Open Knowledge Format
Open Knowledge Format (OKF) is a proposed standard for structuring and managing personal knowledge bases, specifically targeting the interoperability of large-language-model (LLM)-generated wikis. It addresses the fragmentation of individual, unstructured “LLM Wikis” by establishing a unified schema for data exchange and integration.
Core Concepts
- Standardization of LLM Wikis: Moves beyond isolated, proprietary note structures to a common format that allows different LLM agents or tools to read, write, and update knowledge bases seamlessly.
- Interoperability: Enables cross-platform compatibility for personal knowledge management systems, reducing vendor lock-in and data silos.
- Structured Evolution: Represents a shift from ad-hoc note-taking to a systematic approach.
Context and Origins
- Evolution from Karpathy’s LLM Wiki: OKF builds upon Andrej Karpathy’s initial concept of the “LLM Wiki,” transitioning from a theoretical framework to a standardized implementation.
- Google’s Proposal: Google has proposed OKF as the definitive standard for this ecosystem, aiming to unify disparate agent knowledge structures.
- Recent Developments: See Google’s OKF: Standardizing Karpathy’s LLM Wiki for AI Interoperability for detailed analysis of the standardization effort.