NeurIPS 2025, San Diego, US
Artism: AI-Driven Dual Engine System for Art Generation and Critique
Art
Paper


Artism is a dual-engine AI system, which employs multi-agent social simulations and an automated critique generator to reveal the "conceptual collage syndrome" and the shifting boundaries of originality in the post-digital era.

Multi-agent Systems (LLM), RAG-Enhanced Generation, Text-to-Image Synthesis (Flux/Wen), Web-based Simulation, AI-Driven Art Critique, Virtual Art History

IEEE ICME 2025, Nantes, France
Future  Archaeological Archive: Exploring Narrative Entropy in AI-Generated Ecological Speculations
Art
Paper
This paper examines AI-generated imagery as a speculative archival method for exploring collective memory in post-ecological futures. Through a visual journal set in the year 2125, it investigates how algorithmic systems reconstruct historical narratives when source data is incomplete. By treating AI hallucinations as a deliberate methodological strategy, the study questions the boundaries between historical authenticity and speculative fabrication in machine-mediated archives. The research employs prompt engineering with LangChain and image generation workflows in ComfyUI (Flux, IPAdapter) to frame AI as an active agent in future archaeology and ecological speculation.

LangChain(Prompt Engineering), ComfyUI(Flux, IPAdapter), Future Archaeology, Ecological Speculation, Collective Memory