Expressive Machines

“Robots exist in the world, and machines that do things exist in the world, and people are always going to make art with the latest technologies,” says the associate professor of computer music and digital arts in the Technology in Music and Related Arts (TIMARA) department. “That’s always been the case, whether it’s the printing press or electricity. Artists have always been on the front lines of using things not as they were originally designed and intended, but for creative purposes.”

His research spans beyond the theoretical: Kemper actually designs and creates the musical robots that he studies. He is involved in every step of the process—whether it’s brainstorming ideas, researching technology, writing software, or developing hardware and fabricating parts. 

“What is interesting about this?” he always asks himself when creating a new machine. “What can I do that's going to be novel?”

That question is poignant, especially with the growing fears that AI and robots might eliminate certain professions and diminish our ability and need to be creative and produce our own original work. Kemper acknowledges these fears but explains that his work doesn’t necessarily use AI.

On the contrary, Kemper builds machines that humans can use to realize their creative vision. As he outlines in an article in the journal Frontiers in Robotics and AI, there are many different types of musical robots. For example, one of his creations is a performance system that involves vibrating motors. He held the motors, controlling the speed of their vibrations by turning his wrist a certain way, and then used them to play a harp. The machine didn’t remove humans from the process of making music; instead, the machine changed and expanded the sounds we could produce. 

Kemper has also explored the concept of machines making “mistakes” to illustrate robots being musically expressive. He explains that when we watch a human performer, we feel a sense of empathy for (and an inherent connection to) them because they are another human. This is the impact of hearing music, especially when we see it performed live. 

This concept becomes more complicated when it comes to robots, as empathy is a human quality. “How do we empathize with robots?” Kemper wondered. “How do we anthropomorphize these machines?” 

His solution was to program his musical robots to “get stuck” in a loop. The robot will suddenly start repeating the same action over and over. Though this is deliberate on the part of the creator, it gives the appearance of a glitch.

“It’s like a mode of failure,” he explains. “As an audience member, you’re thinking, ‘Oh, is it stuck? Is it working?’ There’s this moment of empathy for this machine, like, ‘Oh, I want it to work.’ But it’s also a critique on mechanization in general—how we view machines and technologies as things that are perfect and always working.”

This critique and alteration of people’s perspectives on robots and machines is an integral part of Kemper’s research and mission. Given the multifaceted nature of his work, it also has many layers of impact. On one hand are the boundaries he pushes with the music he makes through his robots. Because they are machines, they can be sonically distinct from humans. On the philosophical side, Kemper wants people to question their preconceived notions about creativity and machines. 

“As humans, we invented the concept of creativity,” Kemper stated. “You can’t have creativity without human involvement. So one way to think about it is that the design of robots and machines in general is an outgrowth of human creativity. If those machines go on to make art or make music, ultimately there’s still humans in the chain who have contributed creativity to that.”

Kemper’s work lends itself well to student collaboration, and he’s taught relevant classes such as Workshop in Music and Media Technologies. But he views ideas and inspiration as a constant flow—which is why he introduces students to his work and assists them with their own research.

“The more students can be involved with my research and see what I’m doing,” he says, “the more that I can also help them with the projects that they want to pursue.”

Steven Kemper’s research revolves around musical robotics and the impact they can have; this research involves the creation of his own machines. He earned his doctorate in composition and computer technologies at the University of Virginia. Prior to coming to Oberlin, he was an associate professor of music technology and composition and the director of the music department at the Mason Gross School of the Arts at Rutgers University.

About the Illustration

Illustrator: Israel Vargas

Return to Oberlin Research Review