Can disabled people make art in relationship with AI?

Abilitopia is a performance work shaped by disability, technology and a central premise that none of us are independent. Created with dancers Duncan Armstrong, Raven Afoa-Purcell and Julie van Renen, the project asks: What happens when interdependence is treated not as weakness, but as a creative force?

The work brings dancers and artificial intelligence into relationship on stage. It is not about AI replacing humans or producing a seamless technological future. Instead, it explores connection, care and shared agency when disabled ways of working lead the creative process. Abilitopia is crip-led: disability is not added later or framed as something to overcome. It is the starting point. Access shapes how the work is rehearsed, structured and performed.

The AI in Abilitopia is visible and embodied within a mobile robot that moves through the space with the dancers. The robot is AI-enabled and responds in real time, observing, describing and speaking into the room.

Yet it does not function alone. The robot is operated by its creator, Adam Ben-Dror, and its driver, Fiona Saunders. Its movement depends on Fiona’s driving, its technical framework on Adam’s design, and its speech on live prompting and context. Its presence also depends on the dancers who respond to it.

Public conversations about AI often centre on autonomy and the replacement of human labour. In Abilitopia, the opposite becomes visible. The AI’s presence relies on layers of human coordination and care.

The robot sometimes glitches and occasionally needs to be rebooted. Rather than hiding these interruptions, the performance incorporates them. Rebooting becomes part of the choreography of care, reminding us that technological systems — like bodies — require maintenance and support.

The dancers respond to the robot’s descriptions, and the operators respond to the unfolding performance. The work emerges through negotiation rather than control. For many disabled people, unpredictability is already part of daily life — in our bodies, our access and in our interactions with institutions. Abilitopia stays with this reality rather than smoothing it away.

This approach challenges dominant narratives about technology and disability. Many imagined AI futures still centre on fixing, correcting or eliminating disabled bodies — treating disability as a problem technology should solve. Abilitopia refuses that logic.

Disability is not a problem to be solved but the context in which the technology operates. The robot does not measure dancers against a norm; it learns within a disabled-led environment and becomes part of a shared creative system rather than an authority above it.

The project also rethinks audio description (AD), which traditionally explains visual action so blind or low-vision audiences can follow a performance. AD often focuses on what movement looks like from the outside.

In Abilitopia, we experimented with describing internal experience instead: what the movement feels like inside the body. Is it heavy, sharp, slow, electric? What is the dancer noticing within their body?

This approach was influenced by conversations I had in Los Angeles with Indian disability artist Akhila Vimal, a dancer with visual impairment whose research explores expanding the possibilities of audio description. She spoke about AD as a creative practice rather than simply an access service.

In response, audio description became part of the choreography. The dancers’ sensory experiences shaped the language. Sometimes the description revealed something invisible to sighted audiences; sometimes it unsettled what the eye assumed. AD was not an add-on, rather it expanded the performance itself.

Abilitopia suggests that when disabled ways of working lead, different values and aesthetics surface: relationship over control, access over perfection, care over speed. Technology and audio description become tools within this ecology of collaboration rather than systems that stand above it. In this sense, the work offers not just a performance, but a small model of how technological futures might be shaped differently.