Show HN: LLM draws SVG, StableDiffusion polishes it

i had this idea for an AI workflow:

1. describe an svg drawing in great detail

1.5. ask the llm to label its drawing because it's fun to watch

2. generate the svg

3. img-to-img with sdxl

it's a silly demo but kinda fun, source: https://www.val.town/v/substrate/scribble

you can coax Llama3 into generating better SVGs using multiple steps, actually – in a previous version the SVG had gradients, stacked layers, dot grid textures... even animations

but i had to do a roundabout thing to rasterize the SVG in the val sandbox so i removed all that

https://x.com/vprtwn/status/1811710646729277740