DeepHorus is an innovative tech media and intelligence platform focused on AI, blockchain, startups, and emerging technologies.
Published on

Quick Study

Authors
  • avatar
    Name
    DeepHorus
    Twitter

Claude Opus 5.5 High has been used to build "Quick Study," a browser-based artifact that takes an uploaded photograph and repaints it with simulated brushstrokes on an HTML canvas in just a few seconds. The demonstration is dated 23rd October 2026 and was produced from a single natural-language prompt rather than a hand-written codebase.

The prompt asked the model to "build an artifact where you open a photo and it rapidly (in a few seconds) paints a version of that photo using a canvas element and simulated paintbrush." The resulting tool loads an image, analyses it, and renders an animated, painterly reworking directly in the browser.

Under the hood, Quick Study uses Sobel edge detection to identify the contours and structural lines within a photograph before the simulated brush follows those paths. Notably, the artifact ships with no external libraries beyond Google Fonts, keeping the entire painting engine self-contained in the canvas element.

The experiment follows an earlier project in which Claude Fable was prompted to simulate a watercolour illustration using P5.js, suggesting the author is systematically testing how different Claude models handle creative, generative visual coding from short prompts.

The work points to how quickly frontier models now convert brief, open-ended instructions into functioning interactive tools. Tasks that once required a graphics programmer — edge detection, brush simulation, and frame-by-frame rendering — were compressed into a single request and a few seconds of runtime.

Context around the release comes from the author's blog, which in October 2026 also covered "Battle Drones," a vibe-coded real-time strategy game with audio-based multiplayer, a Python 3.15 migration post, and a separate Claude Haiku 5.5 write-up.

For developers, Quick Study illustrates a widening pattern: models acting as rapid prototypers for creative coding. Whether such single-prompt artifacts mature into production-grade image tools remains open, but the pace of these public experiments suggests more will follow.

Read original article →Source: simonwillison