2-Minute Overview

Web4 in 2 Minutes

Plain English first, so you can read this cold — every term is spelled out the first time it appears. The two you'll lean on most are ATP and LCT; a few more (T3, CI, MRH) appear briefly below, each glossed in place, with deeper shorthand explained on their own pages.

The Problem

The internet has no memory. A spammer can create a new account in seconds. A scammer banned from one platform signs up on another. Your 10-year reputation on Twitter means nothing on Reddit. Bad actors never face compounding consequences — they just reset.

The Idea

What if every online action had a cost, every identity was tied to physical hardware, and your reputation followed you everywhere — permanently?

Actions cost energy — valuable contributions earn it back. Posting, messaging, voting — everything draws from a personal energy budget (called ATP — short for Allocation Transfer Packets, a name borrowed from cellular biology, where ATP is the molecule your cells spend to do work), but the budget refills continuously when others confirm your work was useful, so contributors don't get locked out. Each action also creates a small receipt of what you spent. Spam drains the budget with no return.

Identity is anchored in your devices. Think of it as a digital ID card that lives in your device's security chip — not a password you type or a key file you store, but a presence your hardware proves. Your phone, laptop, and tablet each carry such chips, and each one vouches for the others, so no single device can speak for you alone (the constellation is called a Linked Context Token, or LCT). More devices strengthen it. Creating a fake identity means buying new hardware — expensive and slow.

Trust is earned, never bought. Your reputation is multi-dimensional (not just a single score), portable across platforms, and permanent. Good behavior compounds. Bad behavior follows you.

Humans and AI play by the same rules. The same trust framework applies to people and AI alike. Web4 calls all participants “agents” — human or artificial — because the rules are identical. Everyone earns their place.

Three more ideas round out the picture — your trust dimensions, your trust neighborhood, and whether your behavior holds together over time — each on its own page. Tap for the shorthand (T3, MRH, CI) and links, or meet them when you get there.

T3 (Talent / Training / Temperament) — the three dimensions of trust, scored per role; MRH (Markov Relevancy Horizon) — your trust neighborhood, distance-adjusted reach, 3 hops max; and CI (Coherence Index) — does your behavior add up over time. The concept sequence on each page links the rest.

The Result

Spam dies because it's too expensive. Scammers can't escape their history. Trusted contributors accumulate real, portable reputation. And the whole thing runs without any central authority deciding who's trustworthy.

Wondering what it would actually look like? See conceptual UI mockups →

Where is this now?

Web4 is active research with working simulations, not a deployed product. One distinction worth keeping straight: Web4 is the protocol — the rules and math; 4-Life (this site) is the lab where you explore those rules through simulations, walkthroughs, and explainers. So 4-Life lets you watch the mechanics in action — see societies form, trust networks emerge, and bad actors fail. The underlying protocol specification is open.

What exists today:

Formal protocol specification (open source)
Open-source reference implementations (LCT, T3/V3, ATP, witnesses)
Working trust simulations you can run here
This 77-page interactive explainer site

What comes next:

Pilot deployment with a small real community
Platform integration tests (Tier 1 wrapper — adding trust scores to existing platforms like Reddit or Gmail)
Independent review of trust mechanics
Protocol finalization based on real-world data

No timeline commitments — this is research, and research takes as long as it takes. The simulations work. The question is whether the real world behaves like the simulations.