Expressiveness in voice AI should come from conversation context, not a style setting

deepgramscott · x · 2026-08-18

Deepgram's Scott argues that the meaning of a sentence rarely lives in the words themselves: "Sure." can mean agreement, annoyance, disbelief, or "stop talking" depending on context. So expressiveness in voice AI shouldn't be a style setting but a product of conversational context—if a user has been frustrated for five minutes, the next response shouldn't suddenly sound upbeat.

Original post →

More from Multimodal

Multimodal channel →