Leading with Perspective E08: Larry McGrath


What makes a conversation feel natural rather than scripted? What kills it in three seconds? What makes someone feel understood versus processed?

In this episode, we are joined by Larry McGrath, a User Research Lead at Amazon, where he has spent the past four years working on Alexa Plus, Amazon’s conversational AI platform, to talk about how to build AI that actually works for people.

With deep roots in the social sciences, Larry has dedicated his career to understanding how sociology, anthropology, and the study of conversation can shape the design of AI products that feel more natural, responsive, and human.

In conversation with ReD’s Tamara Moellenberg, Larry walks us through the layers of the AI product stack and makes the case for why social scientists belong at every level, from how we design the rhythm of a conversation to how we align an AI’s values for coherence, not just the avoidance of bias.

He also reflects on what product managers most often get wrong: losing sight of the human need they set out to meet in the first place.

Host: Tamara Moellenberg

Guest: Larry McGrath

Introduced by Matthew Janney


 

Next
Next

Building strategy at times of high uncertainty