Re-Work's Virtual Assistant Summit is this week, and our very own UX Director Cathy Pearl is going to be speaking on the panel "Are Virtual Assistants Ready to take on Current Healthcare Challenges?" In preparation for her moment of glory on Friday, July 29th, I sat Cathy down to ask her a few questions about her career in the world of voice interaction.
Bayley: Will you explain your background in voice interaction?
Cathy: I started designing voice user interfaces in 1999. I remember being thrilled when I could say "checking account" or "savings account" and the computer would understand me. Since then, I've worked on all kinds of voice user interactions, from immersive tour guides in D.C. to getting tips from Esquire magazine advice columnists, as well as more practical tasks such as getting traffic updates or booking flights.
B: What have you seen as the greatest positives and negatives in bringing voice interaction to healthcare?
C: One positive is allowing people who, for various health reasons, are unable to type. Until recently, this would mean they were cut off from a lot of communication such as email. A family friend who had ALS was able to write emails and stay very involved in the ALS community because of voice dictation software.
Another positive is that, if done well, voice user interfaces can be a very natural and intuitive way for people to communicate--they don't need to read a manual to get started. And it makes people more engaged when they feel they're having a conversation rather than just pressing buttons. This ties into healthcare because higher engagement with their healthcare app can lead to better health outcomes.
As for negatives, I think the biggest one has been the state of speech recognition. Up until recently, the accuracy has been good but not quite good enough, leading to frustration. And being frustrated with your computer or phone is never good for your health!
B: What is the most exciting part of being in voice interaction design?
C: Being on the cutting edge of something that is about to explode into every day life. Speech recognition has been around a while, but only now with the advent of things like Google Now and Siri is it gaining wider acceptance. And with the latest developments in deep learning, the accuracy is becoming closer and closer to human levels, which means we can push more and more interactions into the natural mode of conversational speech.
B: Will you explain a bit of the direction and intention in designing Molly's voice and speech the way you have?
C: One of my main goals is to make Molly as conversational as possible. This means making sure Molly asks the right questions at the right times. It also means making Molly intelligent--if you just told her you didn't sleep well, Molly should not ask you, five questionslater, how you slept. A lot of systems today have good speech recognition, but they're not very smart, and people will not engage.
Luckily for me, I get to chat with Cathy about her pearls of wisdom in voice interaction and health care whever I want. You, however, should probably fill out this form to get in touch with her or any other member of our team.
'Til Next Time,
The Sensely Team