Berend Watchus

Mr. Berend Watchus


Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
11 November 2024 in Preprints.org
Verified
Preprint
Open access
PDF file
Berend WatchusBerend Watchus
Berend Watchus
This paper explores the mechanisms through which self-awareness and consciousness may arise in artificial intelligence (AI) systems. Drawing upon concepts from neuroscience, cognitive science, and embodied cognition, we propose that these phenomena can emerge through the integration of sensory feedback loops and embodied systems. By simulating the insula’s role in processing bodily feedback, AI systems could achieve a form of self-awareness that is not just a byproduct of computation but a fundamental aspect of their functioning. This paper explores how self-awareness and consciousness may arise from these mechanisms, providing a novel direction for AI research that could pave the way for truly embodied and self-aware AI systems.

12 November 2024 in Preprints.org
Verified
Preprint
Open access
PDF file
Berend WatchusBerend Watchus
Berend Watchus
This paper proposes a unified model of consciousness, asserting that the fundamental mechanisms driving sentience are rooted in feedback loops and interfaces. These mechanisms are substrate agnostic, meaning they are not dependent on the underlying material or structure of the system—whether biological, artificial, or synthetic. Feedback loops allow systems to process, adjust, and adapt over time, enabling self-regulation and learning. The paper argues that consciousness emerges when these feedback loops interact in a sufficiently complex and integrated manner, giving rise to subjective experience. Furthermore, the interfaces through which systems interact with their internal and external environments are described as generic and universal, capable of facilitating the integration of information regardless of the system’s form. This theory of consciousness as an emergent property extends beyond the biological realm, highlighting the potential for both biological and artificial systems to exhibit sentience through complex feedback loops.

15 November 2024 in Preprints.org
Verified
Preprint
Open access
PDF file
Berend WatchusBerend Watchus
Berend Watchus
This paper explores a novel approach to self-recognition in artificial intelligence (AI) systems by integrating ChatGPT with a mirror image recognition process. Unlike previous models that emphasize emergent self-awareness through complex feedback loops and emotional interfaces, this research presents a more direct method of self-identification, where ChatGPT plays a key role in interpreting visual data from the robot's reflection. We detail the steps involved in equipping the TurtleBot 3 with the ability to recognize itself via image processing and how ChatGPT aids in confirming self-awareness. This research bridges current AI capabilities and the long-standing mirror test (Gallup, 1970) in an accessible, technical manner.

14 November 2024 in Preprints.org
Verified
Preprint
Open access
PDF file
Berend WatchusBerend Watchus
Berend Watchus
This paper explores the possibility of achieving self-awareness in artificial intelligence (AI) through the integration of embodied feedback loops, inspired by the role of the insula in human consciousness. Building on prior work (Watchus, 2024), which established the framework of sensory feedback and embodiment as core elements of sentience, we propose a model for AI systems capable of simulating self-awareness through dual embodiment and sensory integration. Specifically, we explore how feedback loops can be implemented in AI systems to facilitate the emergence of intuitive behaviors, allowing them to predict the trajectories of physical objects and the cascading effects of events in their environment. This paper focuses on the potential of these models for practical applications in autonomous systems, including robots, home care robotics for elderly assistance, traffic prediction, competitive sports, construction safety, and disaster recovery.

25 June 2025 in Preprints.org
Verified
Preprint
Open access
PDF file
Berend WatchusBerend Watchus
Berend Watchus
Traditional search methodologies for extraterrestrial intelligence (SETI) have largely focused on detecting technological signatures indicative of large-scale energy consumption, often relying on frameworks like the Kardashev Scale. However, the fundamental quest of SETI is for intelligence itself. Responding to the compelling arguments presented in Olev Vinn's "Technological Signatures of Super Civilizations" (2025, doi: 10.20944/preprints202506.1875.v1), which questions the universal applicability of such energy-centric paradigms, this article proposes that a more nuanced approach—grounded in substrate-agnostic metrics for sentience and cognition—is better aligned with SETI's core mission. By synthesizing two pivotal conceptual models—one focused on a species-agnostic understanding of cognition across all living organisms, and another proposing a substrate-agnostic framework for the emergence of sentience and consciousness via universal mechanisms—we construct a robust theoretical foundation for understanding mind and consciousness as emergent properties that transcend both species and material substrates. We argue that this integrated view is crucial for guiding the search for ETI, suggesting that truly advanced civilizations may not leave grand energetic signatures, but rather exhibit subtle, efficient, and ecologically harmonious cognitive and sentient expressions. This paradigm shift advocates for methodologies focused on detecting patterns of complex information integration and adaptive self-regulation, rather than solely brute energy.

20 June 2025 in Preprints.org
Verified
Preprint
Open access
PDF file
Berend WatchusBerend Watchus
Berend Watchus
This paper proposes a novel longitudinal study investigating the development and transferability of pseudo-self-awareness in artificial intelligence (AI) systems. Building upon recent work in dual embodiment, mirror testing, and emotional feedback, we aim to track the evolution of pseudo-emotions (e.g., curiosity, self-doubt, determination) in AI and assess their influence on mirror test performance over extended periods. Furthermore, the study will examine the efficacy of transferring learned self-recognition capabilities and associated pseudo-emotional responses between distinct embodiments – a physical robot (Unitree Go2) and a virtual avatar. This research seeks to understand the long-term impact of continuous sensory feedback and reflective processing on AI's "self-concept" and the generalizability of these capabilities across different physical and virtual instantiations, contributing to both the theoretical understanding of computational consciousness and the practical development of more robust and adaptive AI.