A new conversation architecture for AI interfaces.

01 — Why Branchscaipe?

Every good question costs you something.

Follow a tangent in a normal chat and you lose your place. So most people don’t follow it. The question goes unasked, and the thread stays tidy — and a little poorer for it.

Branchscaipe removes the cost.

02 — The Problem

Everyone hit the same wall.

One thread. Every question you’ve ever had about the trip, in whatever order you happened to think of it. Nothing groups. Nothing stays findable.

And the thing you actually came to do is somewhere up there, out of reach.

03 — Why Linear Chats Break

The cost isn’t the mess. It’s what the mess makes you stop doing.

• Context pollution

  Unrelated topics share one context. Neither you nor the model can tell what the conversation is actually about any more.

• Scroll fatigue

  The answer you need is up there somewhere. Finding it costs more than asking again — so you ask again.

• Decision friction

  Every follow-up starts with an unwanted question: does this belong here, or somewhere else?

• Anxiety of divergence

  Eventually the safest move is not to ask. Curiosity gets rationed to protect a thread.

04 — The Insight

The problem was never the answers. It was the shape of the conversation.

Every fix I could think of made the thread tidier. None of them let you leave it.

05 — Exploring Alternatives

Four ways to branch. Each broke somewhere different.

• 01

  Indentation

  Depth eats the width

• 02

  Infinite canvas

  Zoom until nothing is legible

• 03

  Floating bubble

  Covers what you were reading

• 04

  Side-by-side split

  Splits into useless slivers

06 — The Final System

One conversation at a time — with a way back.

Every rejected concept tried to show you more at once. This one shows you less, on purpose. Depth moves backward instead of sideways, so going deeper never costs width, never needs a zoom, and never divides your attention.

You are always in exactly one place — and you can always see the way back.

07 — How It Works

Leave, explore, come back with something.

Branch off a phrase and you get a real conversation of your own — with a trail back. Follow it as far as it goes.

Return, and the thread you left is exactly where you left it. Nothing moved. Nothing was lost.

And when the detour turns up something worth keeping, merge it back — without losing the branch it came from.