Zum Inhalt

ChatGPT thread

Spatial interfaces should feel like tools, not demos

Edited from a ChatGPT-exported thread on what makes a 3D UI worth keeping around.

threejsspatialchatgpt

Projects Outline

The shape of the work

  • Why are you doing this? (Purpose)

    Decide when a 3D interface is a tool and when it is a demo.

  • What are you building? (Product/Service)

    A spatial UI note, imported from a ChatGPT thread, then edited.

  • Who are your customers, and how are you helping them? (Customer needs)

    Platzhalter

  • What differentiates your product/service from others in the market? (Value add)

    The canvas is a window, not a document — copy stays in HTML.

  • What financial and social resources and skill sets do you bring to the table? (Resources)

    Platzhalter

  • How much are you willing to risk/lose to make your idea a reality? (Affordable loss)

    Platzhalter

  • What does success look like? (Success)

    Platzhalter

  • Description

    A spatial interface earns its keep when it does something a layout cannot.

  • Feasibility

    Platzhalter

  • Why Now

    Platzhalter

  • Innovation

    Platzhalter

  • Target & Environment/Context

    Platzhalter

  • Costs

    Platzhalter

Wins

  • Wins — Now

    Platzhalter

  • Wins — 5 Years

    Platzhalter

  • Wins — 25 Years

    Platzhalter

  • Wins — 100 Years

    Platzhalter

  • Market

    Platzhalter

  • Alternatives — Macro

    Platzhalter

  • Alternatives — Micro

    Platzhalter

  • People & Requirements

    Platzhalter

  • Outlook

    Platzhalter

Assets

Räume als Werkzeug

Assets — Concept Board
Platzhalter
Assets — Own Visualizations
Clay tower on the city side of the Zielbild landscape.
Assets — Zielbild integration
Räume als Werkzeug in the Zielbild

This project lights up the tower in the city zone.

Notes

This note was imported from a ChatGPT thread, then edited. The importer keeps the questions and answers as Markdown; I cut the rambling and left the claims I still believe.

The useful test

A spatial interface earns its keep when it does something a layout cannot: hold an object, show a relationship in depth, or let you inspect a structure with your hands (or a pointer that behaves like a hand). If the same information is clearer as a list, it should be a list.

The home glyph on this site is a small version of that test. It is not a product. It is a model generated by a local script, lit like a studio object, and easy to replace with a Meshy or Tripo export later. The scene should remain quiet enough to share the page with writing.

What I asked the thread

  • When is a 3D hero decorative, and when is it structural?
  • How do you keep WebGL from stealing the keyboard and the story?
  • What belongs in a GLB versus what belongs in React?

The answers, distilled: decorate less, load one good asset, respect prefers-reduced-motion, and keep the copy in HTML. The canvas is a window, not a document.