> For rephrasing, or for making text flow more easily, or for reorganizing an outline and having the text update to reflect that higher-level reorganization?
Current models can kind of do this, e.g. you could give some text to GPT-3 and ask it to do this. Not sure what the limit on the input length before results degraded would be. But yep, would be nice to see this out there in a nice interface and working well.
> Where is the support for finding other text you've previously written that might be appropriate?
Semantic search could be used for this, e.g. finding blocks of text similar to what you've already written
But yep, this stuff isn't all nicely packaged in one place
The best practice, which prevents this issue, is to use envs, (as well as pinning your dependencies for any given project in a committable artifact).
(Seems like a lot of people are using system Python, which isn't really recommended.)