CMU School of Drama


Saturday, February 14, 2015

Heather Knight’s comedic robots to perform at San Jacinto College

Your Houston News: Living: A robot comedian, robot film festival, and Cyborg Cabaret are all part of Marilyn Monrobot Labs, founded by Carnegie Mellon doctoral candidate and social roboticist, Heather Knight. She will bring her robotic comedy show to the San Jacinto College South Campus on Tuesday, March 3.

1 comment:

Unknown said...

I'm really excited by the robot-human interaction work happening here at CMU (minus the potential robots-taking-over-the-world scare aspect to some of these technologies). There's so much potential for robots to help humans in social services, but that can only happen with the work of people like Heather Knight who are improving robots' ability to communicate/interact with people. It seems natural that these researchers would gravitate towards robots in the performing arts to safely test robots' abilities to both recognize and express human emotions.

That said, the success robots can have as performing artists is directly proportional to their ability to express emotions in a way humans can recognize. For this reason, I'm really curious about how Data is designed. Is it made to look like a human? To what extent can it move appendages? Does it have facial expressions? And based on all of that information, does Data use the expressive tools it has to their best capacity, or is it programmed to make human gestures that don't make sense on its robotic body? This was one of the biggest problems with a similar Robotics/Drama project last year; Sam French presented *Sure Thing* by David Ives with a robot playing Bill and a live actress playing Betty. The Bill robot was only effective when it used gestures in its own movement vocabulary, as opposed to the times that it tried to mimic humans with it's very limited range of motion. I fear the same kinds of problems could happen with Data.

Furthermore, what does this mean about Data "reading" an audience? This robot can think for itself? (If so, I'm not sure whether to be impressed or creeped out.) And even if it can adjust its routine for its audience, can it improvise based on what's happening in the room? Can you really get a live comedian experience from a pre-programmed performer?

All that to say, it's exciting research but I'm highly skeptical of its performance value.