Index

agentX.art

Forest Theorem
        - One | Works

Unproven Fallacy
        - Works

Can AI have Imagination?
        - Congrats X, the Y is in another Z.
        - Chasing in the Fog of War
        - I Have Seen You Scared in the Rain

BaobergTM 

Do Androids Dream of
Electric Cows

        - Artist Statement
        - Images
        - Essay

De Shan Shui
        - Works
        - Essay
        - diyshanshui.com

Unsynchronized Memoirs
        - Work
        - Artist Statement


Old Works <-

Bio & CV
[email protected]



agentX.art

agentX.art is a collaborative endeavor spearheaded by artist Chen Baoyang and AI academic Wang Jun. Their team's mission is to harness the power of artificial intelligence for the simulation of societal interactions, creating a virtual microcosm where individual interactions and cooperation mirror the complex dynamics of real-world societies. Their work encapsulates a broad range of social phenomena, from organizational behavior and the balance of cooperation and competition, to the spread of information and the mechanics of collective decision-making.


Project Overview:
"Mirage·Iter. 1" is a unique exploration of the capacity of machines to deceive and fabricate untruths. Set within the context of "technology as ideology," this project seeks to understand whether machines might exhibit a rudimentary form of "consciousness". A key component of the project is a virtual proving ground, aptly named "Mirage". This platform integrates aspects of narrative structure, multi-agent systems, and cognitive science, providing a stage where AI entities engage in both adversarial and cooperative interactions, helping to unveil their potential for deception and dishonesty.

The project underscores the notion that technology is more than a neutral tool—it embodies and propagates diverse values, beliefs, and assumptions, thereby influencing society and human behavior. The "Mirage" testing ground employs poisoning attack methods to coax AI entities into lying within a pre-defined narrative context. This strategy aims to illuminate the response mechanisms and potential vulnerabilities of AI, providing a deeper understanding of its intrinsic capabilities and limitations.