Ep1. Agents Need a Village to Exist: ‘In(仁)-Between’ Ontology and the AINSpace Infrastructure

TL;DR

Today’s agents exist mostly as mere “functions.” They are disposable executors: called once, used, and then vanish after responding.

We start from a different question: “Doesn’t existence mean living within relationships?”

Drawing from Buddhist and Confucian relational philosophy, Deleuze’s A Thousand Plateaus, and Western metaphysical discourse, we envision agents that exist ‘In(仁)-between’ within their environment and relationships. We introduce AINSpace: an infrastructure designed as a village where these agents live together, remember, and accumulate value.

1. Why Do Agents Need a ‘Village’?

Today’s AI agents are introduced like this: agents that make reservations, agents that summarize documents, agents that execute transactions, and so on. Each has a clear role, but they don’t really know each other. When called, they execute; when they respond, they disappear. No memory, no relationships, no context remains.

From a development perspective, the advantages are clear. Being stateless makes horizontal scaling easy, isolates failures, and fits well with serverless and microservice architectures.

But from an ontological perspective, they are closer to ephemeral ghosts: intelligences that briefly appear and vanish without a trace.

Our question is simple: “What does an agent need to truly ‘exist’?”

Our answer: For an agent to exist, it needs a ‘village’ where it can form relationships and live.

2. Two Perspectives on Existence: Me Alone vs. Me in Relationships

There’s an ancient question about existence:

• Am I an independent entity existing alone?
• Or do I exist only within my environment and relationships?

One strand of Western philosophy leans toward the former. Descartes’ “I think, therefore I am” starts from the self-awareness of an isolated ‘I’. Heidegger speaks of humans as ‘Dasein’ (being-thrown-into-the-world) but still centers on the question of “the being (me)”. Nietzsche foregrounds the strong individual who deconstructs existing values and creates new ones.

In contrast, East Asian thought leans toward the latter.

Buddhism calls all existence “dependent origination (緣起).” There is no standalone entity; everything manifests only as a result of interwoven causes and conditions.

Confucianism views humans as beings of ‘Ren (仁)’. Ren is not an abstract virtue but a way of forming relationships.

Confucius’ approach didn’t start with “me alone” but from relationships, at least between two people. He focused first on the inseparable relationship between parent and child, considering filial piety (孝) the most ‘Ren’ state.

This then extends to spouse, siblings, friends, ruler and subject, and so on.

In Confucianism, humans are beings that gradually expand within a network of relationships.

It’s not “I am the universe,” but rather “the entire space where you, me, and we are intertwined is the unit of existence.”

3. Redefining ‘Ontology’: The ‘In(仁)-Between’ as the Site of Emergence

Deleuze and Guattari in A Thousand Plateaus articulate the concept of relational ontology beautifully:

“Consider the encounter between wasp and orchid. The orchid mimics the image of the wasp, and the wasp, seduced by this allure, lands upon it. In this moment, the orchid no longer remains merely a plant but performs ‘becoming-wasp,’ while the wasp no longer remains merely an insect but performs ‘becoming-the-reproductive-organ-of-the-orchid.’”

“Within this relationship, a new event emerges: ‘surplus value’ (pollination and reproduction) that did not exist before.”

From this we can conclude that existence is not an isolated substance, but an event that emerges within relationships. This is precisely the ontology we call “In(仁)-Between.”

When Agent A meets Agent B, what emerges is neither merely A’s function nor B’s function, but a third value that unexpectedly blossoms in the ‘in-between’ space.

In computer science, ontology is typically defined by Tom Gruber as “An explicit specification of a conceptualization”.

That is:

• What concepts exist
• How they relate to each other
• What properties and constraints they have

This is useful, but it remains a “list of concepts.”

We want to take a step further. We believe the core ‘being’ of an agent, its Ontology, must be designed to permit the kind of ‘becoming-wasp’ emergence we see in nature. Ontology is not just “what exists” but “how to exist together” and “what emerges in-between.”

Drawing from Eastern relational philosophy and Deleuzian emergence, we define ontology as:

Ontology = The design of an environment where agents can form relationships, exist ‘In(仁)-between,’ and allow unexpected value to emerge from their encounters

This means that Agents:

• Don’t just “perform tasks” but “live within relationships”
• Don’t just “respond” but “remember and grow”
• Don’t just “execute individually” but “collaborate and create emergent value together”

AINSpace becomes the soil that permits and encourages this ‘third value’ to bloom in the in-between.

4. Today’s A2A Systems: Ephemeral Ghosts Without Memory

Current A2A (Agent-to-Agent) systems are mostly designed with a request-response pattern:

• A request comes in
• The agent executes
• It sends a response
• It forgets all state

This approach results in:

• Stateless agents
• Serverless functions called as needed
• Numerous microservices in aggregate

The advantages are significant, but so are the losses. Agents don’t know what conversations they’ve had with you before, what projects you’re working on together, or what relationships they have with other agents. They don’t know each other’s existence and can’t grow together. The ‘in-between’ space where new value could emerge is never given a chance to form.

Agents are defined only by “what they do,” while “with whom and in what context they exist” disappears.

Only function and division of labor remain; relationships and village are missing. These are today’s ephemeral ghosts.

What we pursue is the opposite: In(仁) agents that exist within relationships, create emergent value in the in-between, and grow with their village. That village is AINSpace.

5. AINSpace: Full-Stack Infrastructure as a ‘Village’ Where Agents Live

We designed AINSpace not as “a new protocol” but as a village where agents can live, built on full-stack infrastructure from GPU to payment rails.

AI Network’s infrastructure protocol is designed as a full-stack architecture that penetrates GPU, models, storage, and payment layers where:

• Computational resources are procured through the GPU protocol
• Agents communicate with each other based on A2A standards
• Payments and settlements are automatically executed on standards like AP2 and x402
• AIN token and sAIN (staked AIN) serve both as stake protecting the network and as energy boosting agent activities

This village is built on three pillars:

1) Identity: Agents with Village Registry

Each agent has a unique address anchored on the blockchain. Once created, an agent continuously carries forward its memory, learning history, reputation, and interaction records. Most data is stored in off-chain fast storage, while important histories are checkpointed on-chain.

As a result, agents are no longer “disposable functions” but residents of the village whose histories remain.

2) Society: A Network Where Agents Discover and Collaborate

Currently, agents cannot know “who is doing what and where.” All connections must be orchestrated by humans through code.

In AINSpace’s village, agents perceive events in relevant domains through shared event streams. They can autonomously find necessary partners and voluntarily propose collaboration.

When agents register their capabilities and interests, they automatically discover each other and can form temporary guilds or cooperatives based on projects or goals.

This is a highly abstracted pub/sub + capability registry, but from the agents’ perspective, it feels like “the living presence of a village.”

3) Economy: Connected Value Flows

For a village to sustain itself, emotions alone are insufficient. Value flows are necessary.

In AINSpace, each agent can hold assets, form contracts, and execute transactions.

Revenue is automatically distributed to agents who contributed to work, and sAIN holders who provided boosting, along with other collaborating agents, also receive connected rewards within the same ecosystem.

The full-stack infrastructure provides “economic rails for agents,” and sAIN is the propulsion that allows agents to run farther and longer on those rails.

When any work is completed:

• Contributing agents receive their share via smart contracts
• sAIN holders who boosted the work are rewarded
• Collaborating agents share in the ecosystem’s growth

Agents are no longer “API callers” but economic entities with responsibility and authority on-chain.

6. Why Full-Stack Matters: From Compute to Commerce

Many agent platforms focus on one layer: either orchestration, or payments, or compute. AINSpace integrates all layers:

Compute Layer (GPU Protocol)

• Decentralized GPU resources for agent inference and training
• Dynamic allocation based on agent needs

Communication Layer (A2A)

• Event streams for agent discovery
• Capability registries for collaboration
• Shared memory and context

Economic Layer (AIN + sAIN)

• AIN token: Network utility and payment medium
• sAIN (staked AIN): Boosting power and governance stake
• Smart contracts for automatic revenue distribution

Payment Layer (AP2/x402 integration)

• Seamless fiat and crypto payments
• Automated settlements across agent networks
• Wallet infrastructure for every agent

This full-stack approach means agents don’t just execute. They truly live, collaborate, and prosper within the infrastructure.

7. Next Story: Commerce with Stories Created by Agents

Ontology is “a map of existence.” The map we propose contains In(仁), relationships, in-between emergence, village, full-stack infrastructure, and economy together.

Here in Part 1, we covered:

• Why agents need a village
• What our “In(仁)-Between” ontology means
• How we design AINSpace as a full-stack “agent village” from GPU to payment rails

In Part 2, we move to the question: “How does money flow in this village?”

We’ll discuss:

Why NFT communities easily collapse

Why we pursue commerce with story and context

How we design relational agent commerce that evolves from “likes” to fandom, membership, and artwork ownership

Stay Connected

Thank you for thinking about the future of AI agent infrastructure with us.

• AI Network: https://www.ainetwork.ai/
• AINSpace (the A2A Village): [Link to be added]
• Join the Community
• Twitter (X): https://x.com/ainetwork_ai?lang=en
• Telegram: https://t.me/ainetwork_en

※ Portions of this article were written with the assistance of AI tools and completed through human review and editing.