724 Research

← Back to Concepts

CONCEPT

Context OS β€” AI's World Model for Understanding a Person

Context OS is the persistent world model AI uses to understand a person over time. It is built from six context dimensions and bounded by a data veracity hierarchy and a set of privacy principles. It answers the question of who this person is and what is happening now, forming a complementary pair with Goal OS, which answers where the person is headed.

Concept
Article info
v0.1
Published July 18, 2026
~2 min read
Contents

Definition#

Context OS is the persistent world model AI uses to understand a person over time.

Its premise is:

Every useful response is generated from context, not memory alone.

Memory records what happened.

Context explains:

  • who this person is;
  • what they are trying to achieve;
  • what environment they are in;
  • who matters to them;
  • what capabilities and constraints exist;
  • what state they are currently in.

The Six Context Dimensions#

Identity#

  • values;
  • beliefs;
  • preferences;
  • personal history;
  • self-definition;
  • boundaries.

Goal#

  • long-term Goals;
  • current Goals;
  • priorities;
  • conflicts;
  • Desired Future State.

Environment#

  • country;
  • market;
  • institutional environment;
  • family situation;
  • economic situation;
  • legal and cultural background.

Relationship#

  • family;
  • partners;
  • clients;
  • colleagues;
  • investors;
  • trusted experts;
  • conflicts and obligations.

Capability#

  • skills;
  • experience;
  • reputation;
  • assets;
  • capital;
  • network;
  • tools and Agents.

State#

  • ongoing projects;
  • financial state;
  • health and energy;
  • emotional state;
  • deadlines;
  • looming risks.

Context Data Types#

Context OS may include:

  • Identity;
  • Goals;
  • Active Programs;
  • Timeline;
  • Relationships;
  • Organizations;
  • Knowledge;
  • Opportunities;
  • Constraints;
  • Decisions;
  • Evidence;
  • Events;
  • Patterns;
  • Current State.

Context Is Not Surveillance#

Context OS should not attempt to manipulate a person, nor should it attempt to fully predict one.

Its purpose is to:

  • increase relevance;
  • maintain continuity;
  • surface contradictions;
  • support better decisions;
  • help calibrate Goals;
  • help a person understand themselves.

The north star remains:

Know yourself β†’ become yourself β†’ realize yourself

The Data Veracity Hierarchy#

A trustworthy Context OS should distinguish between the following levels:

1. Raw Input
2. Evidence
3. Proposed Structure
4. Approved Context
5. Reasoning Output
6. Decision
7. Outcome and Learning

AI-generated interpretations must not override facts the user has already confirmed without the user's knowledge.

Privacy Principles#

  • private by default;
  • minimum necessary data;
  • clear user approval;
  • source traceability;
  • exportability;
  • correction and reversal;
  • separation of fact and inference;
  • no external commitment without explicit approval.

Relationship to Goal OS#

Context OS answers:

Who is this person? What is happening now?

Goal OS answers:

What future does this person want? What should happen next?

Together:

Context OS understands the person. Goal OS helps the person move toward a future.

Continue Reading

Open questions raised by this article

  • When an AI's inference conflicts with context the user has already confirmed, who holds final authority to decide?

View all open questions β†’