seeing "Star Wars" for the first time,

I've been fascinated by this idea

of personal robots.

And as a little girl,

I loved the idea of a robot that interacted with us

much more like a helpful, trusted sidekick --

something that would delight us, enrich our lives

and help us save a galaxy or two.

I knew robots like that didn't really exist,

but I knew I wanted to build them.

So 20 years pass --

I am now a graduate student at MIT

studying artificial intelligence,

the year is 1997,

and NASA has just landed the first robot on Mars.

But robots are still not in our home, ironically.

And I remember thinking about

all the reasons why that was the case.

But one really struck me.

Robotics had really been about interacting with things,

not with people --

certainly not in a social way that would be natural for us

and would really help people accept robots

into our daily lives.

For me, that was the white space; that's what robots could not do yet.

And so that year, I started to build this robot, Kismet,

the world's first social robot.

Three years later --

a lot of programming,

working with other graduate students in the lab --

Kismet was ready to start interacting with people.

(Video) Scientist: I want to show you something.

Kismet: (Nonsense)

Scientist: This is a watch that my girlfriend gave me.

Kismet: (Nonsense)

Scientist: Yeah, look, it's got a little blue light in it too.

I almost lost it this week.

Cynthia Breazeal: So Kismet interacted with people

like kind of a non-verbal child or pre-verbal child,

which I assume was fitting because it was really the first of its kind.

It didn't speak language, but it didn't matter.

This little robot was somehow able

to tap into something deeply social within us --

and with that, the promise of an entirely new way

we could interact with robots.

So over the past several years

I've been continuing to explore this interpersonal dimension of robots,

now at the media lab

with my own team of incredibly talented students.

And one of my favorite robots is Leonardo.

We developed Leonardo in collaboration with Stan Winston Studio.

And so I want to show you a special moment for me of Leo.

This is Matt Berlin interacting with Leo,

introducing Leo to a new object.

And because it's new, Leo doesn't really know what to make of it.

But sort of like us, he can actually learn about it

from watching Matt's reaction.

(Video) Matt Berlin: Hello, Leo.

Leo, this is Cookie Monster.

Can you find Cookie Monster?

Leo, Cookie Monster is very bad.

He's very bad, Leo.

Cookie Monster is very, very bad.

He's a scary monster.

He wants to get your cookies.

(Laughter)

CB: All right, so Leo and Cookie

might have gotten off to a little bit of a rough start,

but they get along great now.

So what I've learned

through building these systems

is that robots are actually

a really intriguing social technology,

where it's actually their ability

to push our social buttons

and to interact with us like a partner

that is a core part of their functionality.

And with that shift in thinking, we can now start to imagine

new questions, new possibilities for robots

that we might not have thought about otherwise.

But what do I mean when I say "push our social buttons?"

Well, one of the things that we've learned

is that, if we design these robots to communicate with us

using the same body language,

the same sort of non-verbal cues that people use --

like Nexi, our humanoid robot, is doing here --

what we find is that people respond to robots

a lot like they respond to people.

People use these cues to determine things like how persuasive someone is,

how likable, how engaging,

how trustworthy.

It turns out it's the same for robots.

It's turning out now

that robots are actually becoming a really interesting new scientific tool

to understand human behavior.

To answer questions like, how is it that, from a brief encounter,

we're able to make an estimate of how trustworthy another person is?

Mimicry's believed to play a role, but how?

Is it the mimicking of particular gestures that matters?

It turns out it's really hard

to learn this or understand this from watching people

because when we interact we do all of these cues automatically.

We can't carefully control them because they're subconscious for us.

But with the robot, you can.

And so in this video here --

this is a video taken from David DeSteno's lab at Northeastern University.

He's a psychologist we've been collaborating with.

There's actually a scientist carefully controlling Nexi's cues

to be able to study this question.

And the bottom line is -- the reason why this works is

because it turns out people just behave like people

even when interacting with a robot.

So given that key insight,

we can now start to imagine

new kinds of applications for robots.

For instance, if robots do respond to our non-verbal cues,

maybe they would be a cool, new communication technology.

So imagine this:

What about a robot accessory for your cellphone?

You call your friend, she puts her handset in a robot,

and, bam! You're a MeBot --

you can make eye contact, you can talk with your friends,

you can move around, you can gesture --

maybe the next best thing to really being there, or is it?

To explore this question,

my student, Siggy Adalgeirsson, did a study

where we brought human participants, people, into our lab

to do a collaborative task

with a remote collaborator.

The task involved things

like looking at a set of objects on the table,

discussing them in terms of their importance and relevance to performing a certain task --

this ended up being a survival task --

and then rating them in terms

of how valuable and important they thought they were.

The remote collaborator was an experimenter from our group

who used one of three different technologies

to interact with the participants.

The first was just the screen.

This is just like video conferencing today.

The next was to add mobility -- so, have the screen on a mobile base.

This is like, if you're familiar with any of the telepresence robots today --

this is mirroring that situation.

And then the fully expressive MeBot.

So after the interaction,

we asked people to rate their quality of interaction

with the technology, with a remote collaborator

through this technology, in a number of different ways.

We looked at psychological involvement --

how much empathy did you feel for the other person?

We looked at overall engagement.

We looked at their desire to cooperate.

And this is what we see when they use just the screen.

It turns out, when you add mobility -- the ability to roll around the table --

you get a little more of a boost.

And you get even more of a boost when you add the full expression.

So it seems like this physical, social embodiment

actually really makes a difference.

Now let's try to put this into a little bit of context.

Today we know that families are living further and further apart,

and that definitely takes a toll on family relationships

and family bonds over distance.

For me, I have three young boys,

and I want them to have a really good relationship

with their grandparents.

But my parents live thousands of miles away,

so they just don't get to see each other that often.

We try Skype, we try phone calls,

but my boys are little -- they don't really want to talk;

they want to play.

So I love the idea of thinking about robots

as a new kind of distance-play technology.

I imagine a time not too far from now --

my mom can go to her computer,

open up a browser and jack into a little robot.

And as grandma-bot,

she can now play, really play,

with my sons, with her grandsons,

in the real world with his real toys.

I could imagine grandmothers being able to do social-plays

with their granddaughters, with their friends,

and to be able to share all kinds of other activities around the house,

like sharing a bedtime story.

And through this technology,

being able to be an active participant

in their grandchildren's lives

in a way that's not possible today.

Let's think about some other domains,

like maybe health.

So in the United States today,

over 65 percent of people are either overweight or obese,

and now it's a big problem with our children as well.

And we know that as you get older in life,

if you're obese when you're younger, that can lead to chronic diseases

that not only reduce your quality of life,

but are a tremendous economic burden on our health care system.

But if robots can be engaging,

if we like to cooperate with robots,

if robots are persuasive,

maybe a robot can help you

maintain a diet and exercise program,

maybe they can help you manage your weight.

Sort of like a digital Jiminy --

as in the well-known fairy tale --

a kind of friendly, supportive presence that's always there

to be able to help you make the right decision

in the right way at the right time

to help you form healthy habits.

So we actually explored this idea in our lab.

This is a robot, Autom.

Cory Kidd developed this robot for his doctoral work.

And it was designed to be a robot diet-and-exercise coach.

It had a couple of simple non-verbal skills it could do.

It could make eye contact with you.

It could share information looking down at a screen.

You'd use a screen interface to enter information,

like how many calories you ate that day,

how much exercise you got.

And then it could help track that for you.

And the robot spoke with a synthetic voice

to engage you in a coaching dialogue

modeled after trainers

and patients and so forth.

And it would build a working alliance with you

through that dialogue.

It could help you set goals and track your progress,

and it would help motivate you.

So an interesting question is,

does the social embodiment really matter? Does it matter that it's a robot?

Is it really just the quality of advice and information that matters?

To explore that question,

we did a study in the Boston area

where we put one of three interventions in people's homes

for a period of several weeks.

One case was the robot you saw there, Autom.

Another was a computer that ran the same touch-screen interface,

ran exactly the same dialogues.

The quality of advice was identical.

And the third was just a pen and paper log,

because that's the standard intervention you typically get

when you start a diet-and-exercise program.

So one of the things we really wanted to look at

was not how much weight people lost,

but really how long they interacted with the robot.

Because the challenge is not losing weight, it's actually keeping it off.

And the longer you could interact with one of these interventions,

well that's indicative, potentially, of longer-term success.

So the first thing I want to look at is how long,

how long did people interact with these systems.

It turns out that people interacted with the robot

significantly more,

even though the quality of the advice was identical to the computer.

When it asked people to rate it on terms of the quality of the working alliance,

people rated the robot higher