Machines With Charisma

Can robots tell jokes? If they want to be accepted by humans, they can--and should, says grad student Heather Knight

By Mary Lynn Mack

Sure, a person can be charismatic. But a robot?

Heather Knight thinks so. Currently a doctoral student in the Robotics Institute, Knight says that if we want human-robot interaction to be as seamless as human-human interaction, then we'd better make sure robots are more charismatic.

In fact, Knight says, the "icing on the cake" could be giving a robot a sense of humor. Adding humor and a sense of fun are important to creating the connection between human and machine, she says. "I think in the same way that we enjoy spending time with charismatic people, there will be different types of applications that will become open if we can make socially intelligent machines," Knight says. "And it could do much to help us forgive them for their inevitable missteps--particularly if the humor's self-deprecating."

Knight's interest in human-robot interaction was sparked when she was an undergraduate at the Massachusetts Institute of Technology and had an opportunity to work on a Cyberflora, an exhibit consisting of 20 individual robotic flowers that could sense and respond to the movements of nearby museum visitors and an instrumental soundtrack. The robotic flowers, which glowed brightly in shades of green, purple, blue and orange, would bend and stretch toward an oncoming visitor — but if that visitor got too close or attempted to touch the flower, it would suddenly retreat, close up its floral bud and cool its colors.

Since then, Knight has continued to work on the development of several other robots, including The Huggable, a robot companion with affective, relational touch; Trisk, a robot that understands language by sensing its environment; and RoCo, a robotic computer for learning companion research. Knight's current work, for which she recently earned a National Science Foundation graduate research fellowship, is with Data, a performance robot who not only does stand-up comedy, but also is learning to act in collaboration with other performers.

Data is a Nao humanoid robot, designed by Aldebaran Robotics. At just under two feet tall, his gestures may not be grand, but through his acting lessons with Matt Gray, an assistant professor in CMU's School of Drama, Data is learning how a simple drop of the chin can convey shyness or how hand and arm movements can help accentuate a punch line. (In case you were wondering, Data is named after the Star Trek: The Next Generation character Data, who, through his desire to be more human, also explored the arts and acting.)

Gray says actors have to formalize the "process of being human" in order to be effective in their craft, adding that teaching this practice to Data has had many similarities in that he makes many of the same "elementary mistakes" as young actors do. One such example is that at times young or nervous actors will speak aloud text written into a script--such as stage directions or bracketed instructions from the writer--that was not supposed to be read.

Similarly, Data's program does not differentiate between text to be spoken and punctuation, so while working on a line from Shakespeare, he once declaimed, "Woe woe woe is me, exclamation point."

While on stage for his stand-up routine, Data uses real-time audio and visual feedback from the audience to adjust his performance. In a demonstration of the system at December's first-ever TED Women conference, on-stage sensors measured the volume and duration of the audience's ambient response in the form of laughter, applause or chatter. The initial version of the software assumed most feedback was positive, so it didn't differentiate laughing and applause from heckling and booing. Knight says the tracking capabilities will be extended in the future, to allow deeper analysis of the performer-audience relationship.

During the first demo, visual feedback was given through the use of colored cards--green for "Yes" when asked a direct question or for a general "I am enjoying the performance" response and red for "no" or "I am not enjoying the performance." At Data's first live-audience performance, some jokes did better than others, and a few did flop (with almost complete silence from the audience once he reached the punch line). But when he asked the audience about his overall performance, he received almost all green cards and his loudest applause of the performance.

Knight had considered adding "stalling capabilities" to the routine, where Data would pretends to drop something or smoke a cigarette in order to allow for more time for audience feedback and for Data to process the information. But after feedback from the crowd at TED Women, she suspects the best modification would be to have the robot more directly acknowledge the audience, maybe with another joke: "You guys didn't like that one? I'll try a Steelers joke next."

While not all robots are charismatic, Knight is--and it's hard not to catch the enthusiasm she has for her research. Among her role models, she lists Carnegie Mellon's Manuela Veloso; MIT's Cynthia Breazeal, a mentor and a pioneer in social robotics and human-robot interaction; and former boss Bruno Maissonier, the founder of Aldebaran. Knight also cites the work of Jane McGonigal as an influence. McGonigal is a game designer, games researcher and author, specializing in pervasive gaming and alternate reality games.

Last year, Knight made a splash in the mainstream media with her Robot Census. An attempt to count every robot residing in the university's labs, the form designed by Knight and her friend, Los Angeles-based graphic designer Chris Becker, was also a parody of the U.S. Census form. (Sample instructions: "Use blue or black pen. No Binary." and "Do not count any robots that were in a robot nursing home, jail, detention facility, etc., on September 8, 2010.")

In the end, 547 robots were counted on the Carnegie Mellon campus--more than double the number that RI staff had predicted. Included in her findings were that 60 to 70 percent of the robots on campus have been assigned a gender and most have been given a name.

"There were some robots where the names reflected the function of the robot and some were just (called) a number, but the vast majority of the robots were actually named," she said.

After the initial Robot Census at Carnegie Mellon complete, Knight opened the census to worldwide participation and would like to see census committees established at other large robotics campuses to see if there are conclusions that can be made about a campus from the types of robots in use there.

"As researchers, we might all call ourselves roboticists, but 'robot' is a diverse term," she says. "It will be interesting to examine how we think differently about robots, and have diverse cultures of robotics innovation across different institutions. The censuses will certainly start that conversation."

More information about Knight, her research and her robot theater company can be found on her website
For More Information: 
Jason Togyer | 412-268-8721 |