TriKoNet: The Trivalence Model of Potential Co-Creativity in Socio-Technical Networks

Synthesis: Ruhland introduces TriKoNet, an Actor-Network Theory-based model that conceptualizes Creativity not as an individual attribute but as an emergent effect of socio-technical networks, modeling it as a processual triadic interplay of stabilization, destabilization, and re-stabilization among human and non-human actors. In an exploratory two-phase study, pedagogical avatars were designed within a creative network, and the analysis reconstructs how the designs emerged through problematization, interessement, enrollment, and mobilization. TriKoNet allows diagnosing how technical artifacts — including generative AI systems and AI agents, conceptualized as equal, constitutive network participants — stabilize, destabilize, or re-stabilize creative learning processes, yielding four preliminary design principles and a modular design framework.

Co-Creativity as Network Effect

The paper challenges the view of creativity as an individual cognitive trait. Drawing on Actor-Network Theory, it conceptualizes creativity as an emergent effect of socio-technical networks — a processual triadic interplay of stabilization, destabilization, and re-stabilization across human and non-human actors. The study is grounded in an exploratory two-phase investigation in which pedagogical avatars were designed within a creative network, reconstructing how avatar designs emerged through problematization, interessement, enrollment, and mobilization. The findings show the avatar triad concept was not the result of individual creative acts but was constituted through iterative translation processes among students, researchers, and technical artifacts such as collaborative design tools and generative AI systems.

The Convenience Trap and Agency

The paper engages with the "Convenience Trap," which shifts human Learner Agency and creativity toward machine-induced convenience, leading to loss of competence and homogenization. There is a lack of pedagogical and learning-theoretical research explaining how co-creative learning with AI agents can function without a loss of human agency. TriKoNet addresses this by showing that AI-supported avatars are not problematic per se as black boxes — what matters is the co-constitution of their action structure by the learners. This connects to Sociocultural Learning and Collaborative Learning perspectives on learning as mediated by tools and social interaction.

Design Implications

TriKoNet allows diagnosing the extent to which technical artifacts stabilize, destabilize, or re-stabilize creative learning processes under three conditions: shared cultural reference resources, opportunities for meta-communication, and structurally open scripts. From this the article derives four preliminary design principles and a modular design framework connecting stable structural principles (role logic, technical presence, claim to inclusion) with variable appropriation space — a starting point for translating TriKoNet into a design instrument to be tested later. Implications for practice include focusing on technological agency, promoting co-creative learning networks, orchestrating interplay between students and AI agents, and preferring modular tools over ready-made avatars.

What this means for practice

• Designers. Treat technical artifacts as constitutive network participants: what matters is not whether an AI avatar is a black box but whether learners co-constitute its action structure through problematization, interessement, enrollment, and mobilization.
• Designers. Prefer modular tools over ready-made avatars; students unanimously rejected photorealistic representation and used the modular space to author their own role concepts instead of adopting a supplied persona.
• Designers. Engineer the three enabling conditions the model identifies: shared cultural reference resources, opportunities for meta-communication, and structurally open scripts.
• Researchers. Diagnose a design by asking whether an artifact stabilizes, destabilizes, or re-stabilizes the creative process, rather than whether it is good or bad in itself.

Limitations

• The empirical base is small and exploratory: 18 recruited students formed into 9 pairs, one pair dissolved before analysis, leaving n = 8 groups.
• Recruitment ran through invitations to subject representatives and active student registration rather than randomization, and the sought balance of humanities, social-science, and natural-science students was not enforced.
• The triadic Dreamer–Critic–Realist structure is analytically constructed, and the paper states its validity for natural creativity processes still requires investigation.
• Qualitative evidence comes from single sessions of about 90 minutes plus 132 pages of interview transcripts and Miro boards, and the paper notes that the framework organizing its own analysis stayed implicit while other actants were reconstructed explicitly.

Citation

Ruhland, C. (2026). TriKoNet: The Trivalence Model of Potential Co-Creativity in Socio-Technical Networks. EdArXiv preprint.