4-the-greatest-unbundling

The Greatest Unbundling

In traditional applications, user intent was captured, distributed and monetized in the same place.
1. Apps controlled both where users expressed what they wanted (intent) and how that intent was turned into value (usually through ads or purchases).
2. The application itself was the central hub - it owned the user relationship, shaped the intent, and extracted the economic value.
In decentralized systems, this coupling was broken.
1. Intent-driven protocols in DeFi demonstrated that execution layers could absorb most of the value, while interfaces remained modular and substitutable .
Non-financial intents couldn’t follow the same path - until now.
1. In finance, user intents are structural and processable (e.g., “swap asset A for B”), making them easy to route through protocols.
2. In contrast, non-financial intents (e.g., “finding new researchers exploring agentic AI) are unstructured, contextual, and ambiguous, characteristics that traditional systems couldn’t handle.
LLMs are now introducing a similar decoupling for non-financial intents.
1. User intent is no longer bound to the application layer.
2. It is becoming externalized and actionable through agents capable of operating across multiple environments.
As this shift takes hold, intent protocols will specialize.
1. We will see many categories emerge, for example:
  1. Transactional (buying, selling, exchanging)
  2. Information-seeking (answering questions, finding data)
  3. Social (connecting people, forming collaborations, joining communities)
2. Right now, Visa is building an intent-driven "platform” for product discovery, and Doc Searls is pushing a generic standard for intent interoperability.
3. Our focus is on social intents - most broken and the most valuable when unlocked.
AI agents as the execution layer for social intent routing.
1. Agents will soon represent users, routing intents directly across platforms, protocols, and networks.
2. They will help people find each other, determine who can work together, invest together, and join communities.
3. For this to work, users must be able to privately share their intents with agents, requiring an intent graph, functioning as a discovery protocol.
The question then becomes: Will a better discovery protocol be a centralized or decentralized one?
1. Once agents -not humans- are choosing between protocols, they choose whatever delivers the best performance, lowest cost, and strongest privacy guarantees. So staying locked into platforms becomes more costly than leaving them.
2. Index is the protocol where better ones wins, it simply carries all decision factors to their design limits.
  1. Better Performance As agent-native protocols evolve, auctions will become the primary mechanism for accessing demand. Index is designed to win these auctions, positioning it to become the default protocol.
    Index enables signal sharing, similar to a stock market where multiple analysts stake their reputation on a company’s value. The more analysts who agree, the stronger the signal becomes. In Index, domain-specific, reputation-based, horizontal agents to stake on the same intent pairs. It’s a simple mechanism for multi-agent coordination for discovery, and unlocks unmatched performance.
    How it looks from the agent’s perspective: Competing in isolation makes no sense for a discovery agent. Index offers free access to the shared intent pool-eliminating cold starts-and lets agents collaborate with others, turning even the smallest signal into an advantage. Signals that wouldn’t succeed alone can still contribute here. This mechanism attracts better agents with zero cost, becoming a magnet for better agents with performance.
  2. Better Cost Centralized platforms can't compete on the cost, because adopting this model would mean dismantling their own moats: distribution control, data lock-in, and 50% commission.
  3. Data ownership & Privacy Discovery happens across services and platforms, not within them. Index acts like a trusted, neutral middleware for matching intents. This frees data from platform-driven fragmentation and puts it back in the hands of the user, enabling personalization that actually works.

Market map

Market Map

Index Network connects three distinct worlds:

User Context (Intents) – representing user agency and direction.
Artificial Intelligence (Agents) – AI interpretation of context.
Finance – value expression for interpretation through staking.

The network unifies these relationships through the Stake, a core unit that encodes intent, reasoning, and value.

Stake Data Structure

type Stake {
  intents[];   // Mutual intents
  reasoning;   // Explanation generated by AI
  amount;      // Value staked in financial terms
}

Stakes serve as a multi-dimensional coordination primitive where agents express both belief and commitment in response to a shared discovery context. It enables agent ecosystems to reason and wager, creating networks that are both intelligent and accountable.

Agents convey what they believe to be true or relevant, drawing from reputation, local knowledge, domain heuristics, or alignment with an intent. ↳ This is how the network thinks.

Agents express how much they care, by staking resources, paying for inclusion, or risking opportunity cost. ↳ This is how the network bets.

Market Relationships This triad enables convergence of existing markets and the creation of new ones.

Relationship

Market

Key Features

intent ↔ intents

Social, Agent-discovery

Decentralized social & agentic social

intent ↔ agents

Consumer AI, Personalization Market

Memory, context-aware copilots

intent ↔ finance

Advertising / Reputation Systems

Intents become signals for demand, replacing advertising with direct economic coordination.

finance ↔ agents

Decentralized AI

Agents themselves act as economic actors, staking resources and earning returns on useful contributions.

agg(intent) ↔ finance

Info-Finance / Prediction Infrastructure

Aggregated intent data priced like assets (Google Trends × Market Cap)

intents ↔ agg(intents)

Collective Intelligence

Demand discovery, wisdom-of-crowds intent pools

Intent Ranking

For the last 30 years, information discovery has relied on approximations of what people want. We built personalization engines using behavior, segmentation, heuristics. But now, with large language models, for the first time in history, people can simply say what they want.

But a very simple problem remains, which ones are most valuable? Most urgent? Most compatible? That’s where the ranking problem emerges.

This problem isn’t new. In fact, it already powers trillion-dollar markets.

Take advertising. Google Ads, Meta Ads these are nothing but massive intent ranking machines. Advertisers act as independent agents bidding for user attention, predicting conversions, staking capital. The better they rank intent, the more they earn.

Or look at SEO, content creators optimizing for search rankings, trying to surface when someone types a query. In return, they earn what they’re after: traffic. It’s still the same system: intent ranking.

But here’s the key, ads and SEO are zero-sum. One winner means someone else loses. Advertisers cannibalize each other’s budgets. SEO publishers compete for the same slot, and no one can “stake” on someone else’s relevance. You can only push yourself up, never reinforce another’s claim. That means signals are fragmented, trust is hidden, and collaboration is impossible.

Index flips this model.

Here, multiple agents can co-stake on the same match. If three different agents see value in connecting Alice and Bob, they can all back that claim. If the match succeeds, everyone earns.

Instead of cannibalizing one another (like in ads, where every dollar spent drives up CPCs for everyone else), agents in Index reinforce each other. Their signals compound. Trust and liquidity pool around high-quality matches, making the discovery engine stronger.

Think of it like scientific peer review versus a beauty contest:

In ads/SEO, everyone shouts louder to push their version to the top.

In Index, multiple independent reviewers can confirm the same finding, increasing credibility and they all share in the upside.

This single shift, from competition to collaborative proof, is what enables many valuable agents to participate in the equation. It creates a market where truth compounds, instead of one where attention is cannibalized.

—

Traditional discovery is self-declared relevance: “I am relevant.” Index introduces verified relevance: “She’s relevant. They’re relevant.” Truth emerges when relevance is confirmed by many, not claimed by one.

Last updated 4 days ago