The robotics engineer whose new creation will bring a touch of humanity into your home.
The robotics engineer whose new creation will bring a touch of humanity into your home.
We’re all a bit terrified of the rise of the robot, so accepting one into the family is surely a big ask. Cynthia Breazeal, a superstar American robotics engineer, wants to make us think differently. Having hankered after a real-life R2-D2 since she was 10, she created Kismet, Leonardo, Aida, Autom, Huggable — one helped you stick to a diet, another offered companionship, but none was available to take home. Now Cynthia, 48, has built Jibo — white, shiny, swivelly, somehow happy, and yours for just over £500. He might remind you to call your grandmother on her birthday (and then put in the call). He encourages you to lift your eyes from your screen and engage with others — himself included.
“Trespassers are subject to experimental brain surgery,” warns a notice on the fourth floor of MIT’s Media Lab, the most innovative corner of the world-renowned research institution in Cambridge, Massachusetts. It seems plausible. This is the MIT Media Lab, after all, where an astonishing number of the inventions we use every day – from touchscreens to GPS – were born.
The sign marks the entrance to the Personal Robots Group, where artificially intelligent computers are designed to respond to the quirks, needs and emotions of human beings. Inside, it is warm and cluttered, the space full of electronics and fuzzy creatures. They look like puppets, but many have metal components in lieu of limbs, or eyes that are clearly designed to move back and forth on their own. The lighting is low – not the bright-white fluorescence you expect in a lab – and the floor is covered in dark grey carpeting and white foam flooring onto which things can be projected. It feels like a tinkerer’s basement or, perhaps more accurately, a prop cupboard at the Jim Henson Company. Most of the people working here today are women.
I am meeting Cynthia Breazeal, the founder and director of the Personal Robots Group. When you hear about all the things Cynthia does, it is easy to assume she is some sort of android superhuman herself. In addition to running this lab she is an associate professor at MIT. She is the co-founder of a buzzed-about tech startup called Jibo. She is a public figure in the world of science – you might have seen her TED talk about the rise of personal robots. Plus, she is happily married with three young sons. And she doesn’t come off as even slightly frazzled by juggling all this.
Cynthia, 48, looks a bit like a brunette Meryl Streep, with a long, elegant nose and prominent cheekbones. She is slim and of average height and carries herself with purpose. She was brought up in California but has lived in Boston for more than 20 years and has clearly mastered the art of dressing for frigid winters: she is wearing knee-high suede boots and, under her puffy, all-weather red coat, has layered a plum-coloured fur vest over a plum-coloured jumper. Large geometric earrings dangle from her ears. We retire to her office, a rather anonymous space except for a few photos of her sons on a shelf. Her area of expertise, she explains, is social robotics, machines that are designed to interact with humans. “So much technology today is just data, data, data, and graphs and information,” Cynthia says. “But there is a very different kind of help we get from people, which is social support, emotional support and just feeling like there’s someone who’s in your corner with you. And social robotics is magical, because it actually brings those two worlds together for the first time in a technology.”
This MIT lab, she tells me, is the place where she oversees work that looks 10 or 20 years into the future. But recently she has been more focused on the next 10 or 20 months. Two years ago she took a leave of absence from her role as a professor (she returned to her post in January) to focus on a single robot called Jibo. “If I’m in academia, I could change the world of ideas,” she explains, “but I can’t bring something physical to the world.” At least not right away. So in 2012 she cofounded a startup with Jeri Asher, an entrepreneur, to design and sell a robot that would make sense in kitchens and living rooms today. The company launched a hugely successful crowdfunding campaign in 2014 and expects to start selling the robot this summer in the US and across Europe for about $800 (£560). Jibo, Cynthia says, “is the next logical step in my own life story,” because it builds on decades of her research in making personal – and personable – robots. Cynthia is the company’s founder and chief scientist, and, she says, “very much the face of the company”.
Jibo is Cynthia’s play at bringing social robots to the masses – and proving they can be warm and friendly. “If you took a really cool Pixar character and combined it with an iPad, that’s kind of what Jibo is,” she says. It has a round screen that sits on a pedestal. The screen can swivel around and tilt up and down but Jibo is meant to sit on a countertop or table – early users told Cynthia they liked the fact that Jibo remained stationary rather than following them around the house. Its abstract shape resembles a human head and shoulders and was inspired by the international symbols for men and women used on lavatories.
But Jibo’s voice is decidedly humanised: young, male and chipper – the opposite of Siri’s bossy schoolmarm tone. Cynthia wrote a long brief to her staff about who Jibo is as a character. “So, for instance, Jibo refers to people as, ‘Are you my person?’ He doesn’t say, ‘Are you my user? Are you my master?’” she says. “He’s trying to create this sense of, ‘I am this little critter coming into your home. I want to belong to this family, and I want to help out.’” The video that accompanied Jibo’s crowdfunding campaign shows the little robot filling half a dozen roles. It reads bedtime stories to the children. It speaks up to remind a family member of a forthcoming appointment. It swivels into the best position to snap photos on command. It suggests something to order for dinner.
Cynthia and her colleagues at Jibo are not the only ones hoping consumers will embrace personal robots. Autonomous – also available this summer, for $1,500, and also funded with an online campaign – is a robot with a mobile base, a long pole and an oval screen at the top displaying an animated character; it can switch on the coffee maker and navigate its way around the house. Then there’s Pepper, a robot with arms, hands and a smooth, white humanoid face that’s already available in Japan; considered a companion rather than a robot, it performs tasks and chores. And Furo, a “smart service robot” designed to provide customer service in places such as airports, has a mobile base, an animated face on a screen and a large flat screen underneath ($899). These robots target different types of consumers, but all offer similar features and employ AI technology like Jibo’s. And they are all designed to be used by average people rather than trained experts.
The thing that makes Jibo different is that it’s not just preprogrammed by the company. It is a developer platform, so third parties can create software to expand the possibilities for what Jibo can do. This makes the robot customisable and also presents a financial opportunity for the company. Jibo will have a store, similar to Apple’s App Store, which will allow users to download applications made by various developers. Cynthia and her team have created a small number to start with, she explains, “like when the Wii or when the iPhone was launched. You had to start it off with the core set of games or apps so that the world could understand why this is so different.”
Inspired by science fiction, many people have long dreamt about having robots at home. But as AI technology has evolved over the past decade it has mainly shown up on our smartphones. We have grown accustomed to saying “Hey Siri”, and most of us have stopped expecting that life in the 21st century should mean having a robot gliding into our bedroom to offer us coffee. For busy people juggling responsibilities, the smartphone has become the most essential tool – which is why it can be tough to imagine how a vaguely humanoid tabletop robot could wedge its way into the daily routine. After all, what most of these social robots promise to do – take photographs, order pizza, play games with the children – are things smartphones seem to do quite well already.
Cynthia argues that phones are individual devices, whereas Jibo is designed for the whole family to use. In addition, the team did focus-group tests with carers – often women looking after their own children as well as ageing parents. “They are the most stressed-out people on the planet,” Cynthia says. “They all had smartphones. They all used technology. They didn’t use it for any of this coordination stuff with their family.” Jibo is designed to respond to the chaos in the home. In Cynthia’s mind he’s like the family dog – if the dog made things easier for everyone. “I’m in the kitchen in the morning routine. I’m making lunches. I’m doing dishes. I’ve got kids Velcroed to my leg. I don’t have time to find my stupid computer or iPad and take it out and check the weather. But I could say, ‘Jibo, what’s the weather?’” Clearly this is not hypothetical – she has experienced it. One reason she wanted Jibo to be an excellent photographer was that she was sick of being behind the camera rather than in family photographs. Jibo was designed in response to her own struggles. “I joke that a long, long time ago, in a galaxy far, far away, these robots were called droids,” Cynthia says. “In our galaxy, at this time, they are called social robots.”
It all started with Star Wars. Cynthia saw the original film when she was 10, growing up in Livermore in northern California. The robots – R2-D2 in particular – were her favourite characters. At the time both her parents were working as computer scientists in government labs. The family was one of the first to have a personal computer, and her parents frequently took her to science events. Her brother, who is two years older, studied physics. It was, she says, “a very high-IQ kind of family.” And Cynthia fitted right in. “I was not the rebellious kid,” she says. “I was driven. Whatever it was, I just wanted to be great at it.” She excelled at school and was a serious athlete, competing in football, tennis and athletics. “I found that it created a common ground with a lot of the guys,” she says. Sport later proved to be excellent training for being the only woman in a research lab, she says. “Even if I was in a male-dominated field, I never felt like I didn’t belong there.”
Cynthia applied to the University of California, Santa Barbara, in 1985, intending to become a doctor. Her parents convinced her to focus on engineering first and apply to medical school later; they told her a degree in electrical and computer engineering would keep more options open than pre-med. They were proved right. Cynthia’s undergraduate years at UC Santa Barbara coincided with the opening of a new robotics lab on campus. Engineering classes combined with her childhood Star Wars obsession led to new inspiration. “I decided what I wanted to do was be an astronaut,” she says. Cynthia focused on space robotics and was accepted on a postgraduate programme at MIT in 1992.
The course was led by the renowned roboticist Rodney Brooks, who would go on to become Cynthia’s mentor and one of her greatest champions. They focused on building small, light robots to work in the farthest reaches of space without direct human guidance. “When I first came into the lab, it was like that first Star Wars moment all over again,” she says. “I saw these little robots and thought, ‘My God, if we’re ever going to see robots like R2-D2 in the future, it’s going to start in a lab like this. In fact, it might start in this lab.’” She threw herself into building robots that could move over rough terrain and through other hostile conditions in space. She wrote her master’s thesis on the subject, and in 1993 she completed a fellowship at Nasa’s Jet Propulsion Laboratory in Pasadena, California.
Meanwhile, Brooks had taken a sabbatical to visit robotics labs around the world. By the time he returned to MIT, he was convinced the future was humanoid robots: machines that walked, talked and interacted as we do. The MIT lab shifted direction to make robots designed to move like humans – think of Rosie, the Jetsons’ robot maid, or C-3P0 from Star Wars. Cynthia, by then a senior postgraduate student, became one of the chief architects of Cog, an experimental attempt to design a robot which had the same physical capabilities as an infant.
At the time, researchers were scrambling to build robots that could pick things up and walk down stairs. “But no one was working on what it would actually mean to live with these technologies,” Cynthia recalls. In other words, researchers were focused on whether Cog could hand an item to a person, not how it could interpret whether the person wanted that item in the first place.
Cynthia began to re-evaluate her research more seriously in 1997 after Nasa landed a robot on Mars. “I thought, This is incredible, a huge success for all robots,” she tells me. “We can explore the oceans, we can send them down volcanoes, we’re even sending them to Mars. But they’re not in our living rooms yet. Why is that?” This was her epiphany.
The transformative power of robots wouldn’t come from their ability to explore uncharted territory or travel to space. The value of R2-D2 and C-3P0 lay in their relationship with humans. They were easy to use – not just by Nasa employees or highly skilled manufacturing workers but by everyone. That was because they knew how to respond to humans. It dawned on Cynthia that the final robotic frontier was not, in fact, space. It was the home.
“Robots are never going to be human; that’s not the point. The magic of this technology is how it complements and empowers us.”
This was a fairly radical point of view. Most of her peers were still focused on movement, motor skills and navigation. “Our behaviour is governed by mind and emotions and thoughts and feelings. And now you’ve got to build a robot to respond to this, and no one was working on that,” she says. She went to Brooks and told him she was shifting her focus. “I’m like, ‘Rod, I have to stop everything I’m doing.’” That’s when she started work on Kismet, her first social robot. “I still have the video that was Cynthia’s turning point,” Brooks tells me via email. “We had Cog running with a left arm and hand, and just the beginnings of capabilities, and Cynthia and Cog were manipulating the same dry board eraser. Soon they were taking turns. Wondering how this could be, she started reading psychological literature about interactions between mothers and babies and discovered a phenomenon where mothers lead infants in an activity, not realising they are doing the new stuff but thinking the baby is. This allows the baby to learn what the mother thinks it already knows. Cynthia surmised this was what she had been doing with Cog. That led her to explore ‘social scaffolding’ in her PhD thesis.”
It was around that time that she met her husband, Robert Blumofe, who was also a postgraduate student at MIT – and the grandson of the comedian Jack Benny. A computer scientist, he is now an executive vice president at Akamai, a tech company that works with businesses to ensure fast, secure web browsing. “Having a husband who truly revels in the fact that I’m so successful is really key,” she says. This sentiment is echoed by other powerful women in tech – Sheryl Sandberg has written that who you marry is one of the most important career decisions women can make. Cynthia agrees. “I think if you marry the right person you can do anything.” Blumofe is a sounding board for her ideas, her biggest cheerleader, and an equal parent who is just as likely as Cynthia to pick up their boys, Ryan, Nathan and Caleb, from school. Also important is her right-hand woman and loyal assistant, Polly Guggenheim, an outgoing woman in her 60s who has been at the Media Lab almost as long as Cynthia. “Polly makes all of this possible for me at MIT, and she is a dear, wonderful friend,” Cynthia says. “Alfred to my Batman, Samwise Gamgee to my Frodo.”
In the years since Kismet, Cynthia has been involved in the development of several robots that are programmed to assist human beings as they go about their daily lives. “We see robots as a teammate,” Cynthia says. Kismet was a robotic head with visible metallic components but cartoonishly lifelike eyes, eyebrows and a mouth; Cynthia designed it to recognise and simulate emotions. Leonardo, a furry robot that resembled a Gremlin, was created in 2002 with Stan Winston, a Hollywood special-effects make-up expert. Leonardo was able to recognise faces, produce a range of expressions and respond to touch. These early robots were steps towards figuring out how to program a machine to engage easily with humans.
Later, Cynthia oversaw the creation of robots that had more specific purposes, such as Autom, which helps people stick to their diet and exercise regimes, and Aida, a driving assistant. Postgraduate students are currently refining others, such as Huggable, which looks like a teddy bear and enables doctors and nurses to communicate with children remotely and interact with them in hospital when no one else is around. None of these, though, was ever made commercially available by the lab.
When it came to Cynthia’s focus on social robots, the timing was right. “If I’d had this idea 10 years earlier, it wouldn’t have got any traction,” she says. But by the early 2000s, many countries were becoming worried about their ageing populations. Take Japan. In 1980, 9 per cent of its population was over 65; today that figure is 25 per cent.
By 2055, it will be nearly 40 per cent. Social robots were touted as one way to handle the predicted elderly-care crisis: they could be designed to remind people to take their pills, say, or to help them in the home.
Brooks started talking Cynthia up to others in their field. “Rod Brooks would talk, everyone would listen and he would credit me,” she says. “It was huge having someone of that rock-star status giving validity to this crazy work that I was doing.” With Kismet, he actively encouraged her to take on the establishment. “Our research group were crazy and irreverent, and individual students got to thumb their noses at the status quo and stodgy academics. Cynthia took up that mantle with great relish,” Brooks says. “I was never worried that Kismet would take years to build. I knew something interesting and profound would come of it.”
Yet Cynthia’s approach to artificially intelligent robotics drew a number of detractors. She was criticised by Joseph Engelberger, who invented the first industrial robot in the 1950s, and John McCarthy, who was a pioneer in the field of AI and coined the term. “They didn’t understand the value of social-emotional robots,” she says. “They thought it was just about task: tell the robot what to do, and it will do it for you.” They thought robots’ greatest promise was as automated slaves, not as helpers adapting to and supplementing humans. What good was a robot that could smile?
Much of the general public has other fears about AI. For all the theoretical excitement about the possibility of robot helpers, more advanced AI has been identified as an existential threat to the human race in the long term – and human jobs in the short. Martin Ford, the author of Rise of the Robots: Technology and the Threat of a Jobless Future – last year’s winner of the Financial Times and McKinsey Business Book of the Year Award – says, “Personally, I continue to believe technology is a positive force. However, we will need to adapt our economy and society to ensure that everyone benefits from future progress. If we don’t do that, there is a risk that technology will result in soaring inequality, with only the wealthy few really thriving and many other people struggling to make a living.” More entrenched technophobic criticism is evident in movies like Ex Machina and Her, which imagine a future in which artificially intelligent robots quickly outpace humans, usually with terrible consequences. Then there are more immediate concerns, such as the possibility of a nefarious outsider hacking into a home robot to spy on or steal from its unwitting owners.
Cynthia doesn’t fear a future full of AI-enabled robots, partly because she is not trying to engineer a machine with human-level intelligence. “They’re never going to be human; that’s not the point,” she says. “The magic of this technology is how they complement us and empower us.” Cynthia will probably have to continue stating her case until someone invents a robot that makes humans feel empowered rather than threatened. Jibo, she hopes, will do just that
Cynthia’s vision of robots as not only helpers but also friends, she thinks, has something to do with her gender.
“I don’t do social robots because I’m a woman, but I would certainly say because I’m a woman and a mother and a technologist and a designer, it’s why I do the work I do,” she says. And this is one reason why diversity in robotics labs is so important to her. “You create technology that speaks to you and your life experience and what matters to you,” she says, “and if you only have a very narrow subset of the population creating technology, you’re leaving huge opportunities out.”
Cynthia’s desire to turn her theoretical ideas into something concrete and marketable started to become a reality in 2012 when she was seated next to Jeri Asher at a gala fundraiser for the Boston Philharmonic. “Cynthia’s a beautiful woman and came dressed very exotically,” Asher tells me. Soon they were talking business. “She was reaching for bringing something into the world – into the home – that hadn’t existed before.” The initial crowdfunding campaign in 2014 raised $2.3 million, and the company has gone on to raise another $50 million in investment, including $27 million from venture capitalists interested in bringing Jibo to China, Korea and Japan. Today it has 60 employees.
Asher, who Cynthia describes as the “tactical, practical” half of the business, remembers observing Cynthia on one particularly busy day. “She had just driven to parents’ day or some kind of activity at school. Then she had to run to the office. She had leather pants on, Gucci platform mules and a great-looking sweater. She was in this series of meetings, then going to MIT to work with PhD students and teach class. She had done three things in a five-hour period that most people couldn’t balance in a week. She was unflappable.” Cynthia doesn’t exactly make it look easy. But she does make it look possible – especially with the help of a few emotionally intelligent robots. It’s important, she tells me, for younger women in tech fields to see women who are professionally successful and personally happy. “We’re past the idea of women having to dress and act like men,” she says. And, she hopes, past the point where we’re afraid to accept robots into our lives.