Emily Uematsu Banzhaf, Content Designer at WillowTree, talks to Ben McCulloch about conversation design. Emily is also a conversation/VUI designer, she leads the ethical use taskforce at Open Voice Network, and she’s a professional violinist!
We talked about personas, ethics, the data that can be collected from your voice, ear training and the sound of conversations.
In this episode
- 00:00 – Introduction
- 02:00 – 30 seconds, go!
- 04:06 – What does ‘bot persona’ mean to you?
- 06:06 – What’s the purpose of a bot’s persona?
- 06:56 – How ‘human’ should bot personas be?
- 11:12 – The risks of collecting user data
- 12:32 – The issues of speaking legal-ese
- 15:22 – What data about a user can be collected from their voice?
- 19:17 – Should a bot be ready to speak about things other than their use-case?
- 24:56 – How much does the sound of the voice matter?
- 26:12 – Ear training is useful in conversation design
- 27:07 – How similar is phrasing on violin to verbal communication?
- 29:17 – Hearing issues in TTS when you have ‘perfect pitch’
- 30:20 – Do you think of music when you adjust SSML?
- 31:20 – Do you speak your copy out loud to get a feel for the SSML?
- 31:56 – Outro
Join our next webinar, conversational commerce: how AI-powered digital workers can boost sales: Find out more: https://vux.ai/conversational-commerce-how-ai-powered-digital-workers-can-boost-sales/
