Return to Transcripts main page

The Next List

Profiling Social Roboticist Heather Knight

Aired December 04, 2011 - 14:00   ET

THIS IS A RUSH TRANSCRIPT. THIS COPY MAY NOT BE IN ITS FINAL FORM AND MAY BE UPDATED.


(BEGIN VIDEOTAPE)

DR. SANJAY GUPTA, HOST: Welcome to THE NEST LIST. I'm Dr. Sanjay Gupta.

You're about to meet Heather Knight. She's a roboticist with an edge. She's teaching her robot to do stand up comedy, perform dance moves, even recite Shakespeare.

Now some of this might have her academic colleagues sort of scratching their heads, but there's a point to all this fun. Heather wants to better understand what makes people feel connected to technology, and for Heather, the secret lies in this little robot, Data.

This is THE NEXT LIST. Everyone you'll meet is a force in their field. They are agents of change. And that's why they are on THE NEXT LIST.

HEATHER KNIGHT, RESEARCHER, CARNEGIE-MELLON UNIVERSITY: My name is Heather Knight, and I'm a doctoral researcher at Carnegie-Mellon Univerrsity, and I study social robotics.

DATA: They call me Data the Robot. Gosh, I love saying that.

KNIGHT: A social roboticist is someone that makes robots that can interact with people in a human way.

When I tell people I make robots, they're usually like wow, that's really cool. And when I tell people I make social robots, usually they're a little confused at first. And I found that the best way to talk about what I actually do is by example.

ROBOT: Heather, how about you get working on that emotion program?

KNIGHT: We're trying.

ROBOT: Fair enough.

KNIGHT: Usually when people think of robots, they can think of two ends of the spectrum. One of them is these Hollywood sort of "Terminator" films or "Wall-E," like this fully interactive, almost like human in a different robot-shaped box sort of systems.

And then on the other end of the spectrum, you have these machines that might plow our fields or put our packages in the mail or some of those coolest cases, rove around on the surface of Mars. So when you talk about social robots, that's something that's actually somewhere in the middle of those two extremes.

ROBOT: I am a robot, yes, a robot. If you pick us in our battery pack, do we not bleed our alkaline fluid?

KNIGHT: Robots have a lot to learn, right? They're new to environment. They're an alien that's come down to our planet. They don't know our grammar. They don't know our social structures.

So they're going to make a lot of mistakes. Humor can be a nice escape from that. Like, you know, like any like charming international student that doesn't quite speak our language has probably gotten away with a lot by sort of a little laughter at one's self.

ROBOT: My programmer helps that one day I will be an autonomous robotic performer, like Justin Bieber or, perhaps, Charlie Sheen is a better choice.

KNIGHT: The first performer robot that I made was a stand-up comedian. And not knowing very much about comedy at the time, I thought it would be all about watching the audience. I thought the audience would really define what humor was and every audience might have unique characters.

ROBOT: Hello. It's an honor to be here. You guys are looking good out there, ready for some jokes?

KNIGHT: The first system I designed, the robot was watching and tracking their laughter, their applause, and I had given the audience niece these back cards that could be red or green depending on how they liked something.

It was very like -- I mean, it was a robot that cared about other people, you know. Like -- I don't know how many comedy shows you all have been to lately, but there are a lot of ways to do comedy.

ROBOT: So, a doctor says to his patient, I have bad news and worse news. The bad news is that you only have 24 hours to live. That's terrible, said the patient. How can the news possibly be worse? I've been trying to contact you since yesterday.

KNIGHT: So I wanted to see a world where we actually have useful robots that positively impact our lives.

ROBOT: My programmer designs my presentations, with the goal of driving innovation and social robotics, which is the integration of robot helpers into everyday life. So you might as well get used to this.

KNIGHT: Something that is important is actually getting these machines out into the world with real people and so it's awesome that engineers know how to build stuff. We're really good at that. I think that we can also learn a lot from artists about making stuff that people really care about. ROBOT: Something has gotten a hold of me.

KNIGHT: So when I make social robots I'm trying to think about new spaces of applications for technology in our everyday lives in ways that were never possible before, before you've come up with some of this new capacity for them to understand us. New ways for them to communicate with us, that was awesome.

ROBOT: Thank you, everybody.

KNIGHT: Man, he loves an audience.

ROBOT: Coming up, heather is giving me some serious acting lessons.

(COMMERCIAL BREAK)

GUPTA: Welcome back to THE NEXT LIST. I'm Dr. Sanjay Gupta. As you probably figured out by now, Heather is unlike any roboticist you'll ever meet.

In Heather's version of robotics, getting the gestures right is a really big deal. Who better to teach these gestures than a real acting professor? Together, they've programmed "Data" in a way that almost makes it seem like he's alive.

UNIDENTIFIED MALE: Roll film and then roll up on the count of five, one, two, three, four, five. Good.

KNIGHT: People have spent entire careers like lifetimes thinking about what gesture means thinking about how to tap into an audience, how to make a believable interaction on stage.

So as we explore this realm of social robotics for the first time, I believe that collaborating with performers and with different artists can actually help us bootstrap the development and the creation of these technologies.

UNIDENTIFIED MALE: Roll forward, stretch the fingers and arms down. There you go. Excellent.

MATTHEW GRAY, ASSISTANT PROFESSOR OF ACTING, NORTHEASTERN UNIVERSITY, MASSACHUSETTS: My name is Matthew Gray. I'm assistant professor of acting at Northeastern University in Boston, Massachusetts.

I met Heather when I went to a meeting called "dork thought." Before I even met her, her robot, "Data" was sitting, looking pensive at his toes on the stage. I suddenly realized that was the actor I wanted to play Hamlet.

I've wanted to direct Hamlet for a long time. It sort of culminated in this idea we had of what if we could teach "Data," acting lessons, the way that some human actors ask for acting lessons.

We started with this idea of a simple action that's understandable, like the word to push. Right, Data?

ROBOT: That is easy for you to say, but I am a robot.

GRAY: Data, you performed in front of hundreds of people. You know exactly what I'm talking about.

KNIGHT: Don't get stage fright now.

GRAY: Exactly. Come on, let's try some, to push.

ROBOT: Was that OK?

GRAY: That was great.

GRAY: What I found that I've learned through working with Heather is actually my own misconceptions about robotics. I had assumed that well, robots are repeatable. And infinitely repeatable so they don't make mistakes. One of the first things I learned from working with Heather and "Data" is there are wonderful idiosyncrasies to each robot.

Each iteration processes at a slightly different speed, which to me makes "Data" sort of wonderfully, not hugely, but wonderfully simply neurotic tool to a degree. Now there was something sort of wonderful about realizing even robots aren't perfect.

Are you all right, Data? We have to warm up.

KNIGHT: It's not that no one knew that gesture had meaning before. What's new is that I don't think that many roboticists were thinking about tapping into this particular body of knowledge to apply them to robots before.

So, forging that connection actually makes whole new things possible in terms of the success of our social robots in general and so one of the goals of research is uncovering some of those unknowns.

Or uncovering new things, new ways to do things, which is often combinations of stuff that already exists. That's one of the beautiful things about looking across different fields. Sometimes there already exist methods.

For example, in biology or in my case obviously in the arts, that absolutely apply to the problem you're trying to solve if you just look outside of your own little circle.

ROBOT: Next stop is the story of how Heather and I met and know, it was not at a bar.

(COMMERCIAL BREAK)

KNIGHT: So, I -- that could be playful. So I'm a robot. I was born in a factory just outside of Boston. OK, not really. I'm a person, but I do get asked that sometimes. But yes, so I grew up in Lexington, Massachusetts, home of American Revolution.

My mother's family were Irish immigrants, so that's always fun having that sort of history and my dad is from Texas and he's an engineer. My mom speaks a million languages, works in non-profits most of life and really cares about impacting the world.

Yes, my dad's like super smart and like really an amazing engineer. Somehow I fell a little bit in between the two of them. My mom's really, really outgoing and my dad's more introverted, really cares about his family and stuff.

So I always wanted to make technology. I always like the engineering stuff. This idea that you can build something that has never existed before. You come up with something, poof. Not exactly, right? But yes, that's cool. That's an empowering thing.

But I also wanted to make sure when I went home, you know, my mom would be like, so, where is that technology going? But that it was something that people would care about, that people could understand, and that could have real impact so yes.

I liked making technology that makes people's lives better. I didn't fall in love with robots themselves until I was in college. I had been deciding in high school between being a writer and being an engineer.

Like I always liked that sort of creative process and it's just sort of interesting after like years and years and years that I am starting to actually get to do storytelling with robots. Some ways I'm getting to author characters that physically exist in the world and can do way more than just maybe what's on a page.

When I was a freshman at MIT, I decided I wanted to get a campus job and I went home to my like living group or whatever and someone was like, do you want to work in a robotics lab? I went into this lab and they were building these awesome social robots.

It's like, there was something about social robots that captured my imagination and like from the first project that we did in that lab, we had built this interactive sort of terrarium. There was a robot that didn't look like a person, almost had like a hand sort of thing at the top.

If there were people around suddenly that sort of hand becomes a face. Not that it had eyes, just in how it's acting and checks out the people along the way, right? Curious, very simple behavior system, but one that we get right away and so people would come by like felt a little bit Epcot Center or something.

So sometimes it will be like, they wave a lot, get too excited, comes too close, it gets scared and runs back, right. It was really simple, right. But if you design the behavior system like be that, like that's something there's no speech, like but it's totally you just understand right away.

It was really fun watching people interact with that. So from the very beginning, I always gotten to be able to create robots that people cared about and that were installed in places where I could watch that interaction.

Yes, that was really fun. And when you see people laugh or when you see their eyes light up, I mean, that definitely makes you want to keep doing what you're doing. Could I imagine a time in my life where I wasn't making social robots? It's really hard, right, like I tried. I did.

Like when I finished my masters, I like -- you know, got a job with NASA. That's cool. Like the jet propulsion laboratory outside of Los Angeles, and we're like building space systems and I was like, this should be really cool, but there's no people.

It's like not on another planet, but like out, you know, it's cool, it's interesting, I love that space can inspire people, but I had to come back.

GUPTA: Heather got the social robotics bug, so to speak. So she left NASA's Jet Propulsion Lab and launched into the world of social robotics full time. There she met "Data," and the rest, as they say, is history.

KNIGHT: I think maybe my story with "Data" is illustrative. I was working for the French company that creates these robots, El Deberon, and they basically lent me a robot for several months.

Basically, it was like that puppy syndrome, right. They're like you can just take it for a night. Like my -- you know, I was about to start grad school again, time for me to give the robot back. I kept pushing it off.

Call one week, another week, one month went by. Heather, we want our robot back. Eventually, I was like you know what, I can't give it away. I ended up buying it and then, you know, our little robot comedy duo was born.

ROBOT: Do you want to know if I will one day rule the world. Heather explains. Coming up next.

(COMMERCIAL BREAK)

KNIGHT: What I do is really fun. But you might be wondering, what's the big idea behind that? How is this going to impact the world?

I would say there are a couple things that are possible to play with when you start thinking about the social intelligence of machines. We can make applications for robots or for interfaces that are in our everyday life that are effective.

ROBOT: Heather, help me with my stylish scarf.

KNIGHT: I think that as designers, the idea of being able to take robots out into the wild and to do not all of our testing and standardized laboratory environments but out where the people are that we're designing for.

ROBOT: Ready. Show me a postcard.

KNIGTH: I'm very excited about our future where we have robots in our everyday life. I think entertainment robots are probably one of the first applications that could be fruitful because they can pay for themselves if you have a big audience of people on stage.

I think that artists use the medium of their time and right now that medium is technology. I think it's a new interface for being creative.

ROBOT: Times Square, home of the tourists. A lot of those in New York. Well now, any of you guys tourists?

KNIGHT: One of the things that's really important to me is to create technology that brings us together and doesn't divide us. I have this crazy idea that maybe we could come to a world where we replace not people by robots, but computers by robots.

Like how about making technology more human and I think in effect, that will also make us more able to fulfill our own objectives and connect with each other.

ROBOT: Have you seen the naked cowboy? He plays the guitar in his underwear in a cowboy hat. Sorry, I forget those props today, but he moves like this. Yes, just shaking his booty. Tourists love that guy.

KNIGHT: There's something about a cubicle, something about a screen and sitting in the same place all day, that is really new in modern society, and it's not really very natural. I don't know if that allows us to achieve everything that we might be able to.

But I would dream that there is possibility to still have that same access to the benefits of it technology, but by providing a different vehicle of delivering that. Technology is a device for accomplishing goals. It's something that makes life potentially easier or it makes life better.

In the end, I want to create applications for people that, for example, could empower the elderly to stay in their homes for longer in a way that's actually safe whether it's helping them with certain functions.

At the same time, increasing their freedom not like you have to take this pill right now. You know, it could be like to start have these variables where it's tracking like OK, they were supposed to do that, but they haven't done it for a while, maybe I can think about a different approach that will make them more likely to do that.

Robots have already been helping with the nuclear disaster cleanup in Japan. We're still in pretty early stages of development, but there were some robots that came to look for search and rescue victims at the 9/11 site. And there's been tons of investment since then, even for disasters like Hurricane Irene.

ROBOT: My audience, I have a confession. I am a tourist too. I've only been in the city a few months and I am French.

KNIGHT: I don't want to live in a world where people are replaced by robots, but I do believe that human robot teams can accomplish things that humans alone or robots alone would never be able to accomplish any time in the short term.

ROBOT: I am sorry if you feel betrayed, but I am actually comfortable with my condition. Even a robot can dream.

KNIGHT: In the end I think it's really about making people flourish.

ROBOT: Pretty inspiring stuff, thanks, folks. Sanjay, take it away. I have a hot date with a gal named Siri.

GUPTA: Thanks, Data. You know, Heather Knight is one of these innovators who is taking the field of robotics and making it really fun. She's also part of a unique group of people who are driven to do more with what they love doing.

Sometimes they find that passion quite by accident, and other times, it's as if they were born to do nothing else. In the end, though, they are all agents of change and that's what earns them a spot on this show.

For more on this episode and other agents of change, go to cnn.com/thenextlist and join me on my life's team, cnn.com/sanjay. It's a one-stop spot for all my videos, blogs, tweets and behind-the- scenes photos.

Thanks so much for watching. I'm Dr. Sanjay Gupta. Hope to see you back next Sunday for THE NEXT LIST.