I mentioned this somewhere else where same topic was being discussed. If the Capture card had an option for the prompt to be italic or greyed out, then I wouldn't mind putting "listening..." or something like that. It wouldn't "read" as part of the conversation so it wouldn't seem unnatural, but it would let the user know it was their turn to say something. (I'd be fine with just a blank capture card, I am only suggesting this (italic/greyed out option) bc there was a concern that the user wouldn't know the bot was finished, and this would address that.)