CONCEPT
AI as Boundary Object
The thesis that large language models function as reifications and boundary objects rather than as community members — and that treating them as participants produces specific, predictable pathologies.
In 2023, Wenger and collaborators published an analysis of generative AI through the communities of practice framework, arriving at a precise conclusion: AI systems are sophisticated reifications incapable of participatory engagement in communities of practice. They function as the most powerful boundary objects the organizational world has produced — translating instantaneously between community vocabularies, coordinating work that previously required human brokers, generating outputs that have the form of communal knowledge. What they cannot do is participate. They have no self-authorship, no stakes, no identity implicated in the quality of a community's work. The distinction is not academic. It determines whether AI enhances communities of practice or substitutes for them — and the substitution, when it occurs, erodes the social infrastructure of learning itself.
In The You On AI Field Guide
Wenger's 2023 analysis was grounded in the three-decade theoretical framework he had developed: participation and reification as complementary processes, communities of practice as the primary sites of professional learning, the crucial role