I used to spend days rotoscoping people in videos. Generative infill for background painting and automatic rotoscoping have saved probably a year of my life at this point. Image generation relies on CLIP, which needs a language model for conditioning.
josephc
I might have felt sympathy, but I'm too busy caring for people that he himself hurt. They deserve my attention more.
You don't run over a crowd of people, crash your car, and then get to lament that there are people in front of you at the emergency room.
We are kindred spirits. Every so often I reflect on the fact that JS/TS are the most frequently used languages in the world and I feel a bit sad.
Surprisingly, ChatGPT is the mortal enemy of a lot of machine learning researchers. It gives people a bad impression about what the field is, poisons future training data, and gives AI researchers who make truly remarkable and useful things a bad reputation because everyone "hates AI" when they really hate the stupid VC backed circlejerk that becomes inevitable in a society like this.
I've been using Resolve on Ubuntu and after a bit of a bumpy start it's working just fine.
FreeCAD got a 1.0 release late last year. I need to try it again. I might try one of the browser based CADs, but the idea grosses me out.
The one substitute I've yet to find is for Photoshop, so I need to run that through a VM. It annoys me that nothing compares. I respect Gimp but the UX is overtly terrible. Krita has similar issues for photo editing but is slightly better. Affinity Photo doesn't run on Linux.