Journeys for AI
Building the scaffolding intelligence needs — data, evaluation, governance and the human-in-the-loop craft that turns 60% into production-grade.
Indian Knowledge Systems × Artificial Intelligence
Curated AI journeys — for AI, from AI, and of AI. Where the rigour of Nyāya meets the frontier of machine intelligence.
Curated AI Journeys
Every engagement with intelligence moves along one of three ayanas — a path, a passage, a turning of the sun.
Building the scaffolding intelligence needs — data, evaluation, governance and the human-in-the-loop craft that turns 60% into production-grade.
What machine intelligence gives back — insight, acceleration, and a mirror that reveals where our own reasoning is thin.
The inward path — what intelligence is. State, models, reasoning and meaning, examined through the lens of Indian epistemology.
Nyāya · न्याय
Scale gave us fluency. It did not give us pramāṇa — valid means of knowledge. The Indian schools of logic spent two millennia asking exactly the questions modern AI now stumbles over: what counts as evidence, what is a good model, and how do we know what we know.
A fluent answer is not a valid one. We build systems that can show their means of knowing, not merely their confidence.
Extracting reliable behaviour from a probabilistic substrate is the hard problem of our decade. We treat it as an engineering discipline, not a prompt.
Parsimony, falsifiability, composability, transparency — the five-century-old virtues of a good model, applied honestly to LLMs.
A dharmic stance on data, disclosure and displacement — technology that is answerable to the people and society it reshapes.
The Readings
Long-form thinking on the epistemology of AI — each essay distilled into a curated, infographic reading you can traverse in minutes.
Why today's AI is a global lab, and determinism is the prize.
Read journey →Testing LLMs against five centuries of model virtues.
Read journey →Org silos, 60–70% quality, and the labour paradox.
Read journey →The second avatar of multi-agent systems, and "agent babel".
Read journey →Frontier, current and old knowledge — and where you sit.
Read journey →IP when disclosure feeds the very machine you compete with.
Read journey →State as a network of models — never complete, never fully seen.
Read journey →A dharmic conscience
Bengaluru · Chennai · Bay Area. We think AI needs its own Ayanam — a considered path, not a stampede.