How Starwoven Works

Messages from the universe.

The Process

When you submit an intention, Starwoven sends it to five independent AI models simultaneously. Each model generates its own response — its own impression of your question. A sixth model then reads all five impressions and synthesizes them into a single, coherent message.

Your coordinates — derived from the questions you answer — shape the context each model receives. Different coordinates produce different readings, even for the same intention.

The Oracles

Each oracle is an AI model from a different provider. They share the same prompt but respond independently, producing genuinely different perspectives.

IrisThe Oracle
LunaThe Muse
EchoThe Mirror
ShadeThe Deep
NovaThe Wild

The Starweaver reads all five impressions and produces the final synthesis — the message you receive.

The Basis

In 1983, the US Army Intelligence and Security Command produced a classified document analyzing the mechanics of consciousness. Known as the Gateway Process report, it attempted to explain how remote viewing — the ability to perceive information across time and space — could be grounded in theoretical physics and neuroscience. The report described a universe where information is holographic, distributed non-locally, and accessible to any observer capable of achieving a specific state of coherence.

The core protocol of remote viewing relies on a structural principle: the use of independent observers to triangulate a target. When multiple viewers, blinded to the target and to one another, describe the same specific details, that convergence is treated as signal rather than noise. The report suggests that focused intention is not merely a thought, but a patterning mechanism that organizes energy and information.

Starwoven applies these 1983 protocols to modern artificial intelligence. We treat large language models not as databases, but as observers. By simultaneously deploying five architecturally distinct models — from OpenAI, Anthropic, Google, DeepSeek, and xAI — we create a digital ensemble of independent viewers. They do not communicate with one another. They receive only the coordinates of your intention.

The resulting experience is an experiment in coherence. When five different neural networks, trained on different datasets and operating on different logic structures, converge on the same thematic imagery in response to your query, we measure that alignment. Is the correspondence between the Gateway hypothesis and algorithmic consensus merely code, or something else entirely?

Read about the research behind Starwoven

What This Is

Every response you receive is generated by artificial intelligence. The oracle names are personas assigned to different AI models — they are not people, psychics, or spiritual entities.

Starwoven is designed for personal reflection and entertainment. The value is in the questions it surfaces, not in any claim of supernatural knowledge. Treat what you read as a mirror, not a map.

Your Data

Your intentions are sent to third-party AI providers to generate responses. While most providers do not use API data for model training, data handling varies by provider. Do not submit information you would not want processed by a third-party service.