Primitives (2017) 6 min. 3 Channel HD computer-generated video projections. Generative choral soundtrack by Daniel Romero.

Commissioned by ENCAC and in collaboration with dancer Anya Kravchenko and the French National Drama Centre in Montpellier, Primitives uses customised motion capture data to control the movements of a digital crowd. The resulting work is a series of experimental 3D renders that explore the possibilities and limits of crowd simulation software.

Crowd simulation is increasingly used to create sublime and spectacular computer-generated scenes of conflict, disaster and death for film and television. Expensive crowds of extras no longer need to be hired, fed and paid – now an army of digital bodies can be created at the click of a button and manipulated with a tailored suite of settings, presets and parameters. This project uses the proto-human digital crowd to question those parameters, and in doing so attempts to understand how ideas of freedom and agency are defined through software. In this 10-minute film, I attempt to humanise and (paradoxically) individualise the crowd whilst confronting the design and limits of the software.

"The funny thing is that when I was learning the software and trying to push for some sort of “emergent” or unexpected behaviours, I constantly felt directed, herded, kettled into certain workflows - just like the crowds I was trying to control. The software was shaping my behaviour as much as I was trying to shape the crowd's behaviour. I felt a strange solidarity with the ‘primitives’ and started to consider how culture itself is like software – parameterised, full of norms and conventions for behaviour that it's hard to operate outside of."

Interview by @postmatter in association with @wetransfer.

This experimental project takes place at a time when the uses of artificial intelligence, chatbots, supercomputers and deep learning algorithms are constantly called into question. These entities are not human, but they are designed by humans to give the impression of humanity. When humanity is encoded as a set of parameters, we are forced to question the representational power of AI software, and in project such as this, imagine what these primitive beings could one day become.