AI Knowledge Creation Architecture: Toward a Next-Generation SECI Model

Knowledge Activation is the process by which an AI initiates from an input prompt, retrieves relevant knowledge from external sources such as the BoK (Body of Knowledge) and the Retrieval Knowledge Base, as well as from its internal Pretrained Parametric Knowledge, and forms a coherent Context of Reasoning. At this stage, the AI reconstructs static knowledge into a “mobilized semantic space,” performing the core process of preparing for generation and understanding. Knowledge Activation precedes Knowledge Assimilation, functioning as the trigger that dynamically incorporates external knowledge into the AI’s internal state.

At SimpleModeling, the core knowledge system for contextual sharing is called the BoK (Body of Knowledge). The goal of building a BoK is to enable knowledge sharing, education, AI support, automation, and decision-making assistance.

Tacit knowledge implicitly retained within the AI. It exists as weights, internal representations, and reasoning patterns; it cannot be directly externalized, but manifests through generated output.

A process in which knowledge expands spirally through repeated mutual transformation between tacit and explicit knowledge. It is treated as a core concept of the SECI model.

An architecture that organizes the internal processes executed by generative AI: knowledge activation, assimilation, expression, promotion, and circulation.

A temporary internalization of external knowledge retrieved from the BoK through RAG. It enhances reasoning within a session but does not permanently modify the AI model’s parametric knowledge.

The process of producing externalizable outputs—such as text, code, models, or structured data—based on assimilated knowledge.

A long-term integration process in which structured knowledge from the BoK is permanently incorporated into the model’s Pretrained Parametric Knowledge (PPK) through retraining or fine-tuning.

Knowledge Circulation denotes the cyclical process in which knowledge is activated, assimilated, expressed, and promoted between AI and humans, then re-activated. Through this loop, knowledge generated by AI is organized and reintegrated into the SmartDox site or BoK (Body of Knowledge), allowing the whole system to evolve and deepen AI’s understanding and generative ability.

Retrieval Knowledge Base (RKB) is a structured, retrievable subset of Non-parametric Knowledge optimized for use by RAG (Retrieval-Augmented Generation). It contains indexed SmartDox documents, glossary entries, and semantic metadata, allowing AI models to fetch explicit knowledge as contextual input. Through RAG interaction, the RKB serves as a bridge for transforming external explicit knowledge into intermediate assimilated knowledge within the AI model.

Knowledge embedded within a generative AI model’s parameters, obtained through large-scale pretraining. It represents general linguistic, conceptual, and procedural understanding that the model can apply without explicit external input.

Parametric knowledge refers to implicit knowledge embedded in the parameters (weights) of a neural network. It represents statistical or distributed information acquired during pretraining, rather than explicit facts stored in an external knowledge base. In RAG (Retrieval-Augmented Generation), it is contrasted with non-parametric knowledge, serving as the model’s internal “implicit value.”

A generation technique that supplements a language model’s internal (parametric) knowledge by retrieving relevant external information before generation. RAG systems first search knowledge sources such as databases or knowledge graphs and then use the retrieved context as input for text generation.

A structured instruction or contextual representation that bridges retrieved knowledge (RAG) and the AI model’s reasoning process. It transforms the structured knowledge from the BoK into a narrative or directive form that the model can interpret, act upon, and internalize.

A semantically integrated knowledge state within the AI, constructed from the context produced by knowledge activation. Corresponds to “understanding” within the AI.

Artifacts produced through the knowledge-expression process of generative AI, including documents, code, summaries, and structured data.

The process of extracting valuable knowledge from generated output and “returning” it to the BoK. A cyclical operation that reconstructs externalized knowledge as explicit knowledge.

A knowledge-creation architecture structured around AI tacit knowledge and the BoK, organizing the loops of activation, assimilation, expression, promotion, and reflux.

A Literate Model is a “readable model” that integrates model structure with natural-language narrative (structured documentation). It extends the idea of literate programming into the modeling domain, unifying structure (model) and narrative (structured text) into a single intelligible artifact interpretable by both humans and AI. The concept of “Literate Modeling” has been explored previously by some researchers and developers, mostly as an approach to improve documentation or code comprehension. However, those attempts did not establish a systematic modeling methodology that integrates models, narrative, and AI assistance as a unified framework. The Literate Model is a modeling concept newly systematized and proposed by SimpleModeling for the AI era. Building upon the ideas of literate modeling, it redefines them as an intelligent modeling foundation that enables AI-collaborative knowledge circulation and model generation. It is not merely a modeling technique but a framework that embeds human reasoning and design intent as narrative within the model, enabling AI to analyze and reconstruct them to assist in design and generation.

A model in which knowledge emerges through the spiral repetition of mutual transformation between tacit and explicit knowledge.

The SECI phase in which tacit knowledge is shared and a shared experience is formed.

The SECI phase in which tacit knowledge is externalized into explicit knowledge such as language, models, and diagrams.

The SECI phase in which externalized explicit knowledge is internalized as tacit knowledge.

A concrete action or task performed within an Activity Space to move an Alpha to a more advanced state, typically producing or refining Work Products.

The phase in which multiple pieces of explicit knowledge are integrated and reorganized into a more systematized body of knowledge.

A cyclical knowledge-updating process in the AI era, in which documents (SmartDox), models (CML/SimpleModel), vocabularies (Glossary/Category), and design knowledge (BoK) are integrated through RDF, enabling AI to reference, generate, and improve knowledge, which is then reflected back into documentation.