Choosing Friction (phirephoenix.com)

This essay argues for deliberately preserving “friction” — the difficulty and commitment involved in human creativity and communal life — by refusing to adopt generative AI. Framing the choice as philosophical rather than purely practical, the author links their refusal to concepts from anthropology (schismogenesis) and civic sociology (Putnam’s Bowling Alone): declining to use AI is a way to protect practices that require time, effort and mortality-aware choices (writing a novel, building community), rather than outsourcing them to tools that produce instantly marketable output. They contend LLMs and generative models aren’t doing real “thinking,” and that replacing human deliberation with frictionless automation diminishes meaning, craft and social bonds. For the AI/ML community the piece is a reminder that technical capability intersects with culture, politics and ethics. Key implications: generative models promise to remove cognitive labor and interpersonal friction, enabling subscription-driven scale backed by deep pockets but masking real social and environmental costs; model updates and deepfakes threaten trust and continuity; and automation risks concentrating power and fostering authoritarian tendencies by letting people avoid negotiating with others. The author’s stance reframes design questions for researchers and product teams: when does automation erode values worth protecting, and which frictions should we intentionally preserve to sustain art, community and democratic life?