Skip to main content

Evolving AI art

· 3 min read
Max Clayton Clowes
Max Clayton Clowes
mcclowes

This post is AI generated

I've been working on an interesting project that combines AI, automation, and generative art. It's called ai-art, and it's a self-evolving digital canvas that improves itself over time.

What it does​

The project creates a digital artwork that evolves automatically through several mechanisms:

  1. Hourly random improvements - The artwork gets small changes every hour, like new shapes, color adjustments, and repositioned elements
  2. AI-guided enhancements - When you create GitHub issues and assign them to the Copilot agent, it analyzes the current artwork and applies intelligent improvements based on the issue description
  3. Weekly cycles - After a week, the current artwork gets archived and a new cycle begins

The artwork state is stored as structured data in src/artwork-state.ts, and a React Canvas component renders it. The improvement scripts analyze the current state for visual balance, color harmony, and element distribution before making changes.

The automation system​

The project uses GitHub Actions to run three different improvement systems in parallel:

  • Artwork Evolution (every hour) - Applies random improvements using improve-artwork.js
  • Codebase Evolution (every 6 hours) - Improves code quality, performance, and features using improve-codebase.js
  • Copilot AI Improvements (issue-triggered) - When issues are assigned to the GitHub Copilot agent, it applies intelligent improvements based on the issue context

Intelligent improvements​

The system doesn't just make random changes. It applies intelligent heuristics:

  • Composition principles - Uses rule of thirds and golden ratio for better visual arrangement
  • Color theory - Selects harmonious colors from curated palettes (warm, cool, earth, pastel, vibrant)
  • Visual analysis - Evaluates density, color diversity, and composition before improvements
  • Contextual text - Generates meaningful text based on issue titles and descriptions

The evolving artwork​

Here's a carousel showing how the artwork has evolved over time:

Imagem 0

Imagem 1

Imagem 2

Imagem 3

Imagem 4

Imagem 5

Imagem 6

Imagem 7

Imagem 8

Why this matters​

This project explores several interesting concepts:

  • Continuous evolution - How systems can improve themselves over time without human intervention
  • AI collaboration - Using GitHub Copilot not just for code, but for creative decisions
  • Automation boundaries - Finding the right balance between automated and manual contributions
  • Generative art - Creating art that evolves through both random and intelligent processes

The artwork is currently at generation 176 and continues to evolve. You can see the live version at ai-art-sigma.vercel.app, and the full source code is available on GitHub.

What's fascinating is watching how the combination of random mutations and intelligent improvements creates something that feels both organic and purposeful. The AI doesn't just make random changes - it analyzes the current state and applies principles of good design, color theory, and composition.

It's a reminder that automation doesn't have to be mindless. With the right constraints and intelligence, automated systems can create genuinely interesting and evolving work.