Limited beta

Where this actually stands

The engine that decides whether a Moment is worth your attention runs today, but it runs quietly. It is not delivering to people yet. We would rather it stayed silent for another month than have it learn to interrupt badly.

Evidence first, then a sentence

A Moment has to be able to say what it is based on. If the reasons are thin, the correct output is nothing at all, and that is the behaviour we are tuning for.

A suggestion is not an action

Lila proposes; you decide. Nothing is sent, booked or changed on your behalf because a model felt confident about it.

Most of the time: nothing

The measure of this feature is not how often it fires. It is how rarely you wish it had not.

Silence is better than a wrong insight.

Anything can interrupt you. The harder thing, and the thing worth building, is knowing when not to.

Request early access

Lila is in closed beta. No public App Store or Google Play listing yet.