Quick-start Guide

Turn any question into a visual knowledge map

Ask a question, get an AI response, then branch into deeper follow-ups. Every answer becomes a connected node on a shared whiteboard you can revisit and grow.

How it works

Three steps from question to knowledge map

1

Ask a question

Type any topic or question. The AI generates a structured response that becomes the first node on your whiteboard.

Example

“What are the main theories of consciousness?”

2

Branch deeper

Highlight any word or phrase in a response and ask a follow-up. Each new answer links to the exact concept it came from.

Related ideas Pros & Cons Blend Explore all
3

Share & revisit

Your whiteboard persists. Share the link so others can explore the same map, add their own branches, or pick up where you left off.

Share a direct link to the full map or any single node
Your workspace

Everything in one view

The workspace is split into three areas that stay visible at once: a reading panel on the left, an interactive graph on the right, and a toolbar at the bottom.

Reading Panel

The left side shows the full AI response for the currently selected node. Read the content, then interact with it directly:

Select any text to…

Explain

Get a deeper explanation linked to the exact phrase

Ask a question

Type your own follow-up about the highlighted passage

Highlight

Save an important passage for later reference

Comment

Add your own annotation or note to any passage

Graph Canvas

The right side is an interactive node-and-edge map of your entire whiteboard. Click any node to read it, drag to rearrange, and scroll to zoom. You always see how the current idea fits into the bigger picture.

Action Toolbar

The bottom bar gives you one-click tools that operate on whichever node is selected. Each generates new connected nodes on your map:

Related Ideas

Generate connected concepts, counterpoints, or tangential topics branching from the current node.

Pros & Cons

Create a balanced argument map with supporting and opposing points linked to the claim.

Blend

Merge two nodes into a synthesized response that draws on both perspectives.

Explore All

Auto-detect every key point in a response and branch into each one at once.

Reading Tools & Controls

Additional tools sit alongside the action buttons for navigating, saving, and sharing your work:

Notes

Star nodes you want to revisit

Reader

Read the full thread as a page

Translate

Open any node in Google Translate

View Options

Change graph layout and display

Highlights

See all saved highlights in one panel

Share

Share the map or control who can edit

Why a whiteboard instead of a chat?

Structure is visible

See how claims, evidence, and counterpoints connect instead of scrolling through a flat thread.

Collaborative by default

Share one link. Multiple people explore the same map and add their own branches.

Learning compounds

Your map persists. Come back later, review branches, and continue from where you stopped.

Key capabilities

Graph-native answers

Every AI response is a node in a connected structure.

Term-level deep dives

Highlight a phrase to generate focused sub-branches.

Shared whiteboards

Multiple people build understanding on the same map.

Persistent & searchable

Maps remain available, searchable, and linkable over time.

Linkable context

Share the full map or direct links to specific nodes.

Portable output

Export your work for review, teaching, or documentation.

For educators

Set up a classroom whiteboard

Create a structured idea map for a lesson, then share one URL so students explore the same concept together.

  1. Seed the map with key questions, claims, or sources.
  2. Guide exploration — direct students to specific nodes.
  3. Share one link — everyone works on the same whiteboard.

Share this guide

Copy this page URL into your LMS or class chat.

About RationalGrid

Built on a simple idea: learning branches, loops, and connects. RationalGrid transforms AI conversations into visual, persistent knowledge maps so insight accumulates instead of disappearing in a scrollback.

An open source project by Tom Berman. View source on GitHub.