I feel like I could do some fun things with a segmenter tuned to drawn objects. MediaPipe (understandably) isn't quite there though.
Potential to do more conventional computer vision stuff? Maybe.
Niri (scrollable window manager) keeps adding features. They just released an impressive overlay mode - and one of my longtime faves: windowed fullscreen mode.
Really cool results and I also love the feeling conveyed by "complete plasticity with the raster".
Still getting a feel for the shape and possibilities of this
I think I pinpointed a misgiving I have about general collaborative LLM use: that the LLM will never stop and say "I don't know" or "I don't have this mental model, can you help me understand it". That's the fundamental limitation of being trained to plausibly (not correctly) generate text. read on scrawl