IBM Watson Moments | Design Principal and co-creator

Designing for AI: The perception of convenience

Scroll ↓

In July of 2018, we had a Watson problem. People thought it was just a robot from Jeopardy, not the powerful AI we knew it to be. We had to connect with people and understand their perceptions of AI’s presence is in their life. That's a huge task when you're talking about a company as large as IBM and the countless lines of business they support. We not only had to give Watson a personality, and make sure everyone, inside and outside IBM, recognized it. It was about telling a story, making AI relatable, and bringing man and machine closer together. 7 team members, 100+ playbacks, and 1000+ iterations.

Critical challenge

How might we demonstrate compelling and trustworthy applications of AI?

What I loved about this project

Exploring human factors and emotions. Not just the way that AI should look and what colors/shapes to use. I enjoyed asking questions like “How would you feel if a robot smiled at you…with teeth?” For me it was about understanding a humans direct reaction to an expression. Designing for this perception was the nuance we needed to consider. In doing so, we came up with what is known as “The Tiers of Expression”. It’s a framework and core set of principles that helps anyone, designers specifically, think through what expressions they need to consider when designing for AI use cases (e.g. importance + time + value gained = how to apply). It allowed me to go deep on AI, human behavior, and what that might feel like in our products. Ultimately myself and another Design Principal wrote an article and got to do speaking engagements about it. I enjoyed getting to meet and talk with new people about our work.