It has happened to all of us, says Professor Rosalind Picard. “You’ve reached a company’s voice mail system; you’ve gone through 19 layers of directions, and then for some reason the system kicks you back all the way to the beginning, and you just want to scream.

“There’s no reason for a company to treat their customers that way, ignoring their frustration,” Picard says. “To have a decent experience, technology must treat people with respect.”

What if a computer could read your heart? What if it could detect your annoyance, apologize to you, and change its behavior?

Picard, a pioneer in the field of affective computing, says it’s only a matter of time before the computer gets the skills of emotional intelligence. She is now teaching machines to sense and respond more intelligently to people’s emotions, and she is also teaching them to behave in ways that make more expressive communication possible.

Picard is founder and director of MIT’s Affective Computing Research Group and is co-director of the Things That Think Consortium, the largest industrial sponsorship organization at the Media Lab. She is the author of more than 100 scientific articles, and her award-winning book, Affective Computing, lays the groundwork for giving machines emotional skills and addresses the moral and social questions that arise from this technology.

Now, she says, if a computer capitalizes letters we want to be lower case, or if we can’t figure out how to install a new piece of hardware, we get annoyed, and if the problem becomes really aggravating, we may even curse or pound the mouse. But one day, she says, a computer will recognize our facial expressions like a furrowed brow, clenched teeth, or a smile, so the computer can detect if we are upset, angry, frustrated, or pleased.

“For example,” she says, “the Microsoft Office Assistant would know we are annoyed, and it would say, ‘Sorry, it looks like I’ve annoyed you. Click here and I’ll go away.’ It could apologize, especially if you look really annoyed,” she says. “And it may even show respect for you, by simply shrinking off.”


Picard says an MIT professor was once quoted in an article saying that he was on his knees, crying, one day because he and all his best workers from the MIT Lab for Computer Science couldn’t crack a computer problem. “If the computer can do this to PhDs in computer science from MIT, what’s it doing to the rest of the world?” says Picard, adding that we have designed computers for technical people and haven’t thought about the customer as a human being.

“When you deal with people from different countries, you show respect for them by translating the conversation into their language,” she says. “You adapt to them; you don’t require them to adapt to you. But computers are very disrespectful. They expect us to adapt to them, and if not, we are made to feel dumb. It is not people who are stupid, it’s the computer that is stupid, and it is the software that refuses to adapt.”

There are great applications to Picard’s work. Her group is now working on a project to assist autistics, and one goal is to build better learning environments for learners of all ages. An affectively intelligent system could help a child cope with the difficulty in learning math. The system could detect when the child is frustrated, bored, or interested and could help them to better handle their feelings.

One affective system might help a smoker who wants to quit spot the stresses that cause him to smoke again. Or, another system might help artists develop new ways to use emotional information in musical expression. Recently, Picard worked with car companies to adjust the steering mechanism of a car to detect the stress in your hands, and to register muscle tension and blood pressure increases. If the system detects that you are angry in traffic, it might respond by routing your cell phone calls to voice mail until you are feeling calmer.


“Our work isn’t about making machines emotional, it’s about making them more intelligent, more interesting to interact with, and better able to serve us,” she says.

Picard will not say for sure when the computer will be able to read our expressions and respond, but she says, “The more people who work on this, the faster that day will come.”