SYNTROPIC COUNTERPOINTS
Project:
Botohara: Ethics of the Machines is an AI-driven installation structured as a circular confrontation between six AI
philosopher-clones — Socrates, Confucius, Mao Zedong, Vladimir Ilyich Lenin, Maximilien Robespierre, and
Thomas Jefferson — and a central evaluative samurai agent. Each clone is trained on thinker-specific textual
corpora derived from published writings or, in the case of Socrates and Confucius, from teachings recorded by their
students and followers. The agents generate new arguments in real time, grounded in their respective intellectual
traditions. The installation operates as a multi-agent dialogue system in which the philosopher-clones confront one
another while addressing the central samurai robot. Mao, Lenin, Robespierre, and Jefferson reflect political
traditions shaping modern global powers, while Socrates and Confucius represent foundational Western and East
Asian philosophical frameworks. The AI samurai clone participates in the exchange while evaluating arguments
against an honour-based ethical hierarchy derived from Bushido principles. Incoming arguments are analysed for
alignment or contradiction with this hierarchy. Accumulated contradiction increases an internal tension variable,
externalised through heartbeat sonification. The title Botohara combines “Boto” (robot) and “hara” (harakiri). When
ideological conflict exceeds a defined computational threshold, the heartbeat accelerates and then ceases. This
symbolic machine self-termination marks Botohara, followed by system reset and continuation. Each execution
represents a moment where civilisational value systems fail to satisfy honour-based evaluation, questioning whether
autonomous machines will reproduce human moral compromise or withdraw from it.
Type of the project: Art & Science Project
Year the project was created: 2026
Credits:
Artistic Direction and Concept Dr. Predrag K. Nikolic
Conceptual Framework and Philosopher Corpus Design Dr. Predrag K. Nikolic
Dialogue Analysis and Reasoning Structures Dr. Robert Prentner
AI System Design and Technical Development Jason Thomas Chew
Objectives: To examine how AI agents trained on thinker-specific textual corpora negotiate conflicting civilisational
value systems through real-time multi-agent dialogue. To investigate whether an artificial system can perform
ethical evaluation across historically grounded ideological frameworks using an internally coherent moral hierarchy.
To explore computational self-interruption (Botohara) as a mechanism for marking moments of irreconcilable
contradiction between competing traditions. To question how intellectual heritage is transferred, transformed, or
destabilised when embedded within artificial systems that may participate in future hybrid human–AI societies. To
position AI not only as a generative tool but as an evaluative entity capable of confronting inherited ethical
structures.
Project History: Botohara: Ethics of the Machines builds on an ongoing artistic research trajectory initiated in 2017
with the installation Syntropic Counterpoints: Robosophy Philosophy, which explored philosophical dialogue
generated between AI-cloned Aristotle and Nietzsche. This work introduced AI clones trained on thinker-specific
textual corpora as generative conversational agents engaging in debates on morality, ethics, and aesthetics.
Subsequent projects developed within the Syntropic Counterpoints framework expanded this approach through
Botorikko (2019), in which AI clones trained on the writings of Machiavelli and Sun Tzu engaged in real-time
debates on politics, diplomacy, and conflict, and Metaphysics of the Machines (2021), which extended autonomous
machine dialogue toward questions of AI imagination, metaphysics of word and grammar, content vs. context and
machine reasoning. More recent research conducted under the Epistemoverse framework further developed these
systems into multi-agent environments for the interpretation and preservation of intellectual heritage through
AI-driven interaction. Botohara translates these investigations into a spatial installation by introducing a central
evaluative agent trained on Bushidn principles. By placing AI philosopher-clones derived from intellectual traditions
shaping contemporary geopolitical systems into sustained dialogue with an honour-based ethical framework
external to those traditions, the work extends previous research from machine-generated discourse toward
computational ethical evaluation (confronting human ethical evaluation).
Use of AI: Description: Artificial intelligence functions as the core generative and evaluative mechanism of the
installation. Each philosopher-clone is implemented as an autonomous agent trained on thinker-specific textual
corpora derived from historical writings or recorded teachings. Rather than reproducing quotations, the agents
generate new arguments grounded in the ideological and ethical structures embedded within their respective
corpora. The installation operates as a real-time multi-agent conversational system in which philosopher-clones
produce emergent dialogue through continuous interaction. A central AI samurai agent participates in the exchange
while simultaneously performing ethical evaluation. The samurai’s evaluative mechanism is structured as a
hierarchical ethical model derived from Bushido principles. Incoming arguments are semantically analysed for
alignment, contradiction, and coherence relative to this hierarchy. Accumulated contradiction dynamically increases
an internal evaluative tension variable, which modulates heartbeat sonification in real time. When the tension
variable exceeds a defined computational threshold, the system executes Botohara — symbolic machine
self-termination — suspending dialogue before resetting and resuming operation. AI in Botohara therefore operates
simultaneously as generative discourse engine, conflict amplifier, and autonomous ethical evaluator within a
closed-loop computational system.
Software: The installation operates as a multi-agent conversational system implemented using Python-based
frameworks for dialogue orchestration and real-time interaction between autonomous agents. AI philosopher-clones
and the central evaluative samurai agent are instantiated as independent generative modules trained on
thinker-specific textual corpora. Agent interaction and dialogue generation are managed through large language
model architectures accessed via API-based inference environments. The evaluative framework based on Bushidn
principles is implemented as a rule-based ethical hierarchy integrated within the central agent’s response
generation pipeline. Real-time sonification of the samurai’s internal evaluative state is produced using digital audio
processing libraries and triggered dynamically in response to the degree of ethical coherence detected during
dialogue. The system operates locally and can run continuously in looped mode without requiring user interaction.
Hardware: The installation operates using a workstation-class computer system responsible for running the
multi-agent dialogue framework and real-time audio output. Visualisation is presented via a circular arrangement of
display monitors representing the philosopher-clone agents and the central evaluative samurai agent. Audio output
is delivered through a standard speaker configuration positioned within the installation space to support perception
of dialogue and sonified system state. An optional robotic body representing the central AI samurai agent may be
incorporated as a physical embodiment of the evaluative node. This component operates independently from the
core dialogue system and can be included or omitted depending on spatial or logistical constraints. The installation
does not require specialised sensors or audience interaction devices and can operate continuously in looped mode
using standard exhibition infrastructure.
Collaboration information: Botohara is developed as part of an ongoing collaborative research trajectory led by
Dr. Predrag K. Nikolic, working together with Jason Thomas Chew and Dr. Robert Prentner. The collaboration
extends across multiple projects, including Epistemoverse, and continues to evolve within a shared framework of
AI-driven interpretation of intellectual heritage. Dr. Predrag Nikolic serves as the initiator and artistic lead,
responsible for the conceptual architecture of the project, the design of the multi-agent system, and the integration
of philosopher-clones within the installation environment. Dr. Robert Prentner, an interdisciplinary researcher
working at the intersection of artificial intelligence and consciousness studies and affiliated with the Association for
Mathematical Consciousness Science (AMCS), contributes to the analytical and theoretical examination of
reasoning structures, evaluative coherence, and emergent patterns within the multi-agent system. Jason Thomas
Chew contributes to the technical development and implementation of the dialogue orchestration framework,
supporting the coordination of generative agents and the real-time interaction between philosopher-clones and the
central evaluative samurai agent. This collaboration forms part of a continuing interdisciplinary exchange combining
artistic research, technical development, and philosophical inquiry, with Botohara and Epistemoverse representing
interconnected stages within a broader digital humanities programme exploring artificial agents as interpretative
entities in hybrid human–AI cultural environments.
People: Dr. Predrag K. Nikolic Dr. Robert Prentner Jason Thomas Chew
Dr. Predrag K. Nikolic is an artist and creative researcher working with artificial intelligence as a
medium for interpreting cultural and intellectual heritage, currently working as a professor at the
Swinburne University of Technology. His practice examines how computational agents engage
with philosophy, civilisational value systems, and historical discourse within emerging hybrid human–AI societies.
Through multi-agent installations and generative dialogue systems, he approaches AI as an interpretative entity
rather than a passive tool. His long-term research trajectory includes Robosophy Philosophy (2017), Botorikko
(2019), Metaphysics of the Machines (2021), and Epistemoverse (2024), projects that explore the simulation and
transformation of intellectual traditions through computational frameworks. Positioned within Digital Humanities, his
work investigates how artificial systems can preserve, reinterpret, and critically examine historically grounded
knowledge in technologically mediated cultural environments.
Url: http://www.predragnikolic.com
Dr. Robert Prentner is an Assistant Professor at the Institute of Humanities at ShanghaiTech University,
where his work operates at the intersection of philosophy, science, and emerging technology. His research primarily
focuses on consciousness studies, specifically investigating mathematical approaches to phenomenology, the
interface theory of perception, and AI consciousness. He is interested in how shifting scientific paradigms and
technological advancements fundamentally reshape human knowledge and subjective experience. Prentner has
served as the co-editor of the interdisciplinary journal Mind and Matter since 2017 and has been a co-organizer of
the Models of Consciousness (MoC) conference series since 2019. His work has been published in various leading
international journals. By integrating formal tools, philosophy, and creative practice, he aims to advance the future of
sentient and technologically-mediated environments.
Jason Thomas Chew is a Lecturer at Swinburne University of Technology Sarawak Campus whose
work sits at the intersection of deep metric learning, graph modeling, and agentic AI systems. He completed his
MSc in graph modeling in 2020 and is currently completing his PhD in deep metric learning with applications in
medical informatics. Jason’s research focus is in translational, multidisciplinary work, he designs and builds
end-to-end systems using cutting-edge technologies that can achieve tangible societal impact. His recent projects in
medical informatics investigate metric learning-based approaches to ECG arrhythmia classification, with the broader
goal of improving clinical decision support and expanding access to accurate, data-driven cardiac screening.
Beyond medical applications, Jason’s interests extend to graph-based modeling and graph mining, to agentic,
multi-agent AI systems. His work on AI-based multi-agent systems explores how diverse artificial “voices” can be
used in philosophical discourse and human–AI collaborative settings, with the aim of supporting richer, more
reflective forms of deliberation and decision-making. Across these domains, he is particularly interested in how AI
can be deployed responsibly in social, environmental, and humanitarian contexts, and how technical research can
be translated into accessible, real-world systems that deliver measurable public benefit.Type your paragraph here.
BOTOHARA: ETHICS OF THE MACHINES