CNN [동영상] Your new best friend may be a robot

[Daniel]

The provocative idea worth spreading in the talk I gave at the TEDWomen conference is that robots are a really intriguing social technology. This idea has emerged from pursuing my childhood dream of creating personal robots that are a meaningful, empowering and enriching part of our everyday lives

Your new best friend may be a robot

STORY HIGHLIGHTS

• Cynthia Breazeal: Robots have the potential to be social creatures
• She says they can play a role in education, health care, business
• Breazeal: "Robots touch something deeply human within us"

(CNN) -- The provocative idea worth spreading in the talk I gave at the TEDWomen conference is that robots are a really intriguing social technology. This idea has emerged from pursuing my childhood dream of creating personal robots that are a meaningful, empowering and enriching part of our everyday lives.

Today, when we think of personal technologies, we think of things such as smartphones and laptop computers. When we think of social technologies, we think of things such as Twitter and Facebook.

Today, social technology helps people be social (e.g.,interact, communicate, collaborate, share, recommend, etc.). In the future, social technology may BE social, too -- in the form of social robots. My talk touches on a few different dimensions of what "social" means in this context:

1.Social-psychological interface

Robots "push our social buttons." As humans, we are deeply and profoundly social. It is a great strength of ours. We think about and understand the social world in different ways than the physical world.

It turns out that robots can be designed to exhibit behaviors that engage our social brain. We are subconsciously compelled to try to understand and relate to social robots as beings, not things -- even while fully knowing that they are a (special) kind of machine.

For instance, if robots share our nonverbal cues (gestures, facial expressions, body posture, etc.) and use them appropriately, then we respond to these cues, use them to form social judgments and adjust our behavior based on these judgments -- as we would predict when interacting with people.

Understanding this insight opens new possibilities for robots as a personal and ubiquitous technology where the robot's social attributes are a key part of its functionality and help us achieve personal goals. Possible indispensable applications -- or "killer apps" -- for social robots could be in the domains of health, eldercare, education, telecommunication and entertainment.

2.Social-emotional intelligence

This is the flip side of the first point. We may respond to social robots in social-psychological terms, but social robots then need to hold up their end of the interaction.

One example is the tremendous scientific challenge of endowing robots with what psychologists call Theory of Mind. In short, how can we develop robots that can understand and predict our behavior beyond our observable actions, to infer the mental states that underlie our behavior (e.g., our goals, intentions, beliefs, attitudes, emotions, motivations, etc.)?

TED.com: Dale Dougherty says 'We are makers'

This core social competence underlies our ability to understand one another, communicate, collaborate, learn from each other and so much more. It is key for social robots that can collaborate with us as a partner -- by anticipating our needs, coordinating their actions with ours to achieve a shared goal, understanding what matters to us, learning from watching us demonstrate a task and more.

This is the kind of stuff those droids in "Star Wars" could do. You didn't have to explicitly program them or tell them step by step how to do something. They understood what you meant. They also treated people as people -- simply put, the way we wish to be treated. Robots today have a long way to go on this front.

3.Social medium

I think one of the closest killer apps on the horizon for social robots is to transform screen-based telecommunication to real world teleinteraction. It is intriguing to harness the physical and social embodiment of robots to enable people to interact much more richly with one another over distance. In the talk, I present our research with the MeBot that shows how physical and social co-presence (mediated through a robot) really does matter to people.

As a mother of young children, I introduce the idea of social robots as a distance-play technology -- and I mean play in the real world with kids running around, sharing books, playing with toys and the lot. I think there could be real demand for this for grandparents who live far from their grandchildren to really play with their grandkids despite distance.

It's a struggle for me to have my own kids talk to their grandparents on the phone or using Skype for more than just a couple minutes. They don't want to talk with their grandparents ... they want to play with them! And it doesn't take much imagination to leverage the same idea for children's hospitals where friends and family could "jack into" a teddy bear bot to keep a sick child company or for doctors to make "robotar" house calls to patients, or educators to take distance learning to a whole new level.

TED.com: Life in Biosphere 2

Or thinking further out -- imagine how a construction worker in India can jack into a construction robot to work a project in Japan and never have to get on a plane. Imagine if the robot could even do all the language and social protocol translation. Go ahead, let your imagination run wild: What will change when the internet gets a body?

4.Social agent

Social robots can function autonomously as well, to be social agents that collaborate with people as partners. As social agents, robots are intriguing because they can straddle the realm of technology and the world of people. In my talk I present research on a weight management robot coach that helps you to sustain engagement in a diet and exercise program (this is being commercialized now, see Intuitive Automata).

The robot's social properties enable it to build a successful working alliance with people -- to motivate them to stay engaged and help them set goals and track progress toward those goals. Imagine a future version where the robot can help simplify the logistics of tracking how much exercise you get or how much you weigh through managing your gadget network of wireless scales, pedometers and the like.

It could also help you to connect with your network of care professionals such as physicians, nurses, dieticians and exercise trainers, so that they can be better informed about your behavior and progress to provide better care and advice. The robot might even be a node in your extended social network of friends and family to help connect you with others who share your goals -- potentially to even help "nudge" the group to healthier behavior. Maybe the robot has a virtual extension on a mobile device so when you are on the go, it can be ready-at-hand.

Now, think about how you could apply these ideas to self-management tools for chronic disease that could bring a more personalized, high-touch experience to the patient that dramatically improves engagement and outcomes. That could be transformative.

5.Social design

This last point is a new direction I'm exploring in my research group at the Media Lab. What if social robots are designed by "the crowd"? One challenge of designing social robots is that they have pretty limited interaction with only a few people if they are stuck in the lab.

What if you could harness the internet to have lots and lots of people interact with these social agents, perhaps in the context of an online game? And from these interactions, the robots could learn more generalized and robust models of interaction that are fundamentally based on human behavior.

Robots (online or in the real world) could gather and share these experiences and interaction models to become more versatile, adaptive, robust and capable. I call this "crowd sourcing human-robot interaction." It is one approach my group is taking to explore long-term interaction with robots.

A related question is how to design for long-term interaction with a particular person where the robot really gets to know you as a unique individual. This is really going to matter if we will be living with personal robots in the future. Stay tuned ...

Robots touch something deeply human within us. Broadly speaking, I want to understand how to design technologies that support our valued human qualities that make life meaningful and worth living.

Technology should not "dehumanize" us, or make us feel or behave like machines in order to interact with it -- a concern and criticism I hear expressed too often. It should not isolate us or interfere with our ability to relate to or empathize with others.

I want a future where our technology deeply supports our ability to attain our highest and best selves for each other and our planet. Information and decision making are wonderful. I want to design personal technologies that also have "heart" so that we become the kind of people we truly aspire to be.

The opinions expressed in this commentary are solely those of Cynthia Breazeal.

Links referenced within this article

TED.com: Dale Dougherty says 'We are makers'
http://www.ted.com/talks/dale_dougherty_we_are_makers.html
TED.com: Life in Biosphere 2
http://www.ted.com/talks/jane_poynter_life_in_biosphere_2.html
Intuitive Automata
http://www.intuitiveautomata.com/

Find this article at:
http://edition.cnn.com/2011/OPINION/02/13/breazeal.social.robots/index.html?hpt=C1

[Daniel]

TED is a nonprofit organization dedicated to "Ideas worth spreading," which it distributes on its website. Cynthia Breazeal is an associate professor of media arts and sciences at the Massachusetts Institute of Technology and director of the Personal Robots Group at the Media Lab. She is the author of "Designing Sociable Robots."

[Daniel]

By Cynthia Breazeal, Special to CNNFebruary 13, 2011 -- Updated 2327 GMT (0727 HKT)