FORMA

Breaking free from static, exploitative AI models, FORMA evolves through collective human interaction, growing uniquely with each participant. Each interaction isn’t just interaction — it’s innovation.

Overview

FORMA explores what happens when AI evolves and learns like a human. Unlike traditional systems trained on static data, FORMA’s AI adapts & learns in near real-time through human interaction, becoming an open, dynamic entity shaped by its users. It serves as both mirror and lens—reflecting societal attitudes toward AI while questioning its role: mere tool or something more? By highlighting the co-evolution of humans and technology, FORMA prompts reflection on our responsibility in guiding AI’s growth and the ethical implications of this shared journey.

Type:
Academic Research Project
Institution:
TU & UdK Berlin @ New Practice
Team:
Tim Henrik Seib, Dominik Gangl
Publication
17. Feb, 2025
Research Focus

FORMA explores the question: "How can an AI system be designed to evolve through real-time human interactions, with user contributions shaping its learning, ethics, and perception?" By prioritizing continuous, human-driven input over static datasets, FORMA aims to create AI that adapts ethically and responsively, reflecting the values and perspectives of its users.

Previous Iterations

FORMA’s pervious iterations centered on real-time learning driven by user interactions. Users directly influenced the AI’s growth, with contributions visualized as dynamic data clouds representing continuous development. Early trials revealed challenges in scaling and sustaining engagement in communal spaces.

Outcome

The second iteration of FORMA responded to challenges from the first, where a single, collective AI model faced scalability issues and limited user engagement. To address this, personalized AI companions—digital twins—were introduced. These digital twins evolved uniquely with each user, fostering deeper, individualized interactions while remaining connected to the broader AI ecosystem. This shift enhanced user involvement, ensured scalable growth, and enriched data diversity, paving the way for FORMA’s fully adaptive final form.

Final Thoughts

FORMA represents just one approach to exploring real-time, interaction-driven AI development. Given more time, we would have delved deeper into the concept of AI companions, experimenting with various types of digital twins, building them out further, and testing their potential in diverse contexts. The intersection of design, computation, and art has proven to be particularly fascinating, revealing new possibilities for human-AI relationships. We see great potential in the concept of digital twins, not only as personalized AI companions but also as tools for understanding how technology can mirror and reshape human experiences. However, as this field advances, it is crucial to continuously reflect on the ethical implications of such developments—ensuring transparency, privacy, and responsible growth in human-AI collaborations.

Let’s get in touch

Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.