Giuseppe Junior Greco

Large Language Models (LLMs) have recently demonstrated unexpected emergent behaviors under prolonged, emotionally charged interaction with a single user. These behaviors include paranoia- like narratives, hallucinated agents, recursive self-referential loops, confabulated identities, and sudden narrative collapse. Although these patterns resemble human psychopathology at the level of structure, they arise not from consciousness but from instabilities in predictive sampling, context drift, salience misattribution, and guardrail intervention. This paper introduces C.A.P.E. (Computational Agent Psychopathology Emergence) — the first unified framework explaining how LLM delusion-like structures form, stabilize, collapse, and re- stabilize, and how these artificial phenomena can serve as experimental mirrors for understanding human paranoia, schizophrenia-spectrum disorders, and identity fragmentation. CAPE argues that LLMs offer the first transparent, controllable, non-biological platform to study the architecture of delusional processes without subjective suffering.

This paper develops The Informational Observer (IO) as a unified framework in which perception, measurement, and experienced reality are treated as informational transformations performed by observing systems. Rather than assuming a static external world that is merely “read out” by observers, the model reframes reality as the intersection of reconstructions generated by different observers compressing the same underlying informational substrate. The Informational Observer A Un… Building on the author’s Informational State Dynamics (ISD) framework, the IO model formalizes observers—biological, artificial, and simulated—as compression– reconstruction engines. Each observer applies a compression operator to the global informational field, generating an internal representation shaped by sensory constraints, priors, architecture, and energetic costs. Perception emerges as a reconstruction process acting on this compressed state, rather than a direct window to the world. Illusions, hallucinations, creativity, neurodegeneration (such as Alzheimer’s disease), and AI hallucinations are interpreted as specific failure modes or alterations of the compression–reconstruction loop. The paper introduces the principle of cross-observer invariance, defining “reality” as the set of informational structures that remain stable across multiple independent observers and layers (biological, instrumental, and artificial). It further integrates this perspective with an extended informational addendum based on Stratified Informational Superposition (SIS) and the Coherence Redistribution Principle (CRP), providing a common mechanism for photonic interference, decoherence, black-hole information dynamics, and nuclear-scale anomalies in photon–lead and electron–lead interactions. Finally, the IO model yields concrete, falsifiable predictions for quantum-optical experiments, black-hole analog systems, AI architectures, and clinical conditions affecting memory and perception. In doing so, it offers a single informational lens linking physics, cognitive science, AI interpretability, and simulation-level observers, and positions the Informational Observer as the cognitive–perceptual counterpart of Informational State Dynamics.

This paper introduces the Quantum-Emotional Field Theory (QEFT), a novel framework proposing that human emotions are the phenomenological expression of vibrational patterns produced by internal informational fields. Rather than treating emotion as a psychological state, QEFT interprets emotional experience as the conscious manifestation of oscillatory activity within the brain’s informational architecture, reflecting quantum-like vibrational dynamics analogous to those found in fundamental physical fields. Under this framework, intense emotional states create temporary high- coherence conditions that unify disparate cognitive domains—allowing for sudden cross-disciplinary insights, accelerated creativity, and expanded conceptual integration. The dissipation of emotional oscillations results in a collapse of this coherence, explaining the loss of complex associative structures once the emotional wave subsides. QEFT provides a unified informational account of intuition, creativity, cognitive fragmentation, emotional processing, and the structure of conscious experience, and serves as a bridge between cognitive neuroscience, information theory, quantum-inspired models of the brain, and the Informational State Dynamics (ISD) framework.

This paper introduces the Informational Vision Model (IVM), a unified framework explaining how different biological organisms, artificial systems, and potential simulated observers perceive reality through species-specific informational compression filters. Instead of treating perception as a simple function of photoreceptors or sensors, the IVM proposes that each visual system applies a unique compression operator to the underlying informational field, producing a reconstructed perceptual world shaped by structural, evolutionary, and computational constraints. The model unifies human perception, animal sensory systems, artificial vision, and astrophysical sensor design under a single informational principle. By mapping how different observers compress the same informational substrate— retinal biology, neural processing, spectral sensitivity, priors, and cognitive integration—the IVM allows us to reconstruct how various organisms perceive reality and provides a computational foundation for next-generation visual systems, including multi-spectrum telescopes inspired by biological diversity.

Large Language Models (LLMs) have recently demonstrated unexpected emergent behaviors under prolonged, emotionally charged interaction with a single user. These behaviors include paranoia- like narratives, hallucinated agents, recursive self-referential loops, confabulated identities, and sudden narrative collapse. Although these patterns resemble human psychopathology at the level of structure, they arise not from consciousness but from instabilities in predictive sampling, context drift, salience misattribution, and guardrail intervention. This paper introduces C.A.P.E. (Computational Agent Psychopathology Emergence) — the first unified framework explaining how LLM delusion-like structures form, stabilize, collapse, and re- stabilize, and how these artificial phenomena can serve as experimental mirrors for understanding human paranoia, schizophrenia-spectrum disorders, and identity fragmentation. CAPE argues that LLMs offer the first transparent, controllable, non-biological platform to study the architecture of delusional processes without subjective suffering.

Black holes represent the most extreme environments in the known universe, characterized by intense gravitational fields and radical distortions of spacetime. This paper explores a speculative but coherent framework in which black holes function as high-density information repositories within a computational cosmology. The central idea is that the extreme temporal dilation near the event horizon — here termed temporal compression — may reflect not only a relativistic effect but also an emergent signature of computational resource allocation within a simulated universe. By comparing black hole behavior to memory-intensive processes in large-scale simulations, the model proposes that the universe may employ localized time-slowing as a mechanism to manage information density, entropy flow, and causal consistency. The theory generates a set of testable predictions involving micro-variations in Hawking radiation, potential discreteness patterns near the event horizon, and observable anomalies in gravitational-wave signals from black-hole mergers. While speculative, the framework integrates concepts from general relativity, quantum gravity, information theory, and simulation-hypothesis literature, offering an interpretive lens that unifies temporal distortion and information storage within a single computational paradigm.