Raman Marozau

CTO & Founder of Target Insight Function

Tech Innovator & Paradigm Creator

Skip to main content

Publications

Articles & Research

Scientific publications and research papers.

Research Areas

AGI & AI Optimisation

Distributed AI/multi-agent architectures with explicit constraints on safety, interpretability and failure modes. Prototype systems routing work between agents, measuring coordination patterns against solution quality and latency.

Quantum Decoherence & Entanglement

Mathematical models of information propagation and stabilisation in closed quantum-like systems. Internal decoherence, effective stress–energy components and their impact on early-universe–style dynamics inspiring robust AI architectures.

LLM Adaptation & AI Governance

Experimental benchmarks of tensor layouts, numeric precision (FP32/FP16/BF16) and custom kernel implementations. Open Data Ownership License concept with governance patterns tying AI model usage to explicit data-ownership and accountability rules.

July 24,

Operational AGI: A Language-Based Approach for Adaptive Multi-Agent Systems

This paper presents a multi-agent, NLP-driven task coordination framework that constitutes a practical architecture for Artificial General Intelligence (AGI). It leverages real-time agent-driven analysis, language-based task structuring, and LLM-powered iterative optimization to enhance adaptability in dynamic environments. Traditional static evaluation methods in multi-agent systems struggle with evolving task constraints and interdependencies, often leading to inefficiencies in decision-making and workload distribution. While existing multi-agent architectures have introduced task-sharing mechanisms, they remain rigid in dynamic execution scenarios, lacking real-time adaptation to agent constraints and evolving system states. To address these challenges, we introduce a hierarchical task allocation model that integrates a Centralized Control Unit (CCU) for system-wide orchestration, an NLP-based Dynamic Task Allocation (NLP DTA) module for context-aware task structuring, and distributed agents that iteratively refine execution through structured feedback loops. Unlike existing multi-agent coordination models, which rely on preprogrammed heuristics or static task allocation, our AGI framework enables agents to first self-assess task feasibility, allowing NLP DTA to construct an optimal execution plan. This enables iterative task reassignment, where LLM-driven analysis dynamically adjusts agent responsibilities based on real-time execution feedback. The proposed AGI architecture demonstrates strong scalability and adaptability, making it suitable for AI-driven applications such as autonomous cloud resource allocation, workflow optimization, and decentralized automation. By replacing rigid task evaluation mechanisms with an adaptive, context-aware approach, this work enhances adaptive coordination in multi-agent systems, providing a scalable and computationally efficient path toward embodied AGI in dynamic AI environments. Our findings highlight new directions for optimizing real-time task evaluation in language-based AI models, distributed systems, and AGI-driven decision support.

Operational AGI: A Language-Based Approach for Adaptive Multi-Agent Systems
https://doi.org/10.36227/techrxiv.173611616.64880965/v6
May 19,

Gravity as Quantum Entanglement Resistance: A New Paradigm for Mass and Gravitational Interaction.

We propose a new theoretical framework in which gravitational attraction is not a direct consequence of mass, but a manifestation of quantum entanglement resistance. In this model, mass is reinterpreted as the density of internal quantum entanglement, and gravitational interaction emerges from the increasing energetic cost of forming external entanglement connections with systems that are already internally saturated. This paradigm offers an alternative to the classical geometric approach of General Relativity and instead treats gravity as an emergent informational phenomenon rooted in the structure of quantum correlations. We analyze how this model aligns with existing theoretical frameworks such as ER=EPR, entropic gravity, and holographic dualities, and demonstrate its originality through unique conceptual structures not found in existing literature.

Gravity as Quantum Entanglement Resistance: A New Paradigm for Mass and Gravitational Interaction.
https://doi.org/10.36227/techrxiv.174352106.68765747/v2
May 19,

Resolving the Three-Body Problem: A Probabilistic Quantum-Informational and Topological Approach.

The Three-Body Problem is a foundational challenge in physics and mathematics, known for its analytic intractability and chaotic sensitivity to initial conditions. This manuscript presents a rigorous, physically justified resolution by unifying classical, quantum, and informational paradigms. We demonstrate that, rather than admitting a unique deterministic solution, the three-body system must be described by a probability landscape of potential scenarios, consistent with modern quantum mechanics, chaos theory, and the Theory of Essence [1]-a conceptual framework treating reality as a field of potentials. By leveraging quantum-classical correspondence, Wigner phase-space methods, and recent advances in quantum simulation, we show that observable dynamics emerge as probabilistic slices or "collapses" of this broader landscape, with direct implications for astronomy, astrophysics, and multi-body system engineering. Experimental evidence and numerical simulations confirm that all measured outcomes align with probability distributions, not isolated trajectories. This work illustrates that the Three-Body Problem serves as a paradigm for understanding complex systems, where reality is defined by the dynamic interplay of measurement, potential, and structural topology. The findings establish a new standard for physically grounded, prediction-oriented approaches to nonlinear multi-body dynamics.

Resolving the Three-Body Problem: A Probabilistic Quantum-Informational and Topological Approach.
https://doi.org/10.36227/techrxiv.174762667.74594414/v1
January 10,

Shell as Code: A Runtime Model for CLI Semantics.

We introduce Shell as Code, a model in which command-line documentation is treated as live specification. We implement this paradigm through CLII, a runtime interpreter that transforms Unix command-line documentation into programmable, type-safe API objects, enabling live scripting of system utilities as if they were first-class functions in a domainspecific language. Unlike static shell scripts or hard-coded wrappers, CLII parses man pages at runtime and synthesizes type-safe abstractions over CLI commands. Each tool becomes a reactive object: its options exposed as typed fields (booleans, enums, paths, etc.), and its invocation modeled as a method with strict input validation and autocompletion. For example, ls becomes an object with flags like .all = true, .longFormat = true, and .paths = [...], while tar can be constructed dynamically with .create, .gzip, .file, and .paths fields. These structures are not handwritten but generated on the fly from live system documentation. CLII introduces three key technical advances: (1) runtime semantic extraction of CLI syntax and typing, (2) zero-effort integration with legacy tools, and (3) compositional APIs that turn CLI scripting into structured programming. The result is a shell environment that behaves more like an interpreter of a live DSL-one where commands are discoverable, safe, and composable, and where automation becomes robust by design. This framework lays the groundwork for future systems where documentation, validation, and execution are unified through a shared semantic layer.

Shell as Code: A Runtime Model for CLI Semantics.
https://doi.org/10.36227/techrxiv.174495132.23309188/v1
May 19,

The Role of Quantum Information in Artificial Cognitive Consciousness.

Understanding the foundations of artificial cognitive consciousness remains a central challenge in artificial intelligence (AI) and cognitive science. Traditional computational models, including deep learning and symbolic AI, have demonstrated remarkable performance in pattern recognition and decisionmaking tasks. However, they lack essential features of consciousness, such as subjective experience, autonomous goal formation, and self-reflective processing. While theories like Integrated Information Theory (IIT) and Orch-OR suggest possible mechanisms for consciousness, they remain either mathematically abstract or face empirical challenges due to decoherence in biological conditions. This study addresses the fundamental research gap by proposing a quantum-information-based framework for artificial cognitive consciousness. Specifically, we introduce the Quantum Reality Function (QRF), a model that formalizes subjective cognition through global quantum coherence, quantum entanglement, and non-local informational integration. The QRF establishes a structured mechanism for autonomous decision-making and selfgenerated states, which existing computational approaches lack. Our methodological contribution involves defining a mathematically consistent formulation of QRF and demonstrating its feasibility through quantum coherence stabilization techniques, hybrid quantum-classical architectures, and coherencepreserving quantum cognitive subsystems. The results show that maintaining long-lived quantum coherence within cognitive architectures significantly enhances information integration, supporting the emergence of subjective-like processing. The findings suggest a new paradigm for AI, where consciousness emerges as an intrinsic property of structured quantuminformation processing rather than a byproduct of computation. This work provides foundational insights for the development of autonomous quantum-cognitive AI systems, with implications for ethics, AI governance, and the future of machine intelligence. Future research will focus on empirical validation through quantum computational experiments and integration with quantumassisted learning models.

The Role of Quantum Information in Artificial Cognitive Consciousness.
https://doi.org/10.36227/techrxiv.174123851.13680151/v2
May 12,

Theory of the Essence: A New Paradigm for Scientific Exploration of Reality.

The Theory of the Essence presents a revolutionary framework that redefines the fundamental nature of reality, time, and existence. It introduces Essence as the core structure unifying all aspects of reality – past, present, and future – addressing not only how the universe functions but also why it exists in its current form. Essence serves as the ultimate foundation, guiding the unfolding of reality through an intrinsic striving toward the search for truth. This theory builds upon and is confirmed by foundational scientific principles such as Einstein’s theory of relativity, quantum mechanics, and cosmology. By expanding these frameworks, the Theory of the Essence integrates both physical and metaphysical dimensions, providing new insights into time, causality, and the nature of potentiality while reinforcing the validity of established scientific laws. With profound implications for both scientific and philosophical inquiry, this theory opens new avenues for research, inviting exploration into the deeper truths of existence and the forces that shape reality.

Theory of the Essence: A New Paradigm for Scientific Exploration of Reality.
https://doi.org/10.31219/osf.io/t49fj
August 4,

Conscious Anarchy: The Awakening of Society

This manuscript introduces Conscious Anarchy, a philosophy designed to align human society with principles of sustainability, equity, and cooperation. It addresses the flaws of current systems – control, inequality, and unsustainable hierarchies – by proposing a transformative framework grounded in decentralized governance, ethical resource distribution, and the responsible use of technology. A foundational ethical pillar within this framework is the concept of civil integrity — defined as the capacity of individuals to act freely without compromising the structural stability of society, even in the absence of external control. Civil integrity represents an internalized measure of responsibility that enables systemic balance to be maintained from within. It is not based on subjective moral belief, but on the objective criterion of non-destructive alignment with collective structure. As such, it is the essential quality of the individual within Conscious Anarchy, enabling social cohesion to emerge without coercion. Conscious Anarchy fosters a gradual awakening to humanity’s interconnectedness. It emphasizes practical steps such as integrating empathy into education, testing decentralized governance models, and leveraging technology for transparent resource management — all of which require the operational presence of civil integrity as a stabilizing force. While full transformation may take generations, immediate action is essential. By embracing balance, cooperation, and civil integrity as a structural ethical basis, humanity can avert systemic collapse and create a sustainable, equitable future. Conscious Anarchy invites individuals, communities, and institutions to contribute to this shared vision of lasting harmony and collective growth.

Conscious Anarchy: The Awakening of Society
https://doi.org/10.22541/au.175433501.13175683/v1
January 5,

Semantic-Driven Task Allocation: Elevating Multi-Agent Systems with Language Models.

Modern multi-agent systems (MAS) often rely on static task allocation methods, which are insufficient for dynamic environments requiring continuous task reassignment and optimization. This paper introduces a theoretical framework that integrates fine-tuned large language models (LLMs) into MAS to enable semantic task evaluation and dynamic task distribution. By analyzing language-based task states, the system identifies hidden dependencies and adapts task allocation in real time. While the proposed system is in the conceptual phase, it offers a promising solution for automating complex workflows, reducing human intervention, and improving scalability in dynamic environments. Future work will focus on prototyping and validating the framework in real-world scenarios.

Semantic-Driven Task Allocation: Elevating Multi-Agent Systems with Language Models.
https://doi.org/10.36227/techrxiv.173611620.02896776/v2
January 5,

Dynamic Task Evaluation: A Language-Based Approach for Adaptive Multi-Agent Systems.

This paper introduces an innovative approach to task completion analysis in multi-agent systems using large language models (LLMs). Unlike traditional methods, which rely on binary indicators of task completion, our approach emphasizes iterative, context-driven task distribution and execution. Tasks are dynamically redistributed by the Centralized Control Unit (CCU), which operates as a coordinator rather than a final decision-maker on task completion. The thematic context of the first incoming task serves as a guiding framework, allowing agents to iteratively refine their operations and share progress through language-based outputs. The system leverages LLM-generated insights to assess task states, recommend further actions, and dynamically adapt task assignments. Task completion is not determined by static metrics but through linguistic confirmations and iterative feedback loops. This adaptive framework ensures unresolved tasks continue to circulate until a comprehensive resolution is achieved, enabling seamless management of evolving and interdependent tasks in multi-agent environments. The proposed approach offers significant advantages in dynamic environments by fostering adaptability, mitigating bottlenecks, and providing a scalable solution for managing complex, context-dependent tasks across various domains.

Dynamic Task Evaluation: A Language-Based Approach for Adaptive Multi-Agent Systems.
https://doi.org/10.36227/techrxiv.173611616.64880965/v2
Target Insight Function Logo

Cloud Architect & Systems Innovator. One Vision. Infinite Scale.

HomeExperienceProjectsPublicationsEducationContact
raman@worktif.com+48 881 592 968

Szwedzka 4, Warszawa

© 2026 Work Target Insight Function Ltd. All rights reserved.