What is it about?

Gorgon Loop is an art installation made of five rotating smart mirrors that watch visitors and talk about them using AI. When someone enters the space, the mirrors turn inward, scan their appearance, and generate real-time commentary — like a group chat between strangers gossiping about you. Each mirror has a different personality: friendly, caustic, Shakespearean, analytically detached, or mean. The comments stack on each other, building social pressure through accumulation. If a visitor approaches a mirror directly, it briefly greets them politely — then immediately returns to judging them the moment they step away. The work exposes how AI systems reduce human presence to computable data, then generate language that feels like social evaluation but is actually statistical pattern-matching with no understanding behind it.

Featured Image

Why is it important?

As machine vision and language models are increasingly deployed in public and everyday contexts, the processes of algorithmic classification and inference are typically invisible. Gorgon Loop makes these processes perceptible as lived experience — not by explaining them, but by staging them. The work demonstrates how AI systems produce the social appearance of judgement without possessing understanding, intention, or contextual awareness: what the paper calls "false agency." By exposing intermediate visual features, extracted descriptors, and language output as a continuous public performance, the installation reveals algorithmic bias — including racial and gender bias in vision systems — not as an abstract technical problem but as something people can feel, witness, and discuss in real time.

Perspectives

Gorgon Loop started from a simple observation: the gap between how AI systems actually work and how people experience them is enormous — and that gap is where power hides. When a machine scans your face and generates a comment about you, the comment feels personal. It isn't. It's the output of training data, parameter weights, and prompt design. But that feeling of being seen, categorised, and discussed — without being able to respond — is real. That discomfort is the work. I wanted to build something that makes the mechanism of algorithmic judgement impossible to ignore, not by explaining it academically, but by putting the audience inside it.

Patrick Hartono
RMIT University

This work began with a simple question: what happens when we become the subjects of AI’s gaze? As AI increasingly observes, interprets, and evaluates people, much of this process remains hidden inside a “black box.” Through Gorgon Loop, I wanted to make this invisible process tangible and experienceable, using interactive art to encourage people to question how algorithmic judgments are formed and how easily we may accept them as objective. I hope this work invites readers to look at AI not only as a tool, but as a system that shapes how we see, interpret, and understand ourselves and others.

Shuang Li

Read the Original

This page is a summary of: Gorgon Loop: An Interactive Art Installation Revealing Algorithmic Judgment through Machine Vision and Generative Language 390, Proceedings of the ACM on Computer Graphics and Interactive Techniques, June 2026, ACM (Association for Computing Machinery),
DOI: 10.1145/3816086.
You can read the full text:

Read

Contributors

The following have contributed to this page