Research
Research on AI-Mediated Coordination, Safety, and Learning Systems
Conceptual Papers, Institutional Analysis, and Reference Implementations
This page brings together conceptual and analytical research examining how artificial intelligence reshapes coordination, discovery, safety, and learning systems. The work focuses on institutional design, transaction costs, information asymmetry, and the unintended consequences of AI-mediated decision-making across commercial and social domains.
Research orientation
The research presented here is conceptual and historical in nature. It aims to clarify mechanisms rather than evaluate outcomes, and to support analysis across domains where AI increasingly mediates access, choice, and interaction. The work draws on institutional economics, innovation management, and political economy to examine how coordination technologies shape power, incentives, and system behaviour.
Research Themes
Market Coordination and Exchange Infrastructure
Transaction costs, intermediation, platform power, and decentralized alternatives for product discovery and exchange.
AI, Safety, and Child-Centric Systems
Design considerations for AI systems operating in environments involving children, safeguarding, asymmetric vulnerability, and duty of care.
AI and Learning Environments
Institutional and pedagogical implications of AI-mediated guidance, discovery, and feedback in educational contexts.
AI Enterprise Governance and Safety, Cost and ROI
Institutional frameworks for AI adoption, risk management protocols, and the alignment of computational costs with operational ROI in regulated industries.
Working and Conceptual Papers
- The Evolution of Exchange: From Barter Frictions to Decentralized AI Infrastructures, J.D. Linton. A conceptual and historical analysis of how transaction-cost-reducing technologies repeatedly generate new forms of intermediation and market power, and how AI-mediated commerce may be reorganized through protocol-level infrastructures that decouple discovery from platform control.
Reference Implementations and Applied Context
Some of the concepts developed in these papers are expressed through private reference implementations made available for independent use. These implementations are non-instrumented and non-observational: users interact with them privately, without data collection, monitoring, or experimentation. They are not used to generate empirical findings and should be understood as conceptual implementations rather than research studies.
- Maimai Supply. Reference implementation of a decentralized approach to product discovery and exchange infrastructure.
- Studio Clarté. Reference implementation exploring AI-mediated environments related to child safety and learning contexts.
Related Writing and Interpretation
Writing aimed at broader or domain-specific audiences, including interpretive essays and applied discussion, is available on the project blog. Explore More AI Ethics Articles on Our Blog
Governance & independence signal
The research presented here is independent and conceptual in nature. Reference implementations are provided for contextual grounding only and do not constitute validation, evaluation, or endorsement of specific outcomes or uses.