The new protocol is now the Network State Protocol, accepting a social smart contract, minting an NFT for identity, and enforcing an AI‑driven code of conduct with proof‑of‑personhood coming.
Graph fully on‑chain; wallet is your identity, no registry, no signup.
Posts are signed by you, stored outside any single app, and anyone can index them.
Metropolis got 880 submissions!
I decided to adopt a modular feed and moderation system so users can swap components, preventing lock‑in and acquisition risk.
In early October 2026 I posted a rapid stream of X updates that mixed product ideas, market observations and community design notes. The thread shifted toward on‑chain social critique, mentioning a 17‑person TG group, a “fid close to 1200”, and a range of quantitative observations.
I decided to store resolution‑condition logic in the app repository and enforce specificity with a custom LLM‑backed API, preventing vague market creation.
I proposed a peer‑to‑peer prediction market that lets users bet via Claude‑to‑Claude through an MCP server, describing it as “not hard”.
I warned founders to reject angels who don’t understand the model, noting that angel investing “doesn’t make money… bragging rights… lost money”.
I outlined a meritocratic community design with leader judgment, an arbiter role, layer‑2 culture and visible penalties for gaming.
I noted that the “fid close to 1200” metric was observed in my post.
I quoted myself: “this should not be hard”.
I said: “go where the users are. tell me if you are building this”.
I observed: “illusion of progress is the biggest enemy of progress”.
I added: “only hard constraints expose this”.
I joked about roles: “highest paid role: shock absorber”.
Quantitative observations about Jalebi Fafda sales and vendor revenue were shared, but I did not claim those numbers as my own statements, so they remain as contextual notes rather than asserted facts.