How to Read This Paper

📖 Who This Is For: Cognitive psychologists, HCI researchers, AI system designers, and anyone interested in how humans manage knowledge across digital and analog systems.

🆕 What's New: A unified three-domain framework that explains how humans strategically navigate internal memory, accessible ecosystems, and unknown territories—with practical design patterns for AI systems that complement rather than replace human cognition.

💬 Feedback Sought:

• Empirical validation of the entropy-creativity framework

• Additional testable hypotheses for the three-domain model

• Real-world applications of the design patterns

• Cross-cultural validity of the framework

Abstract

It is widely accepted that human memory operates through multiple systems (Miller, 1956; Schacter, 2001), that collaborative groups develop transactive memory where individuals specialize in different knowledge domains (Wegner, 1987), and that digital tools have fundamentally altered how we encode and retrieve information (Sparrow et al., 2011; Hutchins, 1995). This paper synthesizes over 50 peer-reviewed studies across cognitive psychology, human-computer interaction, and information science to develop a unified framework for understanding how humans manage knowledge across three interacting systems: (i) Displaced Internal Knowledge (I) - everything stored within individual cognitive systems; (ii) Undisplaced Proximal Knowledge (P) - trusted colleagues, familiar systems, and accessible ecosystems; and (iii) Undisplaced Non-Proximal Frontier Knowledge (F) - unfamiliar territories requiring active discovery.

Drawing on empirical evidence from memory research, transactive memory studies, and digital cognition experiments, this synthesis demonstrates how metacognitive confidence (Nelson & Narens, 1990), task demands, and resource availability govern dynamic shifts among these domains. We reinterpret "forgetting" as adaptive curation rather than system failure (building on Schacter, 2001), explain when collaborative remembering outperforms individuals (extending Rajaram & Pereira-Pasarin, 2010), and outline why unfamiliar territories amplify both discovery potential and misinformation risk (Kruger & Dunning, 1999).

The analysis expands by leveraging ancient frameworks such as the Buddhist pipeline for information processing—contact, feeling, perception, and mental formation—offering potential pathways for understanding how humans "think" and how AI systems might operate in the future.

Building on neurosymbolic AI architectures and entropy theory (cognitive), the framework reveals that creativity emerges from surprising encounters between knowledge domains—when familiar internal expertise collides with trusted collaborative knowledge and unexpected external discoveries. Maximum innovation occurs at these intersection points, where the "surprise gradient" is optimally balanced: enough novelty to spark new connections, but not so much as to cause cognitive overload. This insight bridges information theory with practical design, showing why breakthrough ideas rarely emerge from a single knowledge source but rather from the unexpected synthesis across all three systems.

The analysis yields five actionable design patterns for AI knowledge systems: 

(1) confidence-gated help—AI intervenes only when metacognitive monitoring signals uncertainty; 

(2) desirable-difficulty ladders—progressive challenges that strengthen memory through optimal surprise levels; 

(3) transparent epistemic status—clearly distinguishing what the system knows from what it infers; 

(4) cross-domain bridges—connecting siloed knowledge to enable surprising discoveries; and (5) misinformation triage—protecting users from false information in unfamiliar territories. Each pattern is operationalized with testable hypotheses and study designs, transforming theoretical insights into practical AI systems that amplify human cognition.

Keywords: cognitive psychology, knowledge management, transactive memory, metacognition, human-computer interaction, distributed cognition, memory systems, neurosymbolic AI, concept graphs

Methodological Note: This framework synthesizes over 50 peer-reviewed studies from cognitive psychology, human-computer interaction, neuroscience, and information science, integrating empirical findings from memory research (Miller, 1956; Schacter, 2001), transactive memory studies (Wegner, 1987; Rajaram & Pereira-Pasarin, 2010), digital cognition and distributed cognition work (Sparrow et al., 2011; Hutchins, 1995), and personal information management research (Barreau & Nardi, 1995; Kwasnik, 1991).

Highlights / Contributions

• Three-Domain Framework: I propose a unified model of human knowledge management across: 

• Displaced / Displaced Internal Knowledge (I), 

• Undisplaced Undisplaced Proximal Knowledge (P), and 

• Undisplaced Non-Proximal Frontier Knowledge (F)

• Entropy-Creativity Connection: Demonstrates how optimal entropy differentials between knowledge domains drive creative insight and innovation

• Strategic Forgetting: Explains how apparent memory "failures" represent sophisticated cognitive optimization rather than system breakdowns

• Collaborative Cognition: Shows how accessible ecosystems extend cognitive capabilities beyond individual limits through transactive memory and familiar navigation patterns

• Neurosymbolic Architecture: Argues for concept-centric AI systems that mirror human cognitive organization through neural-symbolic hybrid processing

• Design Pattern Library: Derives five practical design principles for AI knowledge systems that amplify rather than replace human cognitive capabilities