When apps become too present

A few days ago I tried to see if I could remove Meta AI from WhatsApp.

Not because I had a strong opinion about the feature itself. I mostly just didn’t want it there. WhatsApp is still, in my mind, a simple messaging app; one of those products that I don’t really think about as “software” most of the time, even if of course it is. It feels closer to infrastructure, where family groups live, where people send voice notes, where friends organize things, where work sometimes leaks in.

Because of that, small changes to the interface feel more noticeable than they would in other products. I don’t open WhatsApp expecting discovery, recommendations, experiments, or a new surface to interact with. I open it because I want to reply to someone.

So my reaction to Meta AI was not really about AI. It was more about presence. I wanted to know whether I could make that part of the product disappear from my version of the app.

From what I could find, the answer was not really.

Defaults are not the problem

That is the detail that stayed with me. Not because I think every default is bad. Defaults are necessary. A product cannot ask the user to configure everything manually, and most people would not want that anyway. Good defaults are part of what makes software usable.

But “on by default” should not quietly become “impossible to remove”.

There is a difference between a product introducing a feature and a product deciding that the feature now permanently belongs in the environment we use every day. We may choose not to use it, but that does not mean we can choose not to live with it as part of the interface.

That distinction feels small, but I think it matters.

When habit systems start to pull

I had a similar thought with Duolingo, although in that case my relationship with the product is different. I recently wrote about reaching day 900 of learning Russian, and Duolingo was a big part of that. The app helped me build the habit. The short lessons, the streak, the reminders, the small daily push to come back: all of that worked on me.

I don’t want to pretend otherwise. Without those mechanics, I probably would have stopped much earlier.

At the same time, using Duolingo for a long period also made me notice the difference between being helped by a habit system and being pulled by it. Some days the streak represents real practice, other days it represents five minutes of doing the minimum because the number must not reset. The product still succeeded in bringing me back, but I’m not sure the success always means the same thing.

Useful can still be too present

This is not a criticism of streaks in general. I like seeing progress, and I know that consistency matters more than occasional intensity when learning something. But the same mechanics that help you show up can also make the product feel more emotionally present than you want it to be.

Duolingo is interesting exactly because the pressure often works, the reminders work, the streak works, the widget works, the owl looking worried as midnight approaches probably works too. From a product point of view, it makes sense: the app wants you to keep the habit alive, and it uses the interface around the app to make that habit harder to ignore.

I understand the reasoning. I also sometimes find it tiring.

A product can be useful and still feel too present.

Presence is not just notifications

Usually, when we talk about this kind of thing, we talk about notifications. That makes sense, because notifications are the most obvious way software interrupts us. But I don’t think notifications are the whole story since a product can become more present through the interface itself: a button that is always there, a badge that keeps asking for attention, a widget that changes mood during the day, a reminder inside the app, a new feature placed in a position that makes it difficult to ignore.

Choosing the level of pressure

I don’t think the solution is to remove it all. I don’t want software to become passive in every context. Sometimes I want a product to push me a little. If I am trying to learn a language, exercise, track expenses, or build any habit that depends on repetition, some pressure can help. I know this because I have benefited from it.

What I would like is more room to choose the level of that pressure.

Maybe I still want Duolingo to track my streak, but I don’t always want the emotional layer around it. Maybe I want WhatsApp to keep Meta AI available somewhere, but not make it part of the interface I see every day. Maybe I want reminders to be functional, but not dramatic. Some want more discovery. Some want the product to push them. Others just want the app to stay quiet until it is needed.

Most of us probably want different things at different times.

When capability takes attention

When we add something to a product, are we giving the user more capability, or are we taking a little more of their attention by default?

Sometimes the answer is both.

And when it is both, the user should probably have a say.