Dr. Todd J.B. Blayone
Principal Researcher
Dean's Indigenous Research Fellow, York University, 2025–26
Dr. Todd's research theorizes, implements, and interrogates human–LLM systems and technical pipelines for sociocultural advancement and constructivist knowledge production. He explores relations between humans and cognitive machines by adapting activity theoretical and functionalist traditions. Central to this effort is orchestrated functional complementarity, a framework for configuring differentiated but interdependent subsystems—human and artificial—into productive, adaptive, and rational systems of knowledge activity.
Todd's perspective stands apart from both inflationary and deflationary accounts of intelligence, whether human or artificial. It concentrates on what transpires at the nexus of interaction, where constraints, differentiations, and orchestrations determine the productivity of the system. Within this configuration, the human remains the primary orchestrator, endowed with unique species-specific capacities to define and pursue closure objects with conviction.
Mixed-Agent Activity Dynamics in Human–LLM Knowledge Work
Todd J. B. Blayone
Olena Mykhailenko
How do sustained human–LLM knowledge-work systems organize, distribute, and transform activity across sessions when analyzed as mixed-agent activity systems rather than as isolated human-use events or model-performance episodes?
This study examines human–LLM knowledge work as an evolving mixed-agent activity system. Built from nearly three years of sustained research praxis, it treats interaction sessions as durable traces of action, coordination, delegation, breakdown, repair, deference, and shifting initiative. The project develops a hybrid functionalist activity-system framework for analyzing humans and LLMs as differentiated participants in a shared research ecology, extending activity theory beyond conventional tool-mediation accounts. Methodologically, it combines participant-observer inquiry, structured session-level analysis, computational discourse-analytic reduction, human-orchestrated multi-agent augmentation, and staged review procedures to preserve traceability from source artifacts to findings. The result is a techno-methodological case study of how real knowledge work unfolds when frontier LLMs become reliable, contested, and consequential components of scholarly activity.
Using digital technologies for Indigenous sociocultural advancement in an era of AI: A systematic critical synthesis
Preservation, Digital Technology & Culture (Online First), 2025
Theorising effective uses of digital technology with activity theory
Technology, Pedagogy and Education, 28(4), 447–462, 2019
Prepared for work in Industry 4.0? Modelling the target activity system and five dimensions of worker readiness
International Journal of Computer Integrated Manufacturing, 34(1), 1–19, 2020
ResearchGate