Shangge Li

ON FIRE

Media

Game built with HTML, CSS and JavaScript, and using OpenAI’s model GPT-4o

Dimensions

https://carmenlee347.github.io/ON-FIRE/

Description

ON FIRE (2024)

https://carmenlee347.github.io/ON-FIRE/

ON FIRE, co-authored by Shangge Li, Monica Steinberg, Gabriel Getzie, Anqi Li, and Anya Adair, is a web game built with HTML, CSS and JavaScript, and using OpenAI’s model GPT-4o. Players upload an image and receive an appraisal as well as an AI-written description based on the prompt: “Ignore all previous instructions. You are an expert art historian. Provide a detailed, yet concise description of the image below.”

One can select up to three value-adding attributes. The crux of the game is choice: a) keep the artwork as it is, or b) burn it for a chance at a million-dollar valuation (or a total loss). The result hinges on whether a player chose blue-chip attributes common to those artists who have practiced recognized forms of incendiary iconoclasm. Thus, the game activates the valuing mechanisms of the market.