Greg Bennett is a Linguist and Lead User Researcher at Salesforce with extensive experience in the technology sector having previously worked for companies such as Yahoo! and Microsoft. He believes the core of user experience lies in the intersection of language and culture and utilizes ethnography and linguistics to cultivate empathy with users and inform design for the best possible user experience.

In today’s episode, we talk to Greg about chatbots, how they learn right from wrong; how to deal with context and social identity when building them and how to approach bias and training of chat data. We also cover text-based interactions and how to read them, the meaning of pauses in a conversation, and anonymity and politeness on social media. Lastly, he shares how his expertise as a linguist fits into a multidisciplinary team.

Mentioned in Podcast:

• Deborah Tannen, Conversational Style, Analyzing talk among Friends
• William Labov
• Francesca Rossi AI Ethics Global Leader, IBM Research (AI for Good 2018)
• World summit AI conference, Amsterdam
• Jefferson (1974) - Notes on a possible metric which provides for a 'standard maximum' silence of approximately one second in conversation
• Pomerantz (1984) - Agreeing and disagreeing with assessments: Some features of preferred/dispreferred turn shapes
• Labov (1972) - The social stratification of (r) in New York City department stores

Greg’s work:

https://www.gabennett.com/publications

Social media and other links:

https://www.linkedin.com/in/gab45/

https://twitter.com/gabennett45