Swarming Media is a generative codework that has been produced by simulating a fleet of airborne utility drones, navigating a generic urban environment, and performing myriad “dull, dirty, and dangerous” tasks. In the process, they transmit a continuous stream of terse bulletins conveying their every action, which are collated into a single console readout.

The wall of text yielded by this process—divided into rectilinear patterns modelled broadly on an urban grid—suggests a grey crucible of unrelenting activity. Nevertheless, there are also distinct patterns of behaviour that can be discerned when following the paths of individual machines. Most are engaged in a variety of mundane, frequently repetitive tasks, while also requiring occasional maintenance and recharging. Some are lost to accidents or vandalism, and necessitate disposal, while others are attacked and become infected with varied malware packages, and thus demand swift destruction before they corrupt the remaining drone population. Comments and observations by their human monitoring teams are also appended into the message bulletins, whether to note down especially difficult tasks, highlight potential threats, or simply indulge in weary humour. If this piece was read with the eyes of a machine, then the underlying map of the urban environment could be discerned over time, based on tracing the routes taken by individual drones. Occasional sensor test readouts, however, do provide direct glimpses of a reporting drone’s immediate surrounds, at given time steps.

Swarming Media is a more sophisticated reimagining of an earlier piece by the author, entitled Swarm, which was compiled and printed in January 2020. The look of this piece was inspired by another codework Bad Machine (1998) by Dan Shiovitz.

Example plates from Swarming Media.