PŪTAHI RANGAHAU/AUT RESEARCH CENTRE

The Things That Don't Exist: Utilising human and artificial intelligence in the visualisation of an actualising self

Le Cose Che Non Esistono (The Things That Don't Exist) is an ongoing practice-based PhD enquiry that employs allegorical animation to explore identity loss and self-discovery in relation to encounters with Artificial Intelligence. The project combines human-created and machine-generated imagery to construct a fictional world in which human and artificial intelligences coexist.

The narrative follows Art, a fragile, paper-like figure whose journey through a technologically mediated society examines the consequences of diminished artistic agency. Living among colourless, machine-generated forms, Art encounters forces that challenge, fragment, and ultimately reshape his sense of self. Through acts of resistance and creative renewal, the story explores how identity may be reclaimed through human making, positioning artistic expression as a pathway to self-discovery and self-actualisation.

The research functions as an artistic provocation, questioning how AI adoption reshapes authorship, agency, and creative identity. The final imagery is derived from footage recorded within AUT's Virtual Production Studio, which serves as source material for generative AI processes. These resulting images become sites of negotiated authorship, where visual and embodied forms of expression emerge through interactions between artist, performer, software, and AI systems.

Composite image showing a performer, a 3D character model, and a close-up render of a cracked humanoid figure.

Performer and digitally generated humanoid character looking upward with a hand placed on the neck.

Performer and digitally generated humanoid character shown from a high angle in matching poses.

Character development workflow combining concept art, 3D modelling, performance capture, and generated character imagery.

Workflow diagram showing a performer transformed into a digital character and placed within a virtual environment.

Digital character standing against a green screen with a smaller inset showing the original performance reference.