Alleged to result in the homogenization of innovation, the use of predictive analytics in the creative sphere can evoke negative connotations. After all, isn’t the principle of basing the future on lessons from the past the antithesis of creativity? A collaborative experiment in the performing arts and entertainment industry overturns this idea by showing how artificial intelligence can generate a completely new experience and aesthetic.

The theatre. It’s one of the few remaining places where visitors are politely yet firmly asked not to document and distribute their experiences. While the reasoning involves the protection of copy rights and intellectual properties, the result is a sense of surprise and serendipity for those attending a show, which is rare in the age of social media. Without online spoil alerts, the offline experience keeps its you-should-have-been-there value.

Now, if being surprised is what you are looking for, visiting a dance performance that ‘repurposes’ movements from the choreographer’s existing portfolio might not sound immediately attractive. Yet, Living Archive – a show based on Wayne McGregor’s extensive back catalogue – evoked anything but a sense of recognition. The British director used an artificially intelligent choreographic tool to guide his dancers. Developed in collaboration with Google’s Arts and Culture Lab, the tool is trained on hundreds of hours of video material of an array of bodies dancing McGregor’s choreography. Based on this data input and the dancer’s pose, the system generates and shows suggestions for next gestures. In other words, it predicts the most likely sequence of movements to follow.

The outcome, however, isn’t predictive. ‘The poses offered by Living Archive were a starting point from which Wayne and the company of dancers could make an infinite number of responses, opening up and not restricting the serendipitous nature of the choreographic process,’ explains Damien Henry, programme manager at Google’s Arts and Culture Lab. ‘We collaborated with Wayne around his specific vision to have a different relationship with his archive, one that could feed into the creation of new work. AI became the tool to realize that vision.’

Visual artist Ben Cullen Williams, who developed a suspended video installation for the show that premiered at the LA Music Center last July, believes in the innovative qualities of predictive analytics too. ‘AI can create new dialogues, between human and machine rather than person to person. Usually the computer follows our instructions. AI thinks for itself, responding to our input. And, AI can transform the original material it receives, creating surprises for the human side of the dialogue,’ he says.

If we edit out what we think is strange, we censor the AI, privileging our lens of human values

This, he believes, can even have an effect on our perception of aesthetics. ‘As humans we are socially conditioned to recognize a shared understanding and notion of beauty. The movements generated by the Living Archive tool – which also informed my graphics – were at times incredibly sporadic without observable patterns, their fluidity seeming unnatural, haphazard, awkward, even ugly. We naturally tend to omit elements that aren’t consistent with the conditioned framework of beauty. To counter this, my video contains visual elements that I would usually have excluded. They were left in to create a deeper level of authenticity, an acknowledgement of AI’s working methods. If we edit out what we think is strange, we censor the AI, privileging our lens of human values.’ It’s a sign of the times, thinks Henry: ‘There are many artists exploring algorithmically generated art and as this area grows, so too will our perceptions of what is good, beautiful, aesthetically pleasing.’

Besides introducing a new aesthetic language, Williams hopes that the Living Archive project will make more people question the role of technology in society. ‘I believe it is only through interrogating new technology that we can find the right place and the right relationship we have with it in society. In times that technology is increasingly hidden, my video installation was unapologetic in its use of technology. A large visceral semi-transparent LED screen suspended in the middle of the stage. If technology is used as a creation tool and subject matter, perhaps we will see more visible technology on stage. I think AI could be used to have a more symbiotic relationship between performance and environment. The environment being the machine that expresses itself through screens, lights, surfaces and the like, and performers that respond to that environment in real time.’ 

The use of predictive analytics doesn’t have to result in predictive work. In fact, algorithmically generated outcomes can present an aesthetic that is unexpected and off the beaten track. But, if you employ artificial intelligence in your creative pursuits, it’s important to select from – not censor – its suggestions. 

This piece will be featured in our forthcoming November-December issue, Frame 131.