I remember it being about generative UI in one of the Rabbit keynotes and I am very interested in the concept of it.
Are there some ideas around this for implementation in the foreseeable future?
For example you could have responses with multiple answers and have them in a list that can expand per item. (Rabbit could understand what item is currently open if follow up queries are made)
This ofcourse would also mean that what is on screen would not be read out verbatim.