

Modal editing for just raw text input would actually be slower, because you also enter and leave Insert Mode. I find it’s very fast and powerful for navigating around the text, which you probably do a lot more than actually editing it. And when it does come to editing, there are a lot of higher-level tools (at least in Vim) for accomplishing things more quickly, like the ‘s’ command and ‘q’ macros.
I think getting into a mental “flow” state is really valuable, and muscle memory is important for being able to stay there. If your muscle memory is to navigate around using the mouse, that’s great, but Vim feels faster to me.
Herndon and Dryhurst are frequent collaborators, and xhairymutantx is their work. So they didn’t just prompt an LLM to make the image, they trained the model themselves. And they specifically trained the model on pictures of Herndon (who has distinctive red, braided hair).
I’m personally a really big fan of their work (which I don’t expect everyone to be), but the picture that’s being circulated in articles and apparently sold at auction without context is pretty uninspiring.