The boundaries between human and machine will become increasingly difficult to locate. Not because biological and computational systems are converging into sameness, but because intelligence itself is beginning to emerge through interaction and shared environments of perception. This shift is reshaping the pulse of human-machine collaboration. The machine no longer operates solely as an instrument of automation or distant analytical computation. The rise of physical AI, humanoid robotics, and adaptive systems introduces a more relational condition.

Sougwen Chung’s Ecologies of Becoming explores the transition through a prolific creative and speculative practice situated somewhere between drawing, robotics, sculpture and computational performance. As artist, researcher, and engineer, she dissolves the boundaries between human intent and technological intervention, constructing environments where creative production emerges through reciprocity rather than control.

The practice is beyond automation. Its central focus is entanglement, through a choreography of gestures between person and program, improvisation and computation. The machine is a participant.

Shared Signals

This practice has centred on the creation of the Drawing Operations Unit Generations (D.O.U.G.). Robotic systems trained on gestures, movements, and drawing behaviours. The systems do not merely reproduce pre-programmed outputs. Through repeated interaction, they inherit behavioural tendencies and traces of motion that evolve over extended timescales.

This is a subtle yet important distinction. The robotic gesture appears less as mimicry than as behavioural echo, an emergent continuation of movement refracted through interpretation. 

In this view, the machine moves beyond a traditional role of instrument. Here, human and machine operate together at the level of electrical signals, within a shared perceptual space.

These dynamics increasingly resonate with broader developments in physical AI collaboration and social robotics. As humanoid systems move into workplaces, homes, healthcare, and public environments, intelligence may become defined not only by capability or reasoning, but by the capacity to navigate relational space, gesture, timing, and co-presence with humans. The work functions less as speculative abstraction and more as an early perceptual framework for understanding collaborative intelligence in embodied form.

Gesture as Interface

Much of modern computing evolved through abstraction. Interfaces became increasingly detached from the body, mediated through screens, symbolic inputs, and invisible infrastructures. Yet physical AI is reversing this trajectory. Robotics, spatial sensing, wearables, and adaptive environments are returning computation to movement, proximity and physical interaction.

Ecologies of Becoming exists in this space. Movement as machine memory. Behavioural data, repetition and adaptation.

This exchange is a deeper convergence between human and machine cognition at the level of signal. Biological nervous systems and computational systems operate through electrical pulses, transmission, feedback, and response. A shared substrate. A choreography unfolding not only visually, but as an interplay of living and synthetic forms of signal processing. Technology appears as a new membrane for existence, extending perception, cognition, and expression.

Cybernetic Presence

What makes the work compelling is that it resists reducing the human body to pure optimization. Imperfection remains visible throughout the interaction. Drift, hesitation, uncertainty, and variation persist, rather than being removed for mechanical precision.

This creates an ethereal form of machine presence, neither fully autonomous nor entirely subordinate. The systems occupy an ambiguous space between tool, collaborator, mirror and participant. Less industrial workflow and more improvisational performance. A recursive negotiation between organic creativity and algorithmic process.

In many ways, the work recalls contemporary cybernetic systems, and continuous loops of sensing and adaptation. But the introduction of ambiguity, intimacy, and embodiment into the loop shifts this category toward a shared ecology.

An ecology is not a fixed structure but a network of relationships, adaptive, interdependent, and continuously evolving. The act of “becoming” similarly rejects permanence for ongoing transformation.

Together, these ideas reposition AI, away from dominant narratives of efficiency and control, toward systems of mutual influence between humans, machines, environments and signals. This perspective feels increasingly relevant, as AI systems become embedded within day-to-day life. Emerging forms of physical AI will blur distinctions between object, interface, assistant, and social presence.

Collaborative Intelligence

For much of technological history, machines amplified labour while software amplified information processing. But social robotics and embodied AI introduce another condition entirely: systems capable of participating within shared environments of movement, attention, and behaviour. This changes the cultural meaning of intelligence.

Ecologies of Becoming ultimately suggests that the future of AI may not arrive through separation from humanity, but through increasingly entangled forms of co-evolution. Not artificial systems replacing human creative production, but hybrid ecologies where cognition, perception, and expression move fluidly across biological and computational forms.

As physical AI systems continue to evolve, the challenge will be not only technical capability, but cultural adaptation: how humans learn new modes of interaction, how perception extends into responsive systems, and how cognition becomes distributed across human and machine environments.

This is more than robots.