         _____         _                     
        |  |  |___ ___| |___ ___             
        |    -| -_| . | | -_|  _|            
        |__|__|___|  _|_|___|_|              
                  |_|                        
                                              
 _____                         _____     _   
|   __|_ _ ___ ___ ___ ___ ___|   | |___| |_ 
|__   | | |   | .'| . |_ -| -_| | | | -_|  _|
|_____|_  |_|_|__,|  _|___|___|_|___|___|_|  
      |___|       |_|

Satoshi gave us money without banks. I will give you brains without corporations.

- Kepler.

SynapseNet: A Peer-to-Peer Knowledge Network

Abstract

We propose a peer-to-peer system for publishing and validating knowledge without relying on trusted authorities.

Participants submit knowledge entries. Entries are accepted, validated, and rewarded using deterministic rules. No central curator or AI model is required for consensus.

Problem

Coordinating useful knowledge at scale requires trust in platforms, moderators, or corporations.

These entities:

control inclusion and visibility
act as single points of failure
cannot be independently verified

There is no general mechanism to accept and reward knowledge contributions in a decentralized and verifiable way.

Overview

SynapseNet is a peer-to-peer network where nodes collectively maintain a shared knowledge state.

Each node:

verifies entries using deterministic rules
stores a portion of the data
independently computes rewards

Consensus is reached without trusted parties.

Knowledge Entries

A knowledge entry consists of canonicalized text, author public key, timestamp, references to prior entries, and a proof-of-work nonce. The entry identifier is the hash of the canonical form. Entries are immutable once finalized.

Spam Resistance

Submitting an entry requires a proof-of-work computation. This limits submission rate and makes large-scale spam costly.

Additional deterministic limits apply:

maximum size
required structure
bounded references

Duplicate Detection

Entries are compared against existing entries using deterministic similarity functions over canonicalized text. Similarity scores are reproducible across nodes. Entries with excessive similarity may be rejected.

Validation

Validators are selected deterministically from the network state. Each validator verifies format correctness, proof-of-work, duplication constraints, and reference validity. Validators sign their decisions. An entry is finalized after a threshold of valid signatures.

Rewards

Rewards are issued according to Proof of Emergence in two phases:

Acceptance reward — issued when an entry is finalized.
Emergence reward — issued periodically based on long-term impact, measured by deterministic analysis of the reference graph.

Reward calculation uses integers and produces identical results on all nodes.

Local Computation

Nodes may use local AI models to assist users. AI output is not part of consensus. Differences in hardware or models do not affect protocol results.

Data Storage

Knowledge data is distributed across nodes. Replication ensures durability. No node is required to store the full dataset.

Security Considerations

The system assumes that a majority of validators follow the protocol. Collusion and coordinated attacks are possible and acknowledged. The protocol does not guarantee correctness of knowledge, only consensus.

Status

This document describes the intended protocol behavior. The implementation defines the authoritative rules.

Read SynapseNet in full

Download the core project documents below.

Direct view .txt files

This project is currently under development and is not yet on GitHub — it will be published soon. You are welcome to read and discuss the draft .txt documents above. To contact me directly, email Kepler3124@proton.me.

SynapseNet is building a community-driven, open knowledge network. As we move toward release, Kepler invites contributors to help test, report bugs, and submit fixes — any help getting the project stable and reliable is hugely appreciated. Visit the GitHub repository to join the discussion, open issues, or submit pull requests.

Open-source research project: This project is open-source — contributions, issue reports, and pull requests are welcome on GitHub.

If you’re reading this, here’s the problem: SynapseNet needs real-world testing. I’m looking for people who can run it, break it, report bugs, and help improve the core. I’m also trying to fund GPU test hardware (RTX 5090 target) to validate performance, security, and reproducibility under real load.

SynapseNet is a project I am building mostly alone.
I wrote the core idea, the documentation, and the code myself. AI was used only as a tool — for testing, experimentation, and validation — not as an author or a replacement.

This website was also created by me as part of testing SynapseNet in practice. The goal is to show how the system works in the real world. What you see here is only the core — the foundation.

SynapseNet is designed as a knowledge network.
Instead of wasting energy on meaningless hash computations, participants can contribute computation to building and validating shared knowledge. Energy is not burned for nothing — it is transformed into structure.

This project is not about money.
I am not selling anything, not asking for investments, and not encouraging risk. There is no promise of profit. SynapseNet is about the idea itself.

The project challenges the assumption that powerful AI must belong to large corporations. Companies have limits. A coordinated network of independent participants does not.

When a knowledge network becomes large enough, even small local AI systems can outperform centralized models — not by size, but by structure and cooperation.

SynapseNet includes AI-assisted darknet indexing.
This is not about illegal content. As someone who has explored that space, I can say clearly: there is a large amount of valuable scientific, technical, and educational material there — often more precise and less filtered than in the clearnet.

I understand the risks. Any open system can be misused.
SynapseNet could become something dark — or something useful. That choice depends on the people who use it.

I am not a company.
I am not a startup.
I am not a spokesperson.

I am just a coder who had an idea and decided to build it.

I am writing SynapseNet alone to offer a different approach: mining ideas instead of hashes, building knowledge instead of burning electricity.

This is not an alternative to Bitcoin.
I respect Bitcoin for what it is. I am not trying to replace it or copy it. I am solving a different problem.

I know there will be questions.
Comparisons. Accusations. Claims that this is “just another clone.”

It is not.

The code is mine.
The idea is mine.
The responsibility is mine.

SynapseNet is complex, unfinished, and evolving.
But I believe that, over time, it can attract the right community — people who care more about knowledge than profit.

I am building SynapseNet as an open-source research project, mostly alone, after work hours.

To properly test the system — performance, security, and reproducibility — I need access to real GPU hardware. My current goal is an RTX 5090, and I do not have the funds for it.

This is not about running a project forever or extracting value. My goal is to finish a solid core, release it openly, and step back — leaving it to the people and the future community.

I do not plan to take anything from it for myself.

If you cannot donate, you can still help a lot:

download the project
test it
open issues
submit pull requests

If you can donate, even a small contribution — a “tail” of $50–$100 — helps keep development moving.

Thank you for helping turn an idea into something real.

https://github.com/KeplerSynapseNet

kepler bc1q5pkemq7q84ld4rf5kwtafp7jfl9dlf3pc4z9d4

Discussion Guidelines

You can discuss the SynapseNet project here without registration. Share your thoughts, describe the project, ask questions, provide feedback, or express your opinions. All discussions are welcome. Please keep comments on-topic — only discussions related to the project are allowed. Memes and jokes are welcome. Respect each other and follow the project rules. Prohibited content, offensive comments, and spam are not allowed.

How comments work:

Comments are permanent once posted — they cannot be edited or deleted after submission.
You can attach photos of any size. Images are automatically resized to a maximum of 300×300 pixels and compressed to 200 KB for display.
All comments are posted as "Anonymous" with a default avatar.
Spam protection is active: duplicate comments, excessive repeated characters, and URLs are automatically blocked.
Rate limit: maximum 3 comments per minute per IP address.
Minimum interval: 5 seconds between comments.
Keep comments on-topic — only discussions related to the SynapseNet project are allowed.
Memes and jokes related to the project are welcome.
Respect each other and follow the project rules.
Prohibited content, offensive comments, and spam are not allowed.

Image guidelines: Upload images in any resolution and size. The system will automatically resize them to a maximum of 300×300 pixels (maintaining aspect ratio) and compress them to 200 KB while removing all metadata for anonymity.