Brainstorming In Your Browser

The problem is not the ideas, it is the residue

A brainstorm produces three things at once: half-formed ideas, a trail of pages you looked at, and — if you are lucky — a decision. The ideas end up in a chat window, the pages end up as twenty-six tabs, and the decision ends up in your head. A week later you have all three and can reconstruct none of them.

StickyFriend’s answer is not a better note-taking app. It is that the notebook, the chat and the tabs are the same workspace, and that workspace is the browser you were already using.

Step one: give the idea a project, and let it take the window

A project in StickyFriend binds to a Chrome window. Open the project, you get its window; the tabs open in it belong to that piece of work and to nothing else. This sounds administrative and turns out to be the whole trick — “switch to the other thing I am working on” stops meaning “close forty tabs first”.

Two consequences worth knowing before you start:

Step two: write in the notebook, not in the chat

A chat is a transcript. It is a terrible place to keep a conclusion, because the conclusion is buried at message forty-one between two false starts.

The notebook is a writing surface that you and the AI both write into. You can edit what it wrote — that is the point, it is not a read-only log. Practically:

Step three: quote the turn that actually mattered

Brainstorming with an AI produces a lot of text, and roughly one paragraph in ten is the one you want to build on. Rather than copy-paste it into a fresh message and lose the context, you can quote a specific turn and reply to it. The exchange stays anchored: the reply is visibly about that paragraph, not about the general vicinity of the conversation.

This is the small feature that changes how a long brainstorm reads a week later. A transcript with quoted turns has structure; a transcript without them is a wall.

A working rhythm that holds up

Think in the chat. Decide in the notebook. Quote the turn that produced the decision, so the notebook entry points back at its own reasoning. Then close the chat without a shred of guilt — you kept the part that mattered.

Step four: look at the shape, not the scroll

Every project builds a mindmap of its own ideas — nodes, and the links between them. It exists for one job: to answer “what is this project actually about now?” without scrolling chat history. When a brainstorm has genuinely branched, the map shows you the branch you have been ignoring.

Step five: when it stops being a brainstorm

At some point the thinking is done and the work begins. The same project already holds the pieces:

Nothing has to be migrated to a different tool, because nothing left the project.

What this does not do

Being straight about the edges, because finding them yourself after paying is worse:

If one of those is load-bearing for you, check the roadmap before you buy, not after.