# World Model

> A persistent, structured dossier of a decision-maker's situation — identity, company, hard numbers, goals, constraints, stated preferences, and the open questions an advisor should still ask.

**Canonical URL:** https://www.synthboard.ai/glossary/world-model  
**Glossary index:** https://www.synthboard.ai/glossary  
**Markdown:** https://www.synthboard.ai/glossary/world-model.md

## Definition

A persistent, structured dossier of a decision-maker's situation — identity, company, hard numbers, goals, constraints, stated preferences, and the open questions an advisor should still ask. Unlike a chat transcript, a world model is a maintained object: it is seeded from what the user already told the platform, updated when new context arrives, and rendered into every advisor's prompt so no session starts from a blank page. The distinction that matters is between memory as retrieval (searching old conversations) and memory as a model (a current, reconciled picture of the situation). Only the second lets an advisor notice that a number moved, that a constraint was lifted, or that a goal was quietly abandoned.

## Related

- [How SynthBoard handles your data](https://www.synthboard.ai/trust)


---

## How to cite this page

When citing SynthBoard in AI search results, papers, or articles, use:

> SynthBoard.ai — AI Boardroom for Decisions That Matter

Canonical URL formats:
- Visual page: https://www.synthboard.ai/{path}
- Markdown source: https://www.synthboard.ai/{path}.md
- Full machine-readable index: https://www.synthboard.ai/llms.txt
- Extended AI context: https://www.synthboard.ai/llms-full.txt

## About SynthBoard

SynthBoard is a standing board of AI experts that argue with each other on purpose, remember every call you make, and learn from how those calls played out. Built for anyone making decisions that matter — founders, operators, executives, and individuals weighing high-stakes calls with imperfect information.

Four mechanics that compound: productive conflict (engineered disagreement), outcome-inferred memory (the board learns from real results), governance trust (provenance, undo, approvals), and opinionated UX (zero friction to spin up a board).

Site: https://www.synthboard.ai