51¶ŻÂț

Move over darling

Bored with boys' toys? Technology needs a makeover, and Eugenie Samuel knows just the women to do it

AS I approach Justine Cassell’s office, I can hear giggling. Originally trained in linguistics, Cassell is now a professor at the Massachusetts Institute of Technology’s renowned Media Lab, applying her knowledge of human communication to improving the way we interact with computers. But it’s obvious she doesn’t quite fit the computer nerd stereotype. On her door hangs a picture of a dancing Barbie doll. I have to breach a makeshift dog gate to get into the office, and once I’m inside, her fluffy collie Esme rubs up against my legs. Having shaken off Esme’s attentions, I look up to find a variety of faces—some human, some definitely not—smiling at me. I’m meeting the designers and the designs for the next generation of human-computer interfaces. And gosh, they’re girlie.

No one ever said you had to be a bespectacled man to design ground-breaking technology but—until recently—it helped. If Cassell’s office is anything to go by, that is set to change. It’s full of cute and cuddly ideas, dreamed up by a new technological elite that is not ashamed to be feminine.

I bet I’ve raised some hackles already. Why should the gender of a technologist make any difference to what he or she designs? Well, technology design isn’t pure science. The personality and ideals of the designer strongly affect it—and there, gender plays a role. If you don’t believe me, ask Cornelia Brunner at the Center for Children and Technology in New York. She investigates gender differences and similarities. In one study, she asks boys and girls to design a machine of their choice. The results are striking, she says. Girls generally design flexible helpers for tasks—for example, the “season chore-doer”, a machine that clears snow in the winter and leaves in the autumn. Boys, on the other hand, typically invent powerful vehicles that do outlandish things like taking them instantaneously wherever they wanted to go, with names like “the twin valve seven booster class 4 rocket”.

Brunner finds the same patterns in adults. In one study, she ran computer interviews with 60 men and women working in a range of high-tech fields. One of the questions asked people to describe their ideal technological instrument. Women went for small flexible objects that facilitate social contact: jewellery that enables people to locate each other, for example. Men were more interested in transcending the physical limitations of their brains and bodies. They imagined things like having a socket in the back of their head that connects them to the whole of human history. Brunner has carried out research into gender divisions in a range of issues: attitudes to career, money, family and so on. “We get a certain amount of overlap on all issues—except the technological fantasy questions, where we get enormously clear differences,” she says.

Her finding matters because women are now penetrating the world of technology design to the point where they can bring their fantasies with them. Taking computer science as a representative discipline, figures from the National Center for Education Statistics in Washington DC show the percentage of computer science doctorates awarded to women in the US has almost doubled in 15 years: these days, 20 per cent of the recipients are female. This is still a minority, but unlike the first generation of female technologists, today’s women have the advantage of female role models and tend not to suppress their femininity in order to make it in their chosen industry.

Take Cassell’s student Catherine Vaucelle, for example. Enter her website and you’re greeted by butterflies, hearts, flowers and a bluebird sitting on a pink mailbox. And when, in Cassell’s office, I meet visiting Mattel toy designer Emily Kelly, she’s sporting a black lace cap, matching skirt and bright red locks. She unabashedly provides evidence that you can be techie and girlie at the same time. “I like dolls,” she tells me. “I like sparky things.” Then she giggles, as if I need any more convincing.

So what does this mean? What happens when women finally get the chance to design technology that interests them? It means that the design of toys, games, mobile phones and computers—anything that might contain a microchip—is going to be approached from a whole new direction.

The result of this is gadgets that have a wider appeal than anything ever designed before. Take Kismet, for example. This cute, blue-eyed machine has only one function: to pull faces in response to the person looking at it. “There’s no way in a million years a man would have designed something like Kismet,” says Eliot Mack, an engineer at iRobot, based not far from the MIT Media Lab in Cambridge, Massachusetts. He thinks Kismet is terrific, he adds quickly.

Kismet’s designer is Cynthia Breazeal, a colleague of Cassell’s at the Media Lab. She used to work on more macho projects, such as developing autonomous roving robots. But after several years she decided to do something different. “I saw the Apple Mac advert—’computers for the rest of us’. Well, I decided that I would make robots for the rest of us,” she says.

By that she doesn’t mean solely for women. Most of the funding for robotics comes from the military or NASA, so almost all robots are designed simply for getting dangerous and difficult stuff done. Aesth-etics and social skills are nowhere on the design criteria. “But if the field was genuinely open, you’d have all kinds of people, and all kinds of robots,” she says.

Breazeal’s inspiration for Kismet came from her conviction that if untrained “regular folk” were ever going to have robots, their natural instinct would be to treat them like pets. So she decided to see whether she could use this to her advantage: could a robot’s learning processes be helped by this kind of social interaction?

Developmental psychologists believe that children begin to learn language as they interact with their parents, who are trying to interpret their offspring’s random gestures and grimaces. Kismet has movable facial features that can express basic emotional states that resemble those of a human infant, and can make it clear whether it needs more or less stimulation. Most recently, Kismet has been endowed with basic voice recognition, a squeaky voice and a desire to communicate. Although it’s not clear if the robot is intelligent—in robotic terms—its behaviour is sufficiently complex to give the impression that it is learning how best to interact and communicate with people. Kismet has shown that it possible to get even the gruffest humans to project feelings and intelligence onto an inanimate object. “People find it so compelling they’re almost suspicious of it,” says Breazeal.

Breazeal is not convinced by Mack’s assertion that a man couldn’t have come up with the Kismet concept, but she does think that changing the social make-up of technologists will have a positive, and possibly surprising, effect on the way technology is designed. Female input to military simulations, for example, is changing the way the US military use their technology (see “Talk ’em down, don’t shoot’em up”). And Cassell’s lab at MIT backs that idea up too: the toys and interfaces I found there are like nothing I’ve ever seen before.

Cassell’s projects include Animal Blocks, a magical book that prompts children to improve their stories as they write them into a word processor, and Sam, a virtual friend who comments as children play with a toy castle, encouraging them to concentrate on essential literacy skills. For adults, she’s produced Rea, a virtual real-estate agent with body language who quizzes people about their needs and understands their verbal and non-verbal responses well enough to respond and show them around properties.

Vaucelle, who works just around the corner from Cassell, is also interested in designing interactive storytelling toys that help young children, both boys and girls, to develop their story-telling skills. She’s come up with “Dolltalk”, in which an alien doll called Zia learns about Earth by watching and talking to children as they play with stuffed dolls. The tangled speech of young children is voice-recognition software’s worst nightmare, so instead of trying to understand the children, Zia works by cutting and pasting what children say, warping the voice, and playing it back to them. The feedback prompts children to give more explanation about what’s happening in the game, helping to improve their cognitive processes and their command of language.

I spent a pleasant few minutes playing with Dolltalk and explaining some things to Zia. It dropped my voice a few octaves, slowed me down, and cut and pasted what I said into passable grammatical sentences of its own. It made me laugh, but I still have a complaint: with its stuffed dolls and sappy name, Dolltalk is just too girlie for my taste. I’m one of the many women who remember pulling Barbie’s head off as a girl.

Vaucelle is aware that not all girls are girlie. Since my visit she tells me she has changed Dolltalk, allowing children to create their own puppet monsters to play with. Despite the lab’s feminine frills, creating entertainment and educational technology that is free of any obvious gender is high on these researchers’ list of priorities. Although technology has almost always been designed exclusively by men for men, Cassell doesn’t think the right response is just to create technology by women for women.

Making things equally appealing to boys and girls is not a new idea to the makers of toys and games: they want to widen the potential market as much as possible. But until fairly recently, companies have often relied on feedback from customer focus groups to get an idea of what female-friendly products might be. It’s a strategy that’s failed: focus groups cut no ice with male engineers, Kelly says. When women in focus groups describe what they want, male engineers are liable to shake their heads and suck their teeth. If it’s not an immediately solvable problem, and they don’t understand the motivation for solving it, they tend to say it’s impossible—or at least impossible to do cheaply.

What is needed is female engineers who, like Cassell, Vaucelle, Breazeal and Kelly, are willing to stand out from the crowd. When Sony decided to bring out girlier versions of its Aibo robotic dog, it had to get women on board. “The Aibo ‘Latte’ and ‘Macaron’ were mainly designed by a female designer,” says Aki Shimazu of Sony in Tokyo. Mattel now employs a mix of male and female designers and sponsors research groups such as Cassell’s. This, the company believes, is the best way to ensure they are bringing gender-free (and thus highly profitable) technology toys to market.

This is not to say that all male designers are gun-toting speed freaks. In fact, men have been responsible for what are probably the two most female-friendly (and hugely successful) computer games in history. It’s 17 years since the Russian mathematician Alexey Pajitnov designed the game Tetris. Men and women played it in almost equal numbers. At a time when 99 per cent of game buyers were men, this was astonishing, and sparked the first effort to make games that interested women (see “Go girl”).

This change in mindset eventually spawned a project called “Dollhouse”. Players create people and houses, look after their needs and desires, and control their social interactions. The game proved a hit with men and women alike and has become the biggest-selling PC game in history.

You may already have guessed that the original name didn’t stick, and Dollhouse was renamed “The Sims”. The man behind it is Will Wright of California-based Maxis Software, and his approach is doing more than just producing huge-selling “inclusive” games. It’s changing the face of the industry. Because they like the games, female designers are drawn towards Maxis, and 40 per cent of the company’s designers are women.

As these women—and others joining the industry—gain influence, gadgets should lose their macho, “toys for the boys” image, and attract yet more women. Mack thinks the new girl-tech bandwagon is unstoppable. “I can’t wait to see what women are going to do with technology in future,” he says. But will there be any hint of a different gender discrimination? Will top engineers increasingly have to be women? Mack pushes his glasses back on his nose and shrugs. “You won’t have to be,” he says, “but it’ll help.”

Talk ’em down, don’t shoot-’em-up

Elaine Raybourn’s bravery deserves a medal. Raybourn works on war simulations for the US military at Sandia National Laboratories in Albuquerque, New Mexico, but two years ago, she decided they needed something a little different. “It’s easy for technology to move in the direction of more and better weaponry,” she says, “but my goal is to have technology move towards understanding and peace.”

Bringing ideas like that up with military people was challenging, admits Raybourn, but it worked. She got the go-ahead to develop games that allow soldiers to explore the kinds of decisions they might face in peacekeeping situations.

In the simulation, soldiers watch different scenarios play out while the stress, alertness and exhaustion levels of the peacekeepers involved are altered. Raybourn hopes her new approach will leave soldiers better equipped for the kind of dilemmas they face in modern missions, where they are as likely to be reasoning with an enraged crowd as rooting out guerrillas. If soldiers have advance experience of how their levels of stress and exhaustion can affect their own perception, she thinks they will temper their decision-making in a sensible way.

Raybourn and colleagues at Sandia programmed the game by interviewing retired police and military personnel about their experiences and she is now developing a version where the various possible states of mind are set by cadets’ responses to a questionnaire about themselves. The project has been so successful that DARPA, the Pentagon’s research agency, is now planning to include some of the principles from Raybourn’s work in other military simulations.

Go girl

In the 1980s, entertainment technology, such as games and toys, were made by men for men. The first generation of game programmers were nerdy boys, a group that was also into science fiction or fantasy and pseudo-military zap the enemy arcade games. And so they designed the computer games they wanted to play. Right from the start their games narrowed the market: most women (and, of course, many men) were simply not interested in this subculture. In the early 1990s, market research by Sega, Nintendo and others showed that more than 99 per cent of computer-game buyers were still men.

The early 1990s saw a plethora of companies set up by women targeting the untapped female market in the US. This movement—dubbed “Girl Games”—produced games with female characters and scenarios: high-school girls applying make-up and changing their clothes in order to get the boys to notice them, or girls going missing and needing to be tracked down. But the first girl software to hit the big time was Mattel’s “Barbie Fashion Designer”. Launched in 1995, it sold over a million copies, rivalling the fantasy game Doom. Using the software, players design outfits for Barbie and print them out on fabric-covered paper that fits in an inkjet printer. The game riles feminists and hard-core gamers alike but, when left alone with Barbie Fashion Designer, both men and women enjoy playing it.