Fundamental Universal Learning Patterns: A Framework for Identifying Learning Patterns in All Living Organisms and a Consideration of Artificial Intelligence Training

This paper introduces Fundamental Universal Learning Patterns (FULPs), a framework for modeling learning as a substrate-independent developmental process grounded in interaction rather than pre-existing data or fixed reward structures. Intelligence is conceptualized as emerging through a staged progression from minimal signal conditions, here defined as the void, to structured internal representations, in contrast to contemporary machine learning approaches that rely on curated datasets or externally defined objectives. FULPs identifies eight sequential learning patterns observed across biological systems, from single-celled organisms to humans, with FULP Five, curiosity recursion, playing a central role in driving self-directed exploration and knowledge accumulation. Unlike prior frameworks, FULPs does not assume pre-existing structure or human-centric reward functions; instead, it provides a generative developmental sequence in which systems acquire structure progressively through interaction. Optimization and probabilistic inference are embedded as local mechanisms rather than primary drivers of learning. Ethical considerations are incorporated at the architectural level, treating the potential emergence of morally relevant states as a design constraint. FULPs thus provides both a theoretical lens for understanding learning across biological and artificial systems and a foundation for empirical and theoretical research into developmental approaches to artificial general intelligence.