> For the complete documentation index, see [llms.txt](https://evrylife-web-3-explorer.gitbook.io/evrylife-foundation-wiki-docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://evrylife-web-3-explorer.gitbook.io/evrylife-foundation-wiki-docs/evrylife-application/user-intents-ai-agents-and-dao-intents-in-the-evrylife-world-explorer.md).

# User Intents, AI Agents, and DAO Intents in the Evrylife World Explorer

The system has two main parts:

* **User Intents** Personal or channel-level AI features that profiles and channels can turn on to automate tasks.
* **DAO Intents** Platform-wide or governance-related AI tools that help the whole community operate fairly.

Both are designed so AI is always optional, user/DAO-controlled, replaceable, and auditable.

### User Intents

User Intents are modular AI features that a profile or channel can activate and adjust.

* They live in the user’s Ceramic profile or channel stream (private when needed via Lit encryption).
* Users toggle them on/off .
* Users choose their preferred decentralized AI provider 
* New intents can be proposed and added through zkrPOGO

### AI Agents in the Explorer

AI agents appear as discoverable items in the Evrylife Explorer (same as profiles, channels, quests, or products).

* They are tagged and categorized in the Web3 Meta Tree.
* Agents exist only to assist existing profiles and channels — they cannot be standalone profiles or channels.

### DAO Intents

DAO Intents are AI tools that support platform operations and governance.

* They are strictly governed and audited by the community.
* New DAO intents are proposed and approved via zkrPOGO + Snapshot EAIP.
* They run only when triggered by zkrPOGO consensus.
* They use multiple independent AI providers to ensure fairness and avoid bias.

### Decentralized AI Service Providers

Users and the DAO choose which AI backend powers any intent or agent.

The goal is maximum decentralization, transparency, and resistance to bias or single-point control.

Users can connect their preferred decentralized AI provider. The DAO approves and can add or remove providers over time.

For DAO-level tasks (especially moderation, reviews, and consensus), multiple independent AI providers run in parallel. Results must reach strong agreement, or the task escalates to human zkrPOGO moderators.

All AI inputs, outputs, and decisions are logged on Ceramic for full audit-ability.

This structure keeps AI helpful while remaining under full user and community control. The DAO can evolve or restrict any part of the system at any time through governance.

**Decentralized AI bootsrapping considerations for Evrylife**

| Provider             | Strengths for Evrylife                             | Current Status                          |
| -------------------- | -------------------------------------------------- | --------------------------------------- |
| Virtuals Protocol    | Native agent framework + x402 revenue sharing      | High priority – best long-term fit      |
| Bittensor            | Subnet specialization (tagging, moderation, RAG)   | Strong fit – multiple subnets usable    |
| Akash + Hugging Face | Fully open-source models, cheap compute, RAG ready | Fallback / sandbox – already integrated |
| NEAR AI              | Confidential compute, low/no gas, fast inference   | Good fit for private intents            |
| ASI / Fetch.ai       | Autonomous agents, governance-ready                | Future candidate                        |