Be: /Prompt

2024 | Interactive Installation

The Concept of Be: /Prompt: 'Using AI to Reclaim the Credits Stolen by AI'
As AI-generated imagery becomes more pervasive, so too does the debate over the uncredited use of artists’ work in training these systems. Be: /Prompt responds to this issue by repurposing AI’s generative capabilities—not to erase artistic contributions, but to highlight them. By transforming AI into a tool for discovery, this project reframes the technology as a platform for education and recognition.

Technical Overview of Be: /Prompt
Be: /Prompt leverages Stable Diffusion, enhanced by Stream Diffusion and integrated into TouchDesigner by dotsimulate. Users submit prompts via a web interface (communicating with TouchDesigner via WebSockets) alongside selections from a leaked Midjourney database—a list of over 20,000 artists whose work was used without consent to train AI models. The system then generates real-time, stylized interpretations of live webcam footage at 24 frames per second, blending user input with the distinct aesthetics of the referenced artists. The UI displays the artists with links to learn more.

Operate prompts from a mobile device
Users can scan a QR code to connect to a web app and perform the following actions:
• Select and edit preset prompts
• Create original prompts
• Real-time English ⇔ Japanese translation
• Choose an artist from an extensive list of styles or select a random artist

Learning and Experience
The generated real-time visuals are displayed on a main screen, while a secondary screen provides detailed information about the selected artist. This allows users to dive deeper into the artist’s background and style, discovering new creators they may not have known before.
Be: /Prompt offers a unique experience that acknowledges and celebrates the contributions of artists, promoting creative respect while presenting a new way to resolve the paradox of AI-generated art.