Curator Provenance and Licensing Notice

This document predates its open licensing.
As curator and author, I apply the Apache License, Version 2.0, at publication to permit reuse and implementation while preventing enclosure or patent capture. This licensing action does not revise, reinterpret, or supersede any normative content herein.
Authority remains explicitly human; no implementation, system, or platform may assert epistemic authority by virtue of this license.
(2025-12-18T08:07Z version 1.0 - See the Main Page)

Cognitive Memoisation: Plain-Language Summary (For Non-Technical Readers)

Abstract

This work describes a new way of working with artificial intelligence tools such as ChatGPT that recognises an important limitation by design: these systems do not remember past conversations and must not be treated as authoritative. Instead of trying to make the AI “remember”, the approach extracts stable knowledge, decisions, and definitions into human-controlled documents that can be reused as the starting point for the next round of thinking and exploration. This preserves progress without giving the AI memory or authority, keeping the human firmly in control. The method is novel in that it turns a deliberate limitation of large language models into a strength, enabling safe, repeatable growth of understanding across many separate working sessions.

Explanation (In Everyday Terms)

Ralph’s first two papers are about avoiding a very familiar frustration. Anyone who has seen the movie *Groundhog Day* understands the feeling of waking up and having to relive the same day over and over again. When working with AI tools, the same thing often happens: you explain something carefully, make progress, come back later, and discover you have to explain it all again from scratch.

Rather than fighting this, Ralph treats it as a design reality. The AI does not remember—and should not remember. Instead, whenever something important is discovered, agreed, or clarified, it is written down in a separate, human-owned document. That document then becomes the reference point next time, so work continues from where it left off, without pretending the AI has memory.

Another key idea is recognising how people naturally think when exploring new ideas. Ralph often uses what he calls “dangling cognates”: words or concepts that are partly formed and not fully defined yet. This is not a mistake—it is how real thinking happens. The problem only arises when these half-formed ideas are accidentally treated as settled facts. The solution is not to eliminate this kind of thinking, but to clearly distinguish between ideas that are still evolving and ideas that are grounded and stable.

In simple terms, the papers show how to:

Stop repeating the same explanations over and over
Keep control of knowledge in human hands
Let ideas stay flexible while they are forming
Lock things down only when they are truly decided

The result is steady progress instead of endless repetition, without turning AI into something it is not. The AI remains a tool; the human remains responsible.

— Authored by Ralph B. Holland employing inference, insights and verification by chatGPT, under Ralph's guidance following the Cognitive Memorisation methodology.