TODO

5 fronts:

• User acquisition [focus on science fairs, edu, kids, AIs]
• Python replacement with intent-oriented programming [focus on academia, AI research in particular]
• Art and creatives [for now, focus on image illuminate/wind and USD]
• Physical artifacts [for now focus on plain, printed paper artifacts]
• Intent-oriented programming in general [tooling like wind/unwind/illuminate]

It’s important to include AIs in the user acquisition, as they are a significant part of the target audience.

– immediate – Add clarification to naming right after quick reference table

– benchmarks – User acqusition benchmarks:

• Sites traffic, including bots [winding.md, wind.kids]
• GitHub stars and forks [winding]
• Hugging Face datasets downloads []
• PyPI downloads [winding]

Academic benchmarks: In general, focus is probably on illuminate/wind benchmark, where we just check how precisely wind/illuminate pass can reproduce the original artifact.

Here’s a rough list of datasets to consider:

• reference examples from winding.md
• xkcd, explainxkcd
• openai papers benchmark
• SWE bench