The psychology of sinking: why forgetting is a feature

Most productivity tools fight forgetting. Reminders, notifications, pinned items and starred messages all assume that forgetting is the enemy, and that if something slips your mind the system has failed.

Gravity assumes the opposite. Notes sink over time. If you don't rescue them, they drift towards the bottom of the stream and eventually out of sight. That can look like a flaw, and it's the part of the design I'd defend hardest.

The forgetting curve

In 1885 Hermann Ebbinghaus published the first careful measurements of forgetting. He memorised lists of nonsense syllables and tested himself at intervals. Most of what he learned was gone within a day, and the loss slowed after that. His forgetting curve has held up: a 2015 replication by Jaap Murre and Joeri Dros found much the same shape.

Nonsense syllables are a worst case, since they mean nothing, and you keep far more of material that connects to things you already know. The general pattern still holds. Unless something reinforces it, most of what passes through your head fades.

The usual response is to fight that with flashcards, spaced repetition and ever more insistent notifications. For some things, like learning a language, that's the right call. For the daily stream of thoughts, fighting forgetting costs you. If you remembered everything with equal weight, the few things that matter would be buried under everything that doesn't.

Forgetting is one of the ways your brain filters. Things that aren't reinforced fade. Things that matter tend to get reinforced, because you meet them again, they carry feeling, or you go back to them on purpose.

Gravity's sinking works the same way. It goes along with the filtering your memory already does.

What happens when you review

You scroll through your stream. Most notes you skim and pass. They mattered when you wrote them, and now they're just there.

Then something catches your eye: an idea you'd forgotten, a to-do that suddenly feels urgent, a quote that reads differently than it did last week. You swipe right and it goes back to the top.

The notes you skipped got no reinforcement, in the app or in your head. The one you rescued got some. You noticed it, recognised it mattered and acted on it, and that sort of retrieval is one of the more reliable ways to make something stick.

You didn't decide in advance what was important. Your attention decided during review, and the rescue gesture recorded the decision.

What if I forget something important?

This is the fear underneath most note-taking systems: what if I lose something I'll need?

It helps to be specific about what "important" means. If something matters, one of two things usually happens:

1. Something outside you brings it back. A deadline gets closer, someone mentions the topic, or you run into something related. You remember the note or search for it.
2. It catches your attention during review. The idea still feels alive, so you rescue it.

If neither happens, it's worth asking whether it was important after all. Importance depends on circumstances. An idea that felt urgent in March can be irrelevant by June because the situation changed. Plenty of "important things I forgot" were important at the time, and the time passed.

Sinking doesn't delete anything either. The note is still in the stream, and search finds it.

The Zeigarnik effect

In the 1920s the psychologist Bluma Zeigarnik ran experiments suggesting that people remember interrupted tasks better than finished ones. The story usually told with it is about waiters who could recall complicated orders until the bill was paid and then forgot them. That became known as the Zeigarnik effect.

A better-supported finding is about what happens next. In 2011, E. J. Masicampo and Roy Baumeister reported that unfinished goals kept intruding on people's thoughts, and that making a specific plan for the goal made the intrusions ease off. You didn't have to finish the task. You had to trust it was handled.

Writing a thought down, somewhere you know you'll look again, does something like that. Karpathy describes the feeling in his append-and-review note: "When I note something down, I feel that I can immediately move on, wipe my working memory, and focus fully on something else at that time."

If a thought is still unfinished, it won't stay quiet. When you pass the note on review you'll feel it tug at you, and you'll rescue it. The notes that sink without complaint were usually finished with you.

Your attention is a good curator

Recommendation algorithms guess what matters to you from proxies such as recency, clicks or priorities you set months ago. They can't see what's on your mind today.

You can, even when you couldn't explain why a particular note stopped you. Something in it connected with what you're working on or thinking about. Gravity turns that reaction into the order of the stream: what resonates gets rescued, and what doesn't sinks. Over time the top of the stream becomes a picture of what you care about now.

How it builds over time

The notes that survive many reviews are durable: ideas you keep returning to, to-dos that matter in every context, references you keep needing. They've passed your own test repeatedly.

The notes that sink teach you something too. After months of rescuing some things and ignoring others, you get a feel for what's worth writing down and what isn't. Nobody taught you a system. You learned it from thousands of small decisions.

Letting go

Most tools for thought try to help you remember everything. Gravity helps you forget most things without worrying about it. The sinking is what keeps the stream manageable, keeps your attention on what's current, and lets you trust that losing sight of a note is usually the right outcome.

If you'd like the working-memory side of this, read your brain wasn't built to remember everything.