Algorithmic Imagination Engine: A Conceptual Framework for Creative Computation Beyond Pattern Imitation

This project introduces the Algorithmic Imagination Engine (AIE), a conceptual and theoretical framework that explores how computational systems might move beyond pattern imitation toward structured, imaginative idea generation. Unlike conventional AI models that rely primarily on historical data and optimization, this work examines imagination as an algorithmic process involving abstraction, recombination, constraint-breaking, and internal simulation. The paper proposes a multi-layered architecture for algorithmic imagination, integrating memory abstraction, probabilistic divergence, contextual synthesis, and evaluative feedback loops. Rather than presenting a finished implementation, the focus is on defining imagination as a computable process and outlining how such systems could support creativity in scientific discovery, problem-solving, and human–AI collaboration. This research is intended as an open, exploratory contribution to computational creativity, cognitive systems, and theoretical AI. It is published as a preprint to encourage discussion, critique, and further development by researchers, students, and interdisciplinary thinkers. Keywords: Algorithmic imagination, computational creativity, theoretical AI, cognitive systems, creative computation, human–AI collaboration