Eric Paulos is a man riding the experimental edge between humans and machines. His research in this area, both in his graduate studies in computer science and robotics at the University of California at Berkeley, and with renegade robot troops such as Survival Research Laboratories, is far more adventurous than most researchers in similar space dare to be. “Lethal, anonymous tele-obliteration,” the “I-Bomb,” and several types of tele-embodiment are just a few of his past projects. Danger is definitely not outside the scope of his work.
Soon to finish his Ph.D. in electrical engineering and computer science, look for him to continue to carry the torch that lights the dark area between humans and machines.
Roy Christopher: How long have you been interested in and involved with human/machine interfaces and robotics?
Eric Paulos: I have to admit that my initial fascination was with the machine — the motion, the intricacy, the precision, the simplicity. I’m one of those hands-on builder types always taking things apart to learn how they tick. But you get bored and restless after so long, at which point building your own creations is the only means to satisfy the madness. As for robotics I can clearly remember catching a documentary on PBS before I was even a teenager that mesmerized me. It showed people working on building running and hopping robots. Before witnessing that video it was inconceivable to me that there were actually people that society accepted (and even paid) for such activities. I sort of set it aside for many years but remember it strongly when I was choosing a graduate program. It was all too clear then what passions I should follow. It was also a short time after making that decision that I figured out the work I had seen was done by Mark Raibert and his team at CMU and the MIT Leg Lab. Maybe I’ll actually cross paths with those hopping machines someday.
A combination of fortuitous events occurred almost immediately after I started graduate school. The web suddenly provided a revolutionary interface to people and data. It was obvious to me that this was the perfect tool to begin exploration of the human/machine interface. Not robot interfaces in the traditional sense (i.e., expensive, one-off systems requiring a highly trained operator), but interfaces that were inexpensive and globally accessible to anyone — tele-presence for the masses. This motivated much of the early work with Mechanical Gaze, Legal Tender, and the Space Browsing blimps. On a parallel track, it was at this time that I begin serious work with Mark Pauline of Survival Research Laboratories (SRL). At SRL I had access to perhaps the world’s most diverse, bizarre, unique, and of course lethal machinery — a perfect arena to explore novel interface designs. Mark and I blended perfectly in our dark and wondrous thinking on so many of those early projects. Most notable during that time were the experiments in lethal, anonymous tele-obliteration.
RC: When did you get involved with SRL?
EP: My first involvement there was back in 1993. SRL is much more than it appears to an outsider. Certainly its grand vision and many of its brilliant ideas come from Mark Pauline, founder and director of SRL since 1978. But one of its hidden treasures are the many ragtag members of SRL. This group or family of people holds such a wealth of creative ideas and expert knowledge covering an overwhelmingly immense set of domains. Getting to know each of them over the years as colleges, collaborators, and friends is something for which I am extremely thankful. SRL is one of the rare unique experiences in life. One that I know I will always cherish.
RC: What are your aims with Experimental Interaction Unit (EIU)?
EP: The EIU was founded to directly confront and explore the boundary between humans and machines. My interest in the machine has not diminished, but I have always found the narrow, almost nebulous, region of the human/machine interface to be most interesting. This is where the action and interaction takes place: the locus of communication and understanding, where the human meets the machine, the blurred boundary, the unknown, and epicenter of the unexpected.
There are endless permutations to explore, and recent experiments have examined only a small subset of these themes. For example, the work on tele-embodiment: human communication and interaction at a distance through various specialized mechanical tele-operated systems . Other projects such as the I-Bomb and Dispersion, research human action and reaction to accessible interfaces of more devious technological systems in our immediate future. Look for new projects along these themes. However, I certainly don’t want to be complacent and predictable, so expect deviations in the publicly stated strategies to keep people (and adversaries) guessing.
RC: What are your major educational/career goals for the short and long terms?
EP: Right now my focus is on graduation. I’m nearing the end of the computer science Ph.D. program here at UC Berkeley. I recently took my final oral exams, and I’m now deep in performing final experiments and data collection for my thesis. I’m exploring many options: academia, research labs, companies, startups, etc. I really won’t know which path I’ll be taking until I spend some time visiting people, examining facilities, and understanding the goals of each individual choice. My long-term goal is clear: I want to insure that I continue my creative endeavors which help me maintain my sanity. At the moment, the area of personal tele-presence holds my passion and is ripe for investigation. But I believe it’s still difficult to say where my major contribution will be.
RC: Human/machine interfacing is one of the major growth points of computer science as well as psychology, ergonomics, etc. Where do you see your work fitting into the future of these fields?
EP: These days I spend more time reading sociology and psychology research and literature than almost anything else. For years I’ve been examining the machine and now I need to catch up on understanding the human element. I’m fascinated with all of the human interaction experiments that sociologists and psychologists have been exploring. During regular meetings I’ve been having with them, we’ve found tremendous common ground and certainly equal enthusiasm for researching and understanding the personal and social aspects of computer-mediated communication and interaction.
The larger, looming long-term research issues center around online trust and persuasion. Experiments have shown that current online communication channels are inadequate at supporting means of capturing and exchanging trust. Similarly, successful persuasion online is almost impossible compared to direct face-to-face interaction. Something vitally important is missing. Computers and technology will make significant incremental advances in the years to come, but filling in the human communication and interaction puzzle is what I consider one of the major contributions.
I see my work in two parts. First, collaboration with social scientists and researchers to better understand the nuances and subtle cues necessary for rich human communication and interaction. Secondly, the design and implementation of software and hardware systems to facilitate some of the more important communication cues. Of course, there will be a tremendous amount of iteration through this process to improve these systems as well. This is an enormous project and no one person is going to get there alone so we’re going to need new collaborations and interdisciplinary teams to make real progress. Do you trust me?