The Informed Life Mary Parks on Voice User Interfaces
Dec 21, 2019
Mary Parks, an expert in communications and applied linguistics, dives into the world of voice user interfaces, sharing two decades of design experience. She discusses the intricate workings of these systems, including how they categorize user commands and the challenges they face in understanding intent. Mary reflects on the humorous side of voice interactions, sharing relatable blunders, and touches on the complexities of identity in voice technology. Her insights reveal the fascinating blend of linguistics and technology in shaping our digital communication.
AI Snips
Chapters
Transcript
Episode notes
From ESL Teacher To Voice UI Designer
- Mary Parks described how she pivoted from teaching ESL to voice UI design after discovering tech roles for linguists in the 1990s.
- She credits early startup experience and mentors for accelerating her transition into speech technology.
Input And Output Must Be Designed Together
- Voice UI design splits into two main components: speech input and speech output, which must be designed together.
- Treating input and output independently risks major integration failures.
Architect Speech Input Intents First
- Architect the speech input so utterances map into meaningful buckets or intents before building behavior.
- Ensure the input tagging structure fits how the system must respond to avoid disasters.