Evolving JavaScript: Cultivating Genetic Algorithms for Creative Coding
summary
A hands‑on run through genetic algorithms for creative coding showing concrete tools, results, and failure modes: a canvas + TypeScript + xstate demo evolves random circles into a target using roulette‑wheel selection, reaching a 99.4% match at ~2700 generations (second particle demo climbs to ~87–99% but can stall live because mutation settings break convergence); population sizes, selection method, mutation rates, fitness weights, and frame rate (up to ~200fps) are all specified and tuned.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 3:32Genetic algorithms birth
- 9:32How GAs work step by step
- 13:03Creative coding demo
- 19:45Particle system evolution
- 23:56Canvas tips and resources
- 26:05Key lessons from GAs
- 28:12Closing thoughts
related talks
29mEvolving JavaScript: Cultivating Genetic Algorithms for Creative Coding
29mJavaScript Takes on Genetically Evolving Cellular Automata
38mEvolving Game Development with Genetic Algorithms
12mAlgorithmically Generated Visual Designs
22mGenerative Art Speedrun
25m"Code As A Canvas: Exploring Creative Coding With JavaScript"
28mCreating generative art with Javascript
32mThe joy of creating art with code