Artificial IntelligenceReleasedUpdated Jul 26, 2026

Prism

Most AI image tools force you to start over every time you want a small change, and they give you no way to know what the image actually contains if you can't see it.

Why this matters

Visual creativity should be iterative, not punitive. When a small edit means regenerating everything, people lose control over the result.

Current idea
An AI image editor for generating, modifying, and describing visual content. Users can create new images with AI or upload existing ones, then make in-place modifications without regenerating the whole image from scratch. It also produces descriptive text for any uploaded or generated image.
Accessibility notes
Every generated and uploaded image gets a text description, so the tool is usable without sight. Keyboard-navigable controls and screen-reader-friendly output.
What was learned
Small, controllable edits feel more empowering than a single perfect generation.

Development updates

Progress in public

  1. In-place image editing and automatic description generation released.

  2. Initial AI image generation and upload support launched.

Feedback

Have thoughts on this project?

Head to the Ideas board to share a related barrier or idea. Attribution is optional.

Open the Ideas board