Journal Articles (Published)
● Ramesh, A. (2026). The Industrialisation of Institutional Appearance. AI & Society. DOI: 10.1007/s00146-026-03258-5
Full-text access to a view-only version here.
Conference Papers
● Reconfiguring Pedagogical Authority: Human–AI Feedback Workflows as Sites of TechnoPower in Professional Online Education. 4S Conference, Society for Social Studies of Science, Toronto, October 2026. Abstract submitted.
Book Chapters (Accepted, Chapter Forthcoming)
● Platforming Distribution: Ownership, Infrastructure, and the Consolidation of Television Power in India
Working Papers & Preprints
● Infrastructure Without Mandate: EMMRCs, NEP 2020, and the Monetisation of Public Educational IP in India. SocArXiv, 2026. https://osf.io/preprints/socarxiv. Systematic analysis of governance failure in India's publicly funded educational media infrastructure; introduces AI data extraction as a compounding governance risk.
● AI, Education, and the Architecture of Power. SSRN Working Paper, 2026 — Three interconnected essays arguing that AI governance in education lives in contracts, platforms, and defaults rather than policy frameworks. Link: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=6414938
Selected Public Writing
Independent essays, ar1123.substack.com:
● AI's Embrace, Extend & Extinguish Movement Is Here — primary-source analysis of the 2026 open-weights coalition letter, arguing the real policy fight is over regulatory safe-harbor status, not "open vs. closed" AI.
● What Would You Do? — response to the Dwarkesh AI blog prize on national AI strategy for non-frontier states; proposes a concrete, pre-committed kill-switch trigger and argues institutional capacity to invoke it — not technology — is the binding constraint.
● AI Safety Has a Geography Problem — structural critique of talent and compute concentration in AI safety research, with proposals for parallel research ecosystems, local safety benchmarks, and liability-based enforcement.
Work in Progress
● Designed a multi-agent debate experiment with four AI systems played student, grader, administrator, and auditor across a live AI-grading accountability gap, revised positions after seeing each other's arguments, then proposed fixes; a fifth AI synthesized, and a human expert audited that synthesis and caught it overstating the case for deliberate deception. Formulating a draft for publishing.