> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pond.li/llms.txt
> Use this file to discover all available pages before exploring further.

# What is Pond?

> Pond is a tool for understanding.

## Vision

Pond is founded on two observations:

1. **The prevailing knowledge management paradigm is overly concerned with capture, retrieval, and organisation of information.** We owe our vast and inert archives to this — collections that grow but rarely contribute to thought.
2. **Emerging AI-native modes of engagement outsource critical stages of understanding.** Understanding is often a gradual, temporally extended activity that involves tracing the relations between a set of ideas and allowing them to alter, and be altered by, experience.

Pond's aim is to furnish the conditions for this kind of activity.

Think of Pond as a sort of [Commonplace Book](https://en.wikipedia.org/wiki/Commonplace_book) with a set of AI capabilities that allow it to participate more actively in your thinking.

<Note>
  Pond does not afford the opportunity to search for and chat with large collections of media. In some part this is because there is nothing it can add over, e.g., NotebookLM on that front, but more significantly it is because it is premised on the hypothesis that the chat-response model is not *most* conducive to cultivating understanding.
</Note>

## Pond's Ontology

You can use Pond knowing just three words:

* [Fragments](/fragments) — the passages you mark as significant in the sources you bring in (articles, PDFs, videos, notes, images — the docs and API call these [Artefacts](/artefacts))
* [Threads](/threads) — working spaces where Fragments are related, organised, and engaged with over time
* [Notes](/notes) — the prose you write in a Thread, referencing Fragments and other items by `@`-mention

Everything else in Pond is behaviour you'll notice before it needs a name. When you relate two Fragments, a link forms between them — the docs call these [Connections](/connections). When you tag a Note with how it develops the Thread, it becomes a **Move**. Pond navigates this relational layer to suggest Fragments for your Threads and surface new lines of inquiry — [Pond Intelligence](/pond-intelligence). As you read and navigate, Pond keeps your path so you can pick up where you left off ([Trails](/trails)).

At a glance, this is the loop that Pond enables: texts (and other forms of media) come apart along lines of salience into Fragments, which are then recomposed into Threads that represent themes of interest, topics one aims to understand, which Pond helps you nurture and develop — including by letting your Threads [frame](/frames) other texts (and other forms of media) you encounter in the future.