Many sources predict that in the future, humans and autonomous machines (AIs) will collaborate in networks as complementary equals.

S4NTP has been developing models of such networks where humans and machines share agency: All nodes can make decisions, which will influence the overall evolution of a shared set of processes. Single-authorship disappears, and it becomes tricky even for participants to understand who is responsible for which aspect of the unfolding audiovisual experience.

UTopologies is such a network consisting of eight nodes, each with its own body form, microcomputer and amplification, and open to local influences such as sound and light, and the presence of humans. They can generate behavioral patterns autonomously or in combined human/machine networks by communicating their internal states to each other, thus negotiating and coordinating shared or conflicting acoustic and visual behavior. The network is open to human live coding: Patterns merge, contrast or clash with those already circulating, thus blurring the lines between ›concert‹ and ›installation.‹

UTopologies creates mappings between abstract space (i.e. algorithmic topology) and physical space(s) on multiple levels: the body shapes are materializations of generative processes (rendered in wood), 
and the physical spatial disposition is modeled as an algorithmic graph to be used in the decision-making processes within the network.

Furthermore, physical placement can be tuned to local context: in the exhibition at Kleiner Wasserspeicher, we could put five nodes inside a dark reverberant circular space, one on top of the entrance, and two in trees in the vicinity / within earshot. Such nonlinear connections between heterogenous spaces create a sense of presence and absence: we become aware that we never experience the entire topology in action; there are always unknown connected processes going on elsewhere that invite speculation. 

location: S