Hmm -- interesting/fantastic tool. Feels something like avy in Emacs, but for everything on screen. I think this should be possible in linux with a bit of work

Hmm -- interesting/fantastic tool. Feels something like avy in Emacs, but for everything on screen. I think this should be possible in linux with a bit of work (erm, famous last words?) especially because the whole desktop environment is fundamentally open and you don't need to depend on this providing an API. But I think an even better approach is to have build this functionality using screen parsers backed by recent AI advances. That way, you decouple the source / rendering of content from the sink / consumption of content, and can have more flexible behavior on behalf of the end user. I anticipate (hope) such tools to pop up over the next few years.