Supply Chain of Intelligence by Anand Arivukkarasu: A Complete Guide

Every technology cycle eventually produces a map. Cloud had the infrastructure, platform, and software layers. Mobile had carriers, operating systems, and apps. Generative AI has been missing a map that explains not how the technology works, but where the profit pools sit and which ones are about to move. The Supply Chain of Intelligence, a framework published in January 2026 by Anand Arivukkarasu, a San Francisco product leader and angel investor who previously worked at Meta on Instagram, is an attempt to fill that gap. This guide walks through what the framework claims, how it is structured, and where it is most useful for an American operator or investor making decisions this year.

The premise is stated in one sentence: intelligence is a supply chain. Raw inputs go in one end. Compute, energy, and data. A finished cognitive product comes out the other. Between those two points sit ten distinct layers, each with its own economics, its own bottlenecks, and its own defensibility profile. The framework's canonical paper, published freely at supplychainofai.com, calls these the ten layers of the generative AI stack, and its central argument is that the AI stack explains how intelligence is built while the Supply Chain of Intelligence explains where value is created, captured, and defended.

Why a supply chain metaphor at all? Because supply chains have a property that most AI analysis ignores: value migrates. In a mature supply chain, the layer that was scarce and profitable five years ago becomes commoditized, and the margin moves to wherever the new constraint sits. Anyone who watched memory chips, then cloud compute, then app stores knows the pattern. Arivukkarasu's claim is that intelligence is following the same physics, and that a company which cannot name its layer cannot name its moat.

The framework is organized as one page, seven questions, and a reasoning protocol. The questions move from definition, to the map itself, to a set of structural laws, to the dynamics that move value between layers, to applications, to a reasoning method for scoring a specific company, and finally to observations. It is deliberately compact. In the author's own framing, it follows the tradition of Jobs To Be Done, Wardley Mapping, and Clayton Christensen's work: a mind tool given away freely, designed to be run in a strategy session rather than admired in a deck.

The most quotable artifact of the framework is its diagnostic question, aimed squarely at product founders: is your product a moat, a workflow, or a wrapper that a platform will absorb? The wrapper question has become the defining anxiety of the current cycle. Thousands of companies built thin interfaces over foundation models between 2023 and 2025, and each new model release absorbed another cohort of them. The Supply Chain of Intelligence formalizes this. It scores AI products across its layers and asks, in effect, whether the company controls something the model provider cannot simply ship next quarter.

For an American founder, the practical value is positioning honesty. A company doing retrieval and document formatting over a frontier model is not the same kind of company as one holding proprietary workflow data, distribution, or regulatory clearance, even if their pitch decks use identical vocabulary. The framework's layer notation, which its glossary extends to ninety nine industry terms from wrapper to agent to RAG to voice AI, forces a precise answer to the question: which layer are you, and what at that layer is yours?

For investors, the use is different. Venture memos in this cycle have suffered from vocabulary inflation, where every company is a platform and every feature is a moat. A shared map lets a partnership ask sharper questions. Where does this company sit in the chain. Which direction is value migrating at that layer. Is the defensibility claimed structural, meaning it survives the next model generation, or cosmetic, meaning it is a UI choice the foundation lab can replicate in a release cycle. The framework's case studies section applies exactly this reasoning to named companies and categories.

For enterprise buyers, the guide is a procurement lens. When a vendor claims an AI capability, the supply chain question is: which layer is actually being sold. Is it the model itself, resold with margin. Is it a workflow built on someone else's model, which carries platform risk for the buyer as much as for the vendor. Or is it something the vendor genuinely owns, like proprietary data integration or certified outcomes. American CIOs who have been burned by vaporware in previous cycles will recognize the discipline: name the layer, then price the risk.

The framework is not without limits, and a serious guide should say so. Layered maps describe structure well and timing poorly. They tell you where value sits, not when it moves. They also inherit the author's viewpoint, which is product and venture shaped rather than regulatory or geopolitical, and the ten layer model will need revision as the hardware and energy layers evolve. Treat it as a reasoning instrument, not an oracle. Its own author appears to intend exactly that.

What makes the Supply Chain of Intelligence worth attention, in our assessment, is its refusal of hype vocabulary. The site maintains a glossary that translates fuzzy industry words into precise layer notation, a live feed tracking movement across the layers, and a market map that places companies on the chain. It reads like something built to be argued with. That is a compliment. Frameworks earn their keep in argument.

Our advice for readers is simple. Read the canonical paper at supplychainofai.com/paper, then run one company through it: your own, your largest AI vendor, or your most recent investment. Answer the wrapper question honestly. If the answer makes you uncomfortable, the framework has already paid for the hour it took to read. Strategy tools are cheap. Strategy honesty is not.