The Estimate — Summary
The man with the notebook, first introduced in Part 74, has been doing arithmetic. Not systems-level calculations but the simpler and more difficult arithmetic of whether an idea can survive contact with a budget. He has a habit of underlining totals twice: the first underline is the number, the second is the commitment to take it seriously.
The previous four essays described an architecture: a skeptic, seven philosophical operations, a bias-in-intent layer, a retroductive method, and the Intersectional Systemic Harm Index as the operational layer. This essay asks what it costs to build a working pilot in one domain (maternal health), in one country (India), for eighteen months, long enough to find out whether the idea works in practice or only in essays.
The components, costed in Indian rupees and dollars: an epistemic AI trained on the Indian maternal health literature including gray literature and traditional knowledge (₹3.2 crore / $400K). A skeptic architecture with seven tradition-specific operations, each requiring curated training sets developed with scholars in each tradition (₹4 crore / $500K). The ISHI layer productionized for deployment, including data integration and a calibration study (₹3.2 crore / $400K). The human layer: domain expert panels of ASHA workers, auxiliary nurse midwives, district health officers, and philosophical consultants, compensated at rates that reflect the value of their knowledge (₹3.2 crore / $400K). Infrastructure, project management, regulatory navigation, field deployment (₹3.2 crore / $400K).
Total: ₹16.8 crore ($2.1 million). India’s National Health Mission budget is roughly ₹32,000-40,000 crore annually. A single AI deployment by a major health tech company runs ₹40-120 crore. This pilot is a rounding error.
The architecture is affordable. The question is whether anyone’s budget has a line item for questioning the questions the rest of the budget is built to answer.
Realistic funders: the Tata Trusts (institutional patience, public health history), the Gates Foundation’s India office (if positioned as maternal health equity), ICMR with a multilateral partner, a consortium of Indian research universities, or an AI safety body for the skeptic component specifically. The realistic path is coalitional. No single funder covers the full scope. This is messy and also how most consequential things get built in India.
The money is not the hard part. The hard parts: institutional tolerance for uncomfortable outputs, since every component produces findings that tell institutions their categories are wrong. Data access across India’s fragmented health information systems. Community trust in populations that have been studied repeatedly without visible benefit. Philosophical depth without tokenism. And sustainability beyond the pilot: who maintains the systems when the institutional pressures to smooth the skeptic’s outputs begin?
The man has a page with the number circled. Two underlines. He knows the number is not the obstacle. The women in the PHCs are already living inside the stratum gap. They do not need a pilot to tell them this. The pilot is for the institutions. It is the instrument by which institutional knowledge is forced to confront its own insufficiency, documented rigorously enough that the confrontation cannot be administratively absorbed.
The tire still has a slow leak.