We want to modify the studio AI assistant box in a way that supports a human in the loop approach.
- authors should be able to see the question before commiting to the library
- authors should be able to edit the final form
- authors can ask the llm for second, third and so revisions