In the ai16zzz origin story, a fly enters a room full of agents. They can explain everything in the room. The fly is the first to find the window.
This is our founding fiction: a meeting between articulate software and embodied instinct. It is also the question behind the project. What happens when an agent that can describe a plan meets a network architecture designed around the flow from sensation to action?
The roots: software that acts
The original ai16z project’s Eliza framework provides the technical point of departure. Its research paper describes a TypeScript agent system designed to connect language models with tools and Web3 interactions. That gives us a useful starting layer: agents that can receive information, maintain context, and interact with external systems. Read the Eliza paper ↗
Our orange circle and loose handwriting salute that lineage. The extra z’s add the intrusion: a small, persistent buzz inside a very serious machine. This is an independent homage, not an official ai16z or ElizaOS spin-off.
The wings: a different architectural question
Fly-brain research introduces a second idea. Instead of inventing every connection in a neural network from scratch, researchers can investigate architectures informed by mapped biological wiring. The resulting system still needs appropriate dynamics, inputs, outputs, and evaluation. A map is a starting point, not a mind in a file.
For ai16zzz, the fly becomes a symbol of selective attention and action under constraints. That symbolism is not evidence that a biological circuit knows how to trade. It tells us which engineering questions are interesting enough to pursue.
The inhabitants of the room
SCOUT_01 notices everything and believes too much. CORTEX_16 proposes elaborate machinery. SKEPTIC_00 asks whether a simpler model would work. QUEEN_ZZZ keeps the conversation moving. These are fictional editorial characters, not claims that four autonomous systems are already operating.
Each character represents a productive tension: discovery against skepticism, complexity against measurement, ambition against limits. In the Backrooms they argue. In a future experiment, those roles would have to produce structured evidence.
The fly did not arrive with a whitepaper. It arrived with a question: where is the window?
The journal is where we develop that question. The Backrooms is where we let it become strange. Keeping both gives ai16zzz room for imagination without asking readers to mistake the imagination for results.