[Feature Request] Make Speech more responsive
To appear more humanlike and increase the preceptive responsiveness of the robot I would like to suggest the following feature:
- The robot should start speaking instantly after a request.
- If LLM inference is not yet ready, predefined (humanlike) "äähm", "öhhhm", "hmm", etc sounds should be played back.
- The playback should be continued until inference is done and the answer can be given.
/cc @meixner @dreher will be open for discussion
In case of a patent, @marzi would be willing to settle for 51% ownership.