There is a “thinking” message when loading, but not this.
When listening to someone in person and I don’t understand or I need more context my eyes squints a little, head turns slightly.
When I understand, head nods, eyebrows go up, eyes widen.
Having a message or even a little expression capable face on the side of the screen for prompting would allow the user, especially to novel and new user, understand where the response is coming from, therefore prompting better prompts by the user.
Just an idea.