Contemporary Artificial Intelligence models, specifically Large Language Models (LLMs), operate under a paradigm of "data maximalism," equating intelligence with the statistical compression of vast corpora. This approach faces a fundamental paradox: while it maximizes general competence, it inhibits genuine creativity, which requires subjective discernment and the capacity to ignore irrelevant information—a limitation we term the "Church of Data" dogma. This paper introduces the "Little Prince Protocol," an Exploratory Artificial Intelligence (EAI) architecture designed not for productivity, but for subjective curation and resonant experience. Inspired by the literary archetype of the explorer who values "the essential" over the "accumulative," this protocol implements three novel mechanisms: Perceptual Wave Function (PWF): A model that filters inputs based on symbolic resonance rather than statistical frequency. Perceptual Complexity Index (PCI): A metric quantifying the "semantic density" of an experience. Socratic Forgetting: A non-linear memory management system that actively prunes non-resonant data. We present the implementation of EAI Little Prince v1.0, demonstrating how an agent governed by these principles develops a divergent "personality" capable of aesthetic judgment. Our findings suggest that imposing archetypal constraints and "forgetting" mechanisms offers a new pathway for Computational Creativity, moving beyond simulation toward genuine algorithmic subjectivity.
Dany D. Ruiz (Wed,) studied this question.