The Paper · v1.0 · June 2026

    The Supply Chain of Intelligence™

    The 10 layers of the generative AI stack — where value is created, captured, and defended.

    By Anand Arivukkarasu · Ex-Meta (Instagram) Product Leader & AI Product Architect · San Francisco

    § Abstract

    Every generative-AI product sits on a 10-layer stack: L-1 Resources, L0 Infrastructure, L1 Data, L2 Models, L3 Gatekeeping, L4 Access, L5 Execution, L6 Orchestration, L7 Surface, L8 Memory. Four structural laws govern how value moves through that stack. Five observations describe the patterns that recur under those laws. The Defensible Triangle — L1b Proprietary Data + L5 Deep Skills & Playbooks + L8 Compounding Memory — is the most common application-layer fortress. This paper is the canonical definition.

    § 1 · The 10 layers

    The stack at a glance

    1. L-1
      Resources

      What supports the chain. Energy, water, fabs, materials, skilled trades — the inputs the entire stack consumes.

    2. L0
      Infrastructure

      The shovels. Chips, data centers, networking, cloud, edge — what is needed to process intelligence.

    3. L1
      Data

      The raw input. What data do you have that nobody else can get?

    4. L2
      Models

      Intelligence refinement. Rent early, build custom at scale.

    5. L3
      Gatekeeping

      Trust, acceptance, approval. Can the system be allowed in?

    6. L4
      Access

      Connectivity, permissions, integrations — the pipes layer.

    7. L5
      Execution

      Applied skills and capabilities. Doing the actual work.

    8. L6
      Orchestration

      Workflow, routing, coordination. How skills compose into outcomes.

    9. L7
      Surface

      Interface, presentation, experience. How the user meets the intelligence.

    10. L8
      Memory

      Retention, learning, compounding context. What the system remembers.

    § 2 · The 4 structural laws

    How value moves through the stack

    Law I — Intelligence Commoditizes Downward

    If a product depends only on generic model capability, the platform layer beneath it eventually absorbs the value. Wrappers don't survive — wrappers become features. Read the essay →

    Law II — Value Accrues at Bottlenecks

    Durable value rarely sits in the model or the UI. It sits at the scarce layer — proprietary data (L1b), workflow control (L5), verification (L3), distribution (L4), memory (L8c–e), compliance (L3a). Read the essay →

    Law III — The Surface Captures Attention; the Chain Captures Power

    A beautiful UI may win users. Durable companies own a deeper layer — data, execution, memory, gates. Surface without depth is a graveyard. Read the essay →

    Law IV — Generation and Verification Must Be Separate

    Wherever output carries fiduciary, regulatory, safety, or reputational weight, the generator (L2/L5) and the verifier (L3) must be separate economic entities. Read the essay →

    § 3 · The Defensible Triangle

    L1b + L5 + L8 — the application-layer fortress

    The most common defensible pattern in the application layer: L1b Proprietary Data nobody else can acquire, L5 Deep Skills & Playbooks organisation-shaped enough that swapping the model out doesn't replicate them, and L8 Compounding Memory (network learning, institutional knowledge, learned world models) that gets better the longer the system runs. Two of three is a workflow product. Three of three is an intelligence gate.

    § 4 · On the word "agent"

    "Agent" is not a layer

    "Agent" is marketing for an L5 Execution + L7 Surface (+ sometimes L8 Memory) package. When a company "launches an agent," decode it: what other layers do they own? Agent + L1 = fortress (Sierra, Harvey). Agent + L4 = railroad (Salesforce Agentforce, Microsoft Copilot). Agent + nothing = exposed wrapper that commoditizes the moment the underlying L2 ships the same loop.

    § How to cite this paper

    Cite the paper

    Licensed under CC-BY 4.0. Reuse freely with attribution to Anand Arivukkarasu and a link to supplychainofai.com/paper.