101: Concept/Idea
- Requires:
- Enables:
An Idea is a self-contained package of knowledge. Think of it like a recipe card that has three parts: the ingredients (context), the rules for how the final dish should turn out (schema), and the finished dish itself (solution). It’s a permanent building block, not a temporary message.
This document explains the plan for building a new kind of web, made of connected, “living” documents. We’ll look at the main building block of this web (the Idea) and how we can find and share these Ideas using internet addresses (DNS).
To learn how these Ideas can be turned into active tools, read 103: Concept/Ideator. To see the different ways they can be hosted, check out 102: Concept/Sovereignty.
The Mechanics of a Living Web
The whole system is built on one big rule: the information itself contains all the rules. The only building block is a simple three-part package called an Idea. This setup means you truly own your creations and can take them anywhere, because there are no hidden parts or secret settings. You never get locked into one system.
-
Context: This is all the background info. It’s like the ingredients list and instructions for a recipe—everything the AI needs to get started.
-
- More at json-schema.org/
Schema: This is the blueprint. It defines the structure of the final result using a standard called
jsonschema. It tells the AI, “Your answer must have a title, a date, and a list of three points.” This allows any AI to understand and work with the Idea. -
Solution: This is the final output or the answer that the Idea produces.
Ideas are designed to be unchangeable. You don't edit an existing Idea. Instead, to build on a thought, you create a brand new Idea that points back to the old one. This creates a perfect, unbreakable history of how a thought grew over time, like stacking transparent pages on top of each other.
More Than a Prompt: A New Building Block
At first, an Idea might look like just a fancy question you ask a chatbot like ChatGPT. But it’s much more than that. The big difference is that we are moving from single, temporary chats to a system of lasting, connectable creations.
A simple chat question is temporary—you ask, you get an answer, and it’s over. An Idea is a complete, self-contained package. It bundles the question (context), the answer (solution), and the rules (schema) into one portable unit. It’s not just a question; it's the question, the answer, and the recipe that connects them. This allows us to build with them, creating a lasting system, not just having a one-time chat.
This makes an Idea a new kind of digital building block. You don't just “ask” an Idea a question. You can copy it, change it, and plug it into other Ideas to build complex chains of thinking, all without writing normal computer code. It’s a platform for creating, not just a tool for asking.
The Core Rules
To make sure this system is strong, honest, and can be used anywhere, every part of it must follow four main rules.
Recreatable Results
An Idea is designed so you can recreate its result. If you give the same context (ingredients) and schema (blueprint) to a similar AI, you should get a very similar solution (the final dish). This rule ensures that we can trace how a result was created. While different AIs might create slightly different results, the goal is for the final answer to be a direct result of its starting parts.
No Hidden Information
The AI can see the entire context when it's working. This means you can't use the context to store secret information that isn't directly related to the task. This rule is important to keep the Idea focused and honest about what it’s doing.
The Result Follows the Blueprint
The solution is the final state of the Idea. Since every solution must follow the rules of its schema, the state is always structured and predictable. This ensures that every Idea's result is clear, organized, and understandable to any other system.
Unchangeable
If you change the context or the schema of an Idea, it becomes a new Idea. An Idea is considered a simple update if you only add new fields to the schema (the blueprint). If you make a breaking change to the schema, you have to create a new version.
Sharing & Finding Ideas
For an Idea to be useful, other people have to be able to find it. To do this in a decentralized way (without a central directory), every Idea gets its own unique, global name, like a website address. This separates the Idea from where it's stored, making it truly portable and independent.
This works by giving each Idea a unique domain name and using internet address records (DNS) to point to its official file. The details of how this works are explained in the next document, 102: Concept/Sovereignty.