have to be.

For years I thought reaching for my phone was a personal failing. In February the European Commission put a name to it in a legal finding, and the name was not "weak willpower".

My phone charges in the hallway, not at my bedside. That began as a sleep thing, as sleep is something I've always struggled with. Turned out to be the most effective bit of technology management I have ever done, which is a fairly bleak admission from someone who builds apps for a living.

Wake at 3am, reach over, nothing there. You lie in the dark mildly annoyed about being awake and then, and this is the entire trick, you go back to sleep. Fetching the phone now means standing up, leaving a warm bed and coming back. Nobody should bother doing that, I certainly don't.

None of that is willpower and that is exactly the point, because who has willpower at 3am? Distance and discomfort works where discipline doesn't.

What took me an embarrassingly long time was working out that the reaching for the phone had never been a character flaw, it was the product doing exactly as designed.

Nobody builds a casino by accident

If you've ever been to a casino this will be all too familiar. A casino has no clocks on the walls and no windows worth noticing. The carpet is loud so you look up. The exit is somewhere behind you, past several things that would like a word with you. None of that is an aesthetic choice made by a designer who liked the look of it, it is all one decision, taken commercially, about how long you can be made to stay.

The bit that does the actual work is smaller and quieter than the architecture. It is the not knowing. Pull the lever on the slot machine and something might happen, and the maybe is doing more to you than any win ever could, because a reward you can predict stops being interesting and a reward you can't keeps you playing.

Your phone runs the same machine and it runs it better, because it's been watching you.

Pull to refresh is a lever. It feels like a lever, it's operated like a lever and it pays out on the schedule a lever pays out on, not always but sometimes. Infinite scroll is the room with no clock. Autoplay delivers the next one before you have finished the first, and the notification telling you that somebody you barely know has done something you do not care about is, and I am not stretching this for effect, the floor manager coming over because you had started drifting towards the exit.

The casino is honest about one thing. You know exactly where you are. You walked past a sign, through a door and you can walk back out through that same one.

The phone came home with you and it sleeps on your bedside table. It holds almost everything, from your contacts to your boarding passes and your bank details, and it never once announces which of those it's prioritising.

What the European Commission actually said

In February the Commission published preliminary findings against TikTok under the Digital Services Act, and the features it named were infinite scroll, autoplay, push notifications and highly personalised recommender systems. It found the company had not properly assessed how those could harm the physical and mental well-being of its users, minors and vulnerable adults included. Its words for what the design does to you were that it constantly rewards you with new content, and that it can put you into autopilot.

Autopilot. In a regulatory finding.

11 days later it opened proceedings against Shein over engagement-based design, the points-and-rewards sort. In July it made the same preliminary finding against Facebook and Instagram. And last year it fined 'X' €120 million, under the same regulation, over a different interface question entirely, so this is not an idle threat being made.

The research all of this rests on makes the casino comparison itself, and makes it without flinching. The damage is done by not knowing and it's the same not knowing a slot machine sells.

The enforcement may or may not change a single screen, I have no idea, and it isn't the part I keep contemplating. When these companies defend themselves here they reach for freedom to conduct a business, which is a perfectly respectable thing to reach for and also, if you sit with it for a second, a confession. That defence only works if the features and the revenue are the same object. Nobody is planning to fix the infinite scroll, it's the very thing being sold, and the advertising only pays for as long as you are still on the page.

Recapture notifications

Two years before any of that, the European Parliament passed a resolution listing the design patterns it was worried about, and buried in that list is a phrase I have not been able to put down since.

Recapture notifications.

Somebody sat down and named the category of message whose only job is to get you back. Not to tell you something. Not because anything happened that concerns you. To recapture you, like livestock, or an escapee.

You have had hundreds of them, so have I. "You have unread stories." "See what you missed." "Your streak is about to end." That last one is the purest example ever built, because a streak is a thing an app invents, awards to you, and then threatens to take away, and the entire mechanism runs on a loss you would not have been able to sustain if you had never been given the thing in the first place.

Which apps need any of this

The obvious villains are the least interesting part of this.

The machinery exists because your attention is the revenue. No ads equals no reason to want your evening. That is a clean line and it ought to sort the world into two tidy halves.

It doesn't though, and that is the bit worth sitting with. Open your app library and count the things that have no advertising business whatsoever and still have streaks, badges, a little red dot that appears for no reason, a weekly summary of how you are doing, a gentle note that you have not written anything in 3 days. Habit-tracking apps that shame you. Language apps with an owl. Notes apps, plenty of them, that have decided the thing your writing needed was a leaderboard.

None of that arrived through a business model. It arrived through copying. The patterns worked somewhere with ads and got carried across into places with none, by people building the kind of app they had seen, and a retention chart went up, and nobody stopped to ask what the number was actually measuring. An app you open because it nagged you is not an app you needed.

A tool for writing things down has no business having an opinion about how often you use it.

Where Catchlight sits, and what a subscription wants

Just to be straight about my own position, because it is easy to write at length about other people's incentives and quietly skip your own.

Catchlight has no feed. There is nothing in it to scroll except the things you wrote, and your own stuff runs out. There is no account and no email address, so there is nothing to log into and nothing for me to email you about. Every feature works with no network at all. The app has no analytics in it, none, which means I do not know how often you open it and I have deliberately built myself into a position where I never will.

The one thing it will interrupt you for is a reminder, and only ever a reminder you set yourself, about the thing you asked to be reminded about.

Now the bit I would rather not write. If you ignore one of those reminders it will follow up, up to 3 more times, at whatever snooze interval you chose. That is a nudge. I am not going to pretend it isn't one just because you asked for the first one. What I would say in its defence is that it is about your thing rather than my app, and that it stops, and that there is a switch for it under Reminders. What I will concede is that the switch is on when you get there, and a default is a decision I made on your behalf.

Catchlight costs money, which is the answer to how any of this gets paid for without selling your attention to somebody. A subscription comes with its own incentives, they are simply pointed somewhere else. It wants you to keep paying. The way that pressure shows up in my week is pressure to make the thing worth the money, which is a better pressure than the alternative, but I would be selling you something if I called it no pressure at all.

We are not launched yet. I am building this in the open and everything above is what has been built and how it behaves, not a promise about how it will behave later.

A casino has a door

Almost every app is useful for the 30 seconds you actually wanted it, so useful is a hopeless test and I have stopped using it.

The one I use now is whether the thing seems content to let me leave and whether anybody, at any point, drew a diagram of how to get me back tomorrow. You can usually feel the answer within a week, and it has almost nothing to do with how good the app is, which is the part that took me longest to accept, because some of the stickiest software I have ever used was also some of the best.

If you want the practical version of how I manage all this, none of it is mine and none of it is clever. The phone charges in another room. Most of the apps come off the home screen and live in the App Library, so getting to one means searching for it by name, and having to type the name is often enough to remind you that you did not really want it. Focus modes hide the rest while I work and the thing that actually holds it all together, the part every article on this subject skips, is that you have to put something in the space, because boredom will eventually beat any system you build.

There is more on the shape of the app itself in one Take instead of three apps and why your notes shouldn't need an account, the case for working offline first covers the network side, and the features page lists what is actually built. If you want to line the notes apps up against each other, the comparison pages do that job.

A casino at least has a door, with a sign over it, and a floor manager who is only human and gives up on you eventually.

Nothing on your phone gives up on you. Charge it in the hall.

---