Philosophy

1 min read

We're building the open source technology to make local-first personal models ubiquitous.

Here are five ideals we would like to strive for in local-first personal models:

  1. Private by default: models and data stay within a privacy perimeter you control and can verify
  2. Personalized updates: models learn from your personal data traces, with incremental and cheaply-communicable updates
  3. Built to last: models update without breaking, and old versions still work if you ever want to switch back
  4. Clear how it works: see how models interact, combine, and reason
  5. Always with you: models work offline, stay synced, and available across all your devices

The project has two values: simplicity and pragmatism. These values will help people understand what's going on, helping us to go further and faster.