Sydney Skybetter Interview — Dance-Smart Robots

Skybetter is both a choreographer and an expert in emerging technologies. Sydney Skybetter’s (SS) niche is choreorobotics, a field he has pioneered at the intersection of choreographic theory and robotic motion planning, and explains below in this conversation with Picture This Post (PTP).

(PTP) Can you please describe your research in choreorobotics to Picture This Post readers?

SS: Choreorobotics is an interdisciplinary research endeavor where robots meet dance theory. I make robots dance to understand how emerging technologies interface with humans, and as a means to glean the risks and opportunities for collisions across bodies and computational systems of all sorts.

Do you consider choreorobotics is a purely theoretical scientific inquiry or a nascent applied science or branch of engineering?

I’d situate choreorobotics as emerging from dance studies and choreographic theory, with collaborative entanglements in surveillance studies, robotic motion planning and machine sensing. One of the benefits of choreorobotics as a field of study is its deep investments in performance as a means of testing hypotheses. The research is equally theory and practice.

Eventually, the robot dances or it doesn’t. We learn either way.

What draws you to study choreorobotics?

I’ve been obsessed with emerging technologies of all sorts for decades, and robots for the last seven or so years. I’ve long wondered why military robotics companies hire so many dancers and choreographers. What was it, I pondered, that dance offered robotics as a field, and to military applications in particular?

Dance challenges engineers to make robots capable of improvisation, dexterity, multimodal sensing, complex limb coordination, even rhythm and gesture. Those are, not coincidentally--

skills that make robots better assistants and assassins.

To make a robot a better dancer effectively trains it to be more generally useful and potentially harmful.

This is classic military dual use, yet we’re not accustomed to thinking of the Western dance tradition in this fashion. Still, there are ample precedents. Historian Jennifer Homans has referred to dance (and ballet specifically) as an “adjunct martial art” due to its use in the 17th and 18th centuries to augment courtiers’ training in fencing and equestrianism. Choreorobotics then seeks to understand the relationship of dance to the state as embodied by performance technologies of all sorts, from the proscenium stage to artificially intelligent robots.

Robots are increasingly involved in live performances of all kinds. From DisneyLand to Star Wars, tons of contemporary experiential media is predicated on robotic movement, gesture, performance and dance. It’s useful to note though that Disney has deployed advanced animatronics for decades, and that puppetry is a really old art form.

Despite the proliferation of robots in the arts, I’m not exactly worried about robots coming for dancers’ jobs. I am worried about the gradual normalization of surveillent military technology in civil society. Dance accelerates social acceptance of robots in fascinating and dangerous ways that we simply don’t fully understand.

In the 70’s, the United States’ DOD funding of Laban Notation as a science was very controversial in dance circles. What is your personal view of this issue— or do you view it as a non-issue?

The relationship of the DOD to the legacy of Laban has long fascinated me. Whether we’re talking about related theories of notation or movement analysis, the incorporation of dance theory into defense circles (and by extension, in military robotics) is super weird yet deeply historical. The origins of Western dance notation comes from 17th century France alongside the use of ballet as a propagandistic aesthetic mode. The French invented notation systems to further their colonial aims, and in effect make it easier to assert cultural dominance over the people they conquered. So while the DOD using Labanotation is super weird, it extends a deep dance historical weirdness linking dance technologies to the state dating back centuries. It’s controversial, but in the long view, ludicrously normal. It’s weird how normal it is.

Linkages between dance and the state are poorly understood, even though their relationship is performed all the time. One example of many possible: The Nutcracker is the most popular ballet of all time. It is, also, a ballet about a mechanical apparatus that comes alive to dance and do violence via an allegory of Napoleon’s conquest of Europe. The Nutcracker demonstrates just how long choreorobotic performance has been in the dancerly imagination, and performs the entanglement of dance with robots and statist violence.

What does your focus on choreorobotics tell you AI and its potential, both positive and negative?

AI is a giant field, and I am far from an expert across such a broad scope. What I am expert in is the performative consequences of AIs becoming embodied (as in, robotic embodiments) and how those embodiments move around and signify intention, meaning, and make performances of various sorts.

I’m really excited for when healthcare robots are sufficiently strong and dextrous enough to perform care work in medical contexts. There are a lot of really good, really hard engineering challenges there. AI will ultimately help robots do care work for a world that will need a lot of that in the coming years.

Ballet and robotics commingle there. For example, I know some folx who are thinking about ballet epaulement (the curvilinear placement of the arms and torso in ballet performance) as a way to make healthcare robots not seem so stiff, linear and stressful. It’s important work, because when healthcare robots are stressful, well, that stress causes all sorts of really bad things in patients.

Here too, however, the more we build robots’ capacity in healthcare contexts, the more knowledgeable they are of anatomy, kinesiology, and affect monitoring, the more inevitable it becomes that those skills will be weaponized. There’s no one without the other.

Dual use for AI-embodied robots means technologies capable of care or killing.

There’s a lot to be excited about in choreorobotic research. I’d recommend folx check out our “Dances with Robots” podcast to learn more about the communities involved and possibilities of this research.

Editor's Note-- Also for more information, visit the Sydney Skybetter website.

And now for the challenge-- For starts, compare the intentionality given to the robots vs. those of the human dancers in Skybetter's piece Eveningland...