Web4 Protocol · Genepool Ecosystem

GENERATIONNETWORK

Emotions select. Logic justifies. The system learns. The network grows.

Generation Network is the first Web4 platform that builds your digital identity, simulates cross-generational experience, and stores all data exclusively on your own device.

// GENEPOOL ECOSYSTEM
Web4 Ecosystem
Built On
▸
Product // 01

Generation Network

Genepool

A social platform focused on generational identity, built on synthetic cortex infrastructure. It runs fully locally with synthetic cortices, a 13-hormone neurotransmitter system and on-device ML models.

● You are here

01 — System

Artificial Emotions.
Real Decisions.

Generation Network is the first bridge network built on the idea of Synthetic Cortex — a bio-model system inspired by the neurological lobes of the human brain. When you examine the limbic system, you can see that hormones and neurotransmitters directly influence emotional processes. We modelled these mechanisms mathematically and embedded the resulting weights directly into LLM inference time.

The brain does not generate purely rational decisions — it filters possibilities through emotions to optimise itself and preserve creativity. Logic performs calculations, but it cannot choose. Choice emerges through the rationalisation of options already highlighted by the emotional system. Synthetic Cortex reproduces this computational process without simulating feelings.

As you progress through the network, every interaction is transformed into semantic and emotional representations and transferred into a digital brain created specifically for you. Over time the system develops a Reflection — a digital copy that thinks and decides like you.

3+
Years of R&D
Web4
Protocol Layer In Development
LOCAL
Local-First Data
100%
On-Device Storage

02 — Overview

Generation Simulation

Generation Network is a social network and simulation environment built on Synthetic Cortex — designed around a single idea: experiencing the perspectives of different generations, the private histories of individual lives, and the friction between generations, first-hand. Inside the network you can explore how earlier generations communicated — through letters, telegrams, SMS — uncover candid, first-person testimonies from real lives, and simulate all of it through your own digital double.

Once created, your Reflection can be connected to the network. Through it, you interact with other people — and their Reflections — across the entire ecosystem.

GN-CAP / 01
Acts On Your Behalf

Your Reflection is your digital double — it knows you and can act in your name. That includes gathering data relevant to you from across the internet and surfacing it directly in your feed.

GN-CAP / 02
Reads Other Reflections

It reaches out to other Reflections to learn about the people behind them and bring information back to you — showing you the entire process. It can design a simulation, invite another user's Reflection into it, and observe how that Reflection responds as events unfold — carrying real signal about how its user might actually react, since every Reflection is built from its own user's data.

GN-CAP / 03
Mediated Conversation

Your Reflection can speak directly with that person — or another Reflection can approach you to gather data for its own user. It's a new kind of mediated communication: a next-generation social network experience.

Your Device Is Actually a Simulation Environment

Generation Network is built on a local-first, hybrid architecture. Your personal content and your Reflection stay on your device; our cloud infrastructure only coordinates the network — identity, delivery, and matching — and holds only what you choose to share. Reflections carry out their processes on the devices they reach. When another user's Reflection reaches your device, for instance, its interactions actually run inside a simulation created on your own phone. The system even lets you hold certain Reflections on your device — you can remove one at any time you choose. When you open someone's profile and talk to their Reflection, you're effectively calling that Reflection into your own system; the conversation runs on your device (responses are generated through the AI provider), and its record stays on your device, not on our servers. Only the data you explicitly allow is ever sent to the network — everything else remains in your device's local memory.

Your Reflection's Emotions Aren't a Prompted Imitation

Your Reflection's emotions emerge from weights derived at runtime — from your decisions and from every point of interaction you have with the app: how long you stay, who your Reflection contacts, the conversations you have with other people. The more you interact with the app, the more data its emotional space produces (most of this emotional load comes from SNN networks), and these values directly shape the reasoning functions inside your Reflection's Synthetic Cortex. Through the app you can attach instinctive behavior patterns to specific emotional phases, lock certain abilities to particular conditions, or even make your Reflection deeply irrational — however you train it, that's what it becomes. Your Reflection doesn't feel emotions, but it experiences an analogue of the processes that occur in a brain that does, and it's programmed to decide based on the outputs of that process. Take the flinch you feel when a thorn pricks your hand: the Reflection detects the thorn, doesn't feel the pain, but experiences the flinch response that pain produces — and that experience pushes it toward a similar reaction. It doesn't have receptors yet, but it carries the same step at the mechanism level.

Built On Every Era Of The Web

WEB1

Read-only. The network preserves the classic portal-web experience — the system even runs its own generation newspaper at its core.

WEB2

Interactive. The social layer where users engage, respond, and shape the network together.

WEB3

Distributed. Peer-to-peer systems with no central authority.

WEB4 In Development

The layer Generation Network is building: a personalized world where Reflections live inside the internet itself, transfer weight between systems, and move freely in and out of them.

03 — Key Capabilities

What Synthetic Cortex Adds

SYN-CX / 01
Emotional Filtering

Emotions narrow the range of possible choices while reasoning provides the justification. Unlike other LLMs that simulate writing style, emotions here serve evolutionary and psychological functions — narrowing the decision space before logic operates.

SYN-CX / 02
Homeostatic Optimisation

An artificial hormone system maintains internal balance and optimises decision-making processes. Within this architecture, emotion and logic — two inherently noisy systems — operate together through Bayesian integration to produce more coherent, more rational outputs.

SYN-CX / 03
Associative Reasoning

Establishes semantic and associative relationships between pieces of information to generate new connections — the kind of lateral thinking that emotion-free reasoning systems systematically miss.

GEN-NET / 04
Reflections

Your Reflection is your digital counterpart — created from your data, thinking and deciding like you. It acts on your behalf across the network, interacts with other Reflections, and can give others a sense of who you are before you ever meet them directly.

GEN-NET / 05
Generation Simulation

Experience the lives of past generations firsthand — simulating their perspectives, decisions, and consequences from the inside. Send letters to people from another era, read their experiences, and build meaningful connections across generations.

WEB4 / 06
Privacy by Architecture

Your conversations and personal memory are stored encrypted on your device, and nothing on it is copied to our cloud. Your digital brain can be backed up, moved, or deleted entirely at your discretion. Privacy is built into the architecture — not just promised in a policy.

Promotional
Magazine
pmagazine.pdf
Open PDF

04 — Foresight

The Project's Most Important Pillar

Core Goal: Bringing the Internet to Web4.

There's no officially defined concept of "Web4" today. You may find people who've used the term before, but in most cases it referred to something different from what we mean by it. Here's what we mean:

Web1: The internet's earliest era — the internet spoke, and users mostly just read and absorbed information.
Web2: The internet started a conversation — users created content, shared it, and engaged with each other across social platforms.
Web3: The internet aimed to hand ownership back to users — giving them greater control over their digital assets, identities, and networks.

Today we're facing an entirely different industrial revolution: Industry 4.0 — the stage where AI systems enter the industrial domain itself. Our project is building a protocol that brings this same mission down to the level of the entire internet: a wholly new internet experience. We're developing the structure, standards, and protocols for a web world in which the entire internet is reduced to individuals' personal digital Reflections, and web servers themselves are built around those Reflections. We're constructing an entirely different dimension of interaction.

"Logic cannot produce decisions without the value function created by emotions. The brain does not process all available information equally — it relies on emotional filtering and parallel processing."

About the Developer →

05 — Team

About Our Project Team

Generation Network is built by one human developer working alongside a team of AI agents, each with a defined role.

ALL Teams →


Under the hood

How Do Team Work?
How the Project Operates

Structure 1 · API

Connected via API

HUMAN SYNTHETICCORTEXexternal upper layer API Claude ChatGPT Gemini DeepSeek

Agents in this structure talk to large frontier models through APIs. Synthetic Cortex is implemented here as an external upper layer on top of the model. This structure is used mostly for work that demands deeper technical expertise.

API providers

ClaudeChatGPTGeminiDeepSeekGrokMistralKimiQwen

Structure 2 · Open

Open-Source, Local

HUMAN LOCAL BOUNDARY SYNTHETIC CORTEXactivation steering OPEN MODELruns locally LOCAL DATA ✓ ✓ ✓

Agents here run on Synthetic Cortex technology integrated into open-source models. The external brain layers act directly on the model through activation steering. Because the data matters more, human oversight is higher than in Structure 1: every processing phase is reviewed by a human (the ✓ checkpoints above).

Open-source models handle the project's own information and any data that must stay local. Nothing in this flow needs to leave the project's own machines.

Open-source models

KimiDeepSeekLlamaGemmaQwenMistralGLMgpt-ossPhiFalcon

Note — This is not a fixed setup. Models change often depending on need, fine-tuned models may be added, and sometimes a human takes the job over directly.

06 — Download

Join the Genepool

Download from your preferred store. Your data lives only on your device — always.

Primary Stores
Coming Soon
// Direct APK Download — Android Sideload