"I Just Don't Want My Work Being Fed Into The AI Blender": Queer Artists on Refusing and Resisting Generative AI

Quantitative signals: this appears to be a very new/low-adoption artifact (0 stars, ~6 forks reported but effectively no ongoing activity/velocity). With an age of ~2 days and velocity of 0.0/hr, there’s no evidence of a community, reuse, or operational footprint. What the project is: the README context indicates an arXiv paper (“Queer Artists on Refusing and Resisting Generative AI”) built from 15 semi-structured interviews. That makes the “core function” a research contribution (insights/findings) rather than software or an infrastructure component. Defensibility (why score=2): - No code moat / no production artifact: There’s no indication of a library, dataset tooling, governance engine, or reproducible pipeline that others would need to integrate. - The contribution is knowledge-oriented: Qualitative findings in a paper are valuable, but they are inherently replicable by other researchers (similar methods, similar interview studies, different cohorts). That is typically low defensibility for “software project” scoring. - Low adoption trajectory: 0 stars and near-zero velocity strongly suggest the repo is not yet serving as a de facto reference implementation or standard dataset. Frontier risk (low): Frontier labs are unlikely to build this as a direct product feature because it is primarily a social-science/rights framing and interpretive work, not a technical system they must compete with. They could read it and incorporate high-level policy/ethics considerations, but they are not likely to “integrate and replace” it as a deliverable. Three-axis threat profile: - Platform domination risk = high: Large platforms (Google/AWS/Microsoft, and also major GenAI model providers) can readily absorb the *conceptual* findings into internal policies, training-data governance processes, or product-level consent/dispute workflows. While they can’t “replace the paper” one-to-one, they can neutralize its practical impact by adopting generic governance measures without needing the repository. This makes the platform risk high in terms of real-world displacement of influence. - Market consolidation risk = medium: Ethics/governance research and policy documentation can consolidate into a handful of dominant actors (major labs, standards bodies, influential NGOs). However, qualitative research communities remain somewhat fragmented because they depend on cohorts, fieldsites, and narrative authority. - Displacement horizon = 6 months: Because the artifact is interpretive/research-based and not a tool with persistent integration, other papers, reports, or internal platform governance docs could overshadow it quickly—especially if more well-funded institutions publish similar or broader interview-based work. Key opportunities: - It could become a citation anchor for dataset governance, consent frameworks, and community-informed refusal practices in generative AI policy discussions. - If the authors later release interview instruments, coding rubrics, or a reproducible annotation framework, the defensibility could rise modestly (from purely theoretical to a reusable research toolkit). Key risks: - Low operationalization: Without open methods, instruments, or a standardized artifact that others can reuse programmatically, the repo’s impact will remain mostly academic and therefore easier to supersede. - Generic absorption by platforms: Platforms may adopt “best practice” language that renders the specific community findings less actionable or less distinctive.