Multimodal outputs

[dadahl]

how to include html, video, audio, and image outputs

[dadahl]

should these really be considered utterances?

[dadahl]

Maybe these should be considered "presentations"? Don't take "utterance" out, but add a new event. Have to consider where images and videos go on the screen.

[dadahl]

How to display avatars? The avatar information would be in the manifest (image, voice). Is there a reason to change the avatar in the middle of a conversation? No, you should just switch to a new agent. For example, in a doctor's office you have a receptionist, nurse and doctor, who would be separate agents.