1. Title
A Public Layer for Ideas in AI — an “Idea Commons”
2. Author
Konstantin Fioktistov
3. Publication Date
October 4, 2026
4. Description of the Idea
AI already helps millions of users create new ideas. Many of these ideas remain inside private conversations and disappear, even though some of them could be useful to other people, companies, or could eventually become real products.
An idea does not necessarily have to be implemented by its creator to have value. It could be more useful to make it possible for good ideas to spread freely, be discussed, developed, combined with other ideas, and find people who can turn them into reality.
I propose creating a public layer for ideas inside an AI platform — an “Idea Commons.”
AI would help users identify, formulate, structure and critically evaluate ideas. Users could then deliberately publish selected ideas into a shared public space where other people could discover, evaluate, discuss, develop and implement them.
The Idea Commons would not necessarily need to be a separate website. It could become a structured public layer integrated directly into the AI platform.
5. Problem It Addresses
Today, many ideas are created during conversations with AI but remain isolated inside private conversations.
This creates several problems:
- potentially valuable ideas are difficult for others to discover;
- ideas are often forgotten by their creators;
- there is no common public record of their development;
- people who have useful ideas but lack the resources or motivation to implement them have few ways to pass them on;
- AI-generated ideas are disconnected from subsequent human evaluation and development;
- when someone asks an AI about possible solutions to a problem, the AI cannot easily point to a structured history of ideas proposed and developed by other people.
There is also a broader cultural problem: ideas are often considered valuable only when their creator personally implements them. But an idea can have value independently of who ultimately turns it into reality.
6. Proposed Solution
Create a structured public space for ideas directly within an AI platform.
During a conversation, AI could recognize that the user has formulated a potentially interesting idea and suggest:
Share this idea publicly →
This should be a deliberate human action rather than automatic publication.
The user would review the AI-structured version, log in if necessary, and explicitly confirm that they want to publish the idea under their name.
A published idea could contain:
- title;
- author;
- publication date;
- description;
- problem it addresses;
- proposed solution;
- possible applications;
- AI analysis;
- existing analogues;
- community ratings and comments;
- related and derivative ideas;
- information about subsequent development or implementation.
AI evaluation should not simply produce a flattering score. It should provide a critical analysis of strengths, weaknesses, risks, existing solutions and factors that could change the assessment.
Other users could:
- discover and search for ideas;
- rate them;
- discuss them;
- improve them;
- create derivative ideas;
- combine them with other ideas;
- report implementations.
This could create an evolving history such as:
Idea A → A.1 → A.1.1 → implementation
Over time, AI could become an interface to this accumulated idea space.
For example, a user could ask:
“What interesting ideas exist for automotive diagnostics?”
The AI could retrieve relevant ideas from the Idea Commons, including highly rated, recently proposed, actively developed or already implemented ideas.
The important principle is that AI helps create and structure the idea, but humans decide what enters the public space.
Ideas would remain freely usable. They could be developed, combined, implemented and used commercially. Attribution and provenance could remain part of their public history.
Voluntary rewards or other mechanisms for recognizing contributors could be added later, but they are not the foundation of the concept.
7. Possible Applications
The same infrastructure could support many types of ideas:
- product and startup ideas;
- scientific and technological hypotheses;
- engineering solutions;
- social and organizational innovations;
- creative concepts;
- improvements to existing products and services;
- open-source projects;
- educational ideas;
- ideas generated through collaboration between humans and AI.
The system could eventually become a searchable public knowledge layer specifically focused on ideas and their evolution.
It could also support AI agents in the future: agents could discover relevant ideas, evaluate them, combine them, develop implementation plans, or identify people and organizations that might be able to implement them.
8. AI Analysis of Strengths and Weaknesses
I asked OpenAI to evaluate this concept as a ruthless critic, using criteria including user motivation, content quality, AI evaluation, human validation, discovery, integration with AI, network effects, retention, scalability and risks.
Overall assessment: approximately 87/100 as a product concept for a major AI platform.
Strengths
1. Natural integration with AI
AI is already present at the moment when many ideas are created. The proposed feature builds on an existing user behavior rather than requiring users to visit a separate platform.
2. Low friction for publication
The idea can move from a private conversation to a public structured object with relatively little effort.
3. Human-controlled publication
Requiring an explicit human action can distinguish publicly endorsed ideas from the enormous amount of content that AI can generate automatically.
4. Potential network effects
The more ideas, evaluations, discussions and implementations the system accumulates, the more useful the idea space could become.
5. Recognition as a user motivation
Users do not necessarily need financial rewards to contribute. Recognition, reputation, feedback and seeing one’s ideas evolve can provide intrinsic motivation.
6. Long-term value of the accumulated data
A sufficiently large collection of structured ideas, their relationships, evaluations and implementation histories could become a valuable intellectual resource.
Weaknesses and Risks
1. Content quality and noise
A public system could quickly accumulate enormous numbers of trivial, repetitive or low-quality ideas.
2. AI-generated spam
Even if publication requires human confirmation, users could potentially publish large numbers of AI-generated ideas with little personal contribution.
3. Difficulty of evaluating ideas
AI may be unable to reliably distinguish genuinely valuable ideas from ideas that merely sound convincing. Human ratings would therefore remain important.
4. Unclear boundary between an idea and ordinary content
The system would need mechanisms for distinguishing an actual idea from a question, opinion, observation, suggestion or ordinary piece of content.
5. Uncertain long-term motivation
Recognition may be sufficient to attract contributors, but it is uncertain whether it would be strong enough to make users return regularly.
6. Discovery at scale
If millions of ideas are accumulated, simply publishing them is not enough. Ranking, search, relationships between ideas and high-quality recommendation mechanisms would become critical.
Despite these challenges, the concept is worth exploring because it leverages something that AI platforms already have: the moment when people formulate new ideas.
9. Existing Analogues
There are no obvious direct equivalents to the proposed system, but several existing platforms demonstrate parts of the concept:
- GitHub — ideas and technical work can be publicly developed, forked, improved and attributed, although the primary object is code rather than ideas.
- Wikipedia — demonstrates how a shared public knowledge space can be collaboratively developed and continuously improved.
- Reddit — demonstrates how public contributions can receive community evaluation, discussion and social recognition.
- Stack Overflow — combines user-generated content, community evaluation, reputation and structured discovery.
- Open-source communities — demonstrate that people can voluntarily contribute intellectual work without requiring direct financial compensation.
- Scientific publishing and preprint platforms — demonstrate the value of publicly attributing ideas and establishing a visible history of intellectual contributions.
The proposed Idea Commons would combine elements of these models, but with the idea itself — rather than an article, question, answer or piece of code — as the primary object.
PS
**This idea itself was prepared by me in accordance with the principles described in the proposal: it is being shared openly, without seeking exclusive ownership, so that it can be evaluated, discussed, developed, or used by others.
PPS
The “community ratings and comments” section is currently empty. I invite the community to be the first to fill it in — for the first idea in the Idea Commons.**