Modal and Gray Area showcase 11 artworks built with AI

Modal and San Francisco nonprofit Gray Area have opened <em>Scaling Intelligence</em>, an exhibition of 11 new artworks made with machine intelligence and cloud scale compute. The project brings artists together with an AI infrastructure company to explore models and agents as creative tools.
- Scaling Intelligence features 11 new AI artworks.
- Modal provided engineering help and compute credits.
- Gray Area is a San Francisco nonprofit supporting art and technology.
- The works use models, sensors, agents and robots in different ways.
- Artists were selected through a global open call.
An exhibition built with engineering support
Gray Area has supported artists working at the intersection of art and technology for nearly two decades. For Scaling Intelligence, Modal worked with the selected artists over several months, providing engineering help and compute credits as they built their projects from scratch.
The artists were chosen through a global open call for practices that use AI with intention and imagination. Modal said the collaboration challenged assumptions on both sides. Artists were surprised that an AI infrastructure company cared about their creative work, while people in technology were surprised that artists wanted to work with an AI company.
The exhibition responds to a widespread view of AI as a technology for automation and producing uniform or low quality material. Modal presents the works instead as examples of models trained at scale and autonomous agents becoming part of an artist’s toolkit.
From nonexistent islands to digital ecosystems
The 11 works use AI in markedly different ways. In Searching For (Bradley Land), Cezar Mocan imagines a satellite repeatedly searching for an Arctic island that an explorer claimed to discover, but that never existed. One model generates possible satellite views while another checks them against the explorer’s description.
In What is it like to be a river?, Colin O’Donnell, Justine Sun Dela Cruz and Will Gardiner connect a model to sensors on a creek in the Sierra foothills. The work creates an ongoing version of the river using seasonal data and the movement of people in the gallery.
Other projects include Grayson Earle’s robot arm in 00100, which learned to make an offensive gesture in simulation; Iana Chugunova’s nine model choir, which learns from visitors singing with it; and Ira Greenberg’s digital ecosystem, where creatures are born, compete, breed and die while language models write field notes and obituaries.
AI as translator, collaborator and subject
Kristina Pulejkova’s The Metabolic Choir uses sensors to track live yeast and turns its readings into cat voices. Miguel Novelo’s (Φ ᆺ Φ) asya transforms visitor drawings into Maya style glyphs, adding them to a growing book. Parag Mital’s Between Machine and Machine observes groups of up to a thousand agents as they form committees, write constitutions and argue over pixels.
Rashed Haq’s Activation Studies explores unnamed concepts inside an image model without using words as prompts. In Sasha Sorokina’s IMPRINT, an entity called Immy learns from conversations with visitors and is retrained each night on what people said that day. Spencer Chang’s Free Agents gives phone based agents personal goals and lets them scroll, journal and message without stopping.
For visitors, the exhibition offers direct encounters with systems that respond, learn, simulate ecosystems and make decisions. According to Modal, the projects show how artists can use AI to investigate memory, language, ecology, politics and responsibility, rather than treating it only as a way to automate creative work.
