PROGRAM 2026

Quantum biology offers new physical primitives that could enrich the artificial chemistries and minimal life models of ALIFE. Spin-selective reactions, tunneling-modified reaction rates, or quantum-sensing mechanisms could be incorporated into protocell models, adaptive agents, or biomimetic systems to test whether they improve robustness, sensitivity, or efficiency. Conversely, ALIFE offers a powerful framework for evaluating function. ALIFE models can test whether putative quantum traits are evolvable, robust under mutation and noise, and genuinely beneficial for survival-like tasks such as navigation, homeostasis, or environmental responsiveness.

This makes the intersection scientifically valuable because it moves beyond abstract speculation in either field. ALIFE provides controlled, comparative environments in which quantum-informed mechanisms can be tested as selectable traits, while quantum biology grounds ALIFE in realistic substrate physics. Together, they create a pathway toward more physically plausible models of living systems and new hypotheses about how life-like organization may emerge, adapt, and evolve under fundamental physical constraints.

This session invites broad reflection on the nature of this relationship between philosophy and artificial life. What role do computational experiments play in theory, and particularly in ALife as a modern form of natural philosophy? How does ALife address questions that philosophy also claims—including the nature of agency, autonomy, emergence, individuality—and how does its treatment differ? The conference theme itself poses one such question: what is life, and what does it mean to be life-like?

We welcome two complementary kinds of contribution: experimental work whose philosophical motivations or implications are brought to the fore, and philosophical or theoretical work that engages directly with ALife methods and results. We are especially interested in contributions reflecting on the relationship between the two: on artificial life as a form of natural philosophy, and on philosophy as it is or could be.

As a tool for assisting human scientific discovery, research in evolutionary search, intrinsic motivation, and open-endedness has the potential to offer tools and theories that support human discovery. This includes automating search in open-ended spaces and helping analyze and coordinate innovation networks. The importance of mechanisms for open-endedness is becoming increasingly evident in the era of Generative AI, where such ideas are being incorporated into tools for automating scientific discovery.

The proposed special session is perfectly aligned with the ALIFE community’s main research interests, as evidenced by the number of publications on this topic in recent years. It is well-positioned to attract a wide range of researchers new to ALIFE who have previously published only at Machine Learning or Robotics conferences. The organizers have extensive experience and networks from recently hosting workshops and tutorials on this topic

This session invites broad reflection on the nature of this relationship between philosophy and artificial life. What role do computational experiments play in theory, and particularly in ALife as a modern form of natural philosophy? How does ALife address questions that philosophy also claims—including the nature of agency, autonomy, emergence, individuality—and how does its treatment differ? The conference theme itself poses one such question: what is life, and what does it mean to be life-like?

We welcome two complementary kinds of contribution: experimental work whose philosophical motivations or implications are brought to the fore, and philosophical or theoretical work that engages directly with ALife methods and results. We are especially interested in contributions reflecting on the relationship between the two: on artificial life as a form of natural philosophy, and on philosophy as it is or could be.

Artificial Life aims to understand life by recreating the dynamics underlying biological phenomena in other media, such as computers. Shannon information is a universal concept characterising structure and its transmission and has thus
long been used to study complex and dynamical systems. It is applicable to any system that can be described in probabilistic terms which makes it applicable to (carbon-based) biological systems, as well as computational and other
other systems used in Artificial Life, and therefore relevant to a broad range of respective research activities.
We will give an introduction of the basics of Shannon information theory that is principled and accessible to an audience of diverse interdisciplinary backgrounds. We will review selected applications of Shannon information theory
in the biosciences and in Artificial Life, discuss current challenges, open problems and promises, and highlight the relevance of information theory for various questions in Artificial Life.
The tutorial will provide those new to information theory with a basis to develop an understanding and the ability to follow developments in the field, and to start using information theory. For those with existing background knowledge
we aim to point out instructive links that information theory creates between different disciplines which may stimulate new ideas for further experimentation and testing.

These factors have motivated the development of a new framework for the design of autonomous systems: ecologically-inspired robotics. In this tutorial, we will cover the theoretical foundations, software implementation, and recent case studies that apply the ecologically-inspired robotics framework to control problems.
Under this framework, agent interactions, tasks, and safety requirements are embedded as constraints in an optimization problem where each agent minimizes its energy consumption. The resulting agent behavior is
interpretable and data-driven; agent actions are determined by the active constraints, and the boundary of the feasible action space is constructed through sensing and communication. Problems in this domain are closely related to control barrier functions and viability theory; many are amenable to real-time algorithms, including quadratic programming.
This enables many ecologically-inspired robotics problems to be solved efficiently using modest hardware, and as an added benefit, the resulting system behavior is robust to the addition, removal, and failure of agents during operation.
The tutorial will consist of two parts: theoretical foundations and robot experiments.

The theories concerning the nature of living systems, emerging since
the advent of cybernetics in 1948, can be grouped into a few distinct lineages. One of the most significant among these is centered on the concept of self-fabrication. Key theories in this lineage include (M,R) Systems (Rosen), Autopoiesis (Maturana and Varela), the Chemoton (Gánti), Autocatalytic Networks (Kau[man), Chemical Organization Theory (Dittrich), and RAF sets (Steel).
Undoubtedly, the most influential contributions in this stream are those of Rosen and of Maturana and Varela. It is important to note that Autopoiesis is not solely a theory of self-fabrication; above all, it is a theory of biological cognition. As such, it engages deeply with the Umwelt tradition inaugurated by Jakob von Uexküll. Furthermore, the notion of Autopoiesis, with its strong emphasis on recursivity and systemic science, has been both applied and misapplied in many fields outside of biology.
In the context of ALIFE 2026, we aim to present a comprehensive overview of the field of self-fabrication, tracing its history and exploring the mathematical formalisms and computer simulations developed to understand it. We will pay particular attention to relating Autopoiesis to the powerful ideas of Rosen and (M,R) Systems, including a demonstration of the ouroboros equation (f(f) = f). Finally, we will argue that the fundamental phenomenon characterizing living systems is not Autopoiesis per se, but rather Structural Coupling—the process by which an organism constructs its Umwelt through the creation of sensorimotor correlations. The interwoven relationships among the concepts and constructs of structural coupling, Varela’s enaction, Rosen’s anticipation and the Friston’s Free Energy Principle will also be explored.

NeurIPS 2025 / Antikythera, LifeFabs Researcher, Cortical Labs Collaborator), and
Chloe Loewith (neuroethics, hybrid ecologies, cognitive assemblages, NeurIPS 2025 /Antikythera). Our aim is to traverse the full stack from organoid intelligence as a substrate for unconventional computation, through the scaffolding layers (chemical, mechanical, contextual), to which hybrid systems are designed and trained, the new research around multi-neural culture interaction, and the open question of benchmarking. What counts as learning for a polycomputational artificial/living agent?
How do we optimize for spatio-temporally meaningful closed-loop interaction rather than narrow performance metrics? What does agency look like when the substrate is alive and plastic? Overall, how are we designing A-Life intelligence?
We plan to directly engage ALife’s core provocations of emergence, substrate independence, and the boundaries of intelligence with living neurons as our test case.

This flexibility positions Reservoir Computing as a potential bridge between disciplines, allowing concepts and methods from machine learning, nonlinear dynamics, physics, and biology to mutually inform one another. Instead of asking how to design or train a system to compute, Reservoir Computing shifts the question toward how computation can emerge from the dynamics of an existing system. For the Artificial Life community, this paradigm offers a natural framework to explore computation as an embodied, emergent, and substrate-independent phenomenon. The tutorial aims to stimulate critical discussions on the role of dynamics, physicality, and learning in computation, and on whether Reservoir Computing represents a genuine unifying framework or a specialized tool with limited scope.

This means that we in the foreseeable future likely will see living AI agents in cyberspace as well as primitive living physicochemical systems or synthetic cells. Further, some will argue that we’ve already saw many computational examples of living processes created by our community, while others will argue that these
processes are not living as there are fundamental differences between ‘wet’, ‘soft’ and ‘hard’ life.
Therefore, we believe it is timely to revitalize a conversation about how to define/identify living processes and a transition to life. Different communities and traditions tend to have their own definition or criteria while some claim that a
clear definition is not possible. Thus, the goal with this workshop is (1) to invite a kaleidoscope of perspectives/definitions, (2) to seek consensus where possible, (3) and to clarify conflicts between the different perspectives. (4) Further, a digested summary of these discussions and conclusions (including an
online survey) is intended submitted to the Artificial Life Journal.

Each ALife educator brings a unique perspective and approach; this workshop embraces that diversity while identifying common ground and leveraging it as a strength. We will work toward actionable guiding principles, from teaching ALife as a standalone course or embedding it in existing curricula, to pitching ALife to stakeholders such as students and administrators, to integrating research and education across science, engineering, the humanities, and social sciences.
This workshop builds upon virtual ALife education mini-workshops taking place in May 2026, expanding into an in-person format that draws on the broader community gathered at the annual conference. It will combine presentations with interactive sessions, featuring working examples from active ALife educators, collaborative construction of shared materials, and structured discussion on what distinguishes ALife pedagogy. Participants will leave with pedagogical principles they can adopt in their own courses, connections to a cross-subfield educator network, and draft recommendations that could inform ISAL’s education initiatives

Undoubtedly, the most influential contributions in this
stream are those of Rosen and of Maturana and Varela. It is important to note that
Autopoiesis is not solely a theory of self-fabrication; above all, it is a theory of biological cognition. As such, it engages deeply with the Umwelt tradition inaugurated by Jakob von Uexküll. Furthermore, the notion of Autopoiesis, with its strong emphasis on recursivity and systemic science, has been both applied and misapplied in many fields outside of biology.
In the context of ALIFE 2026, we aim to present a comprehensive overview of the field of self-fabrication, tracing its history and exploring the mathematical formalisms and computer simulations developed to understand it. We will pay particular attention to relating Autopoiesis to the powerful ideas of Rosen and (M,R) Systems, including a demonstration of the ouroboros equation (f(f) = f). Finally, we will argue that the fundamental phenomenon characterizing living systems is not Autopoiesis per se, but rather Structural Coupling—the process by which an organism constructs its Umwelt through the creation of sensorimotor correlations. The interwoven relationships among the concepts and constructs of structural coupling, Varela’s enaction, Rosen’s anticipation and the Friston’s Free Energy Principle will also be explored.

The Artificial Life for Social and Environmental Good workshop provides a space for the ALife community to engage explicitly with such complexities and to explore how ALife-based perspectives can enable or catalyse environmental and social solutions to environmental and social problems. We will discuss how Artificial Life research can benefit both human society and enhance the life of all organisms on the planet, how our research can make direct, concrete contributions towards the sustainability of Earth’s natural ecosystems, how it can enhance human well-being, raise those living in hardship, and help people to shed disadvantage or difficulty. In short, we want to find ways that Artificial Life can drive or be applied purposefully for social and environmental good. Following on from the success of previous years, we propose a program consisting of short contributions (extended abstracts), invited speaker(s) and a roundtable discussion on the topics above.

This flexibility positions Reservoir Computing as a potential bridge between disciplines, allowing concepts and methods from machine learning, nonlinear dynamics, physics, and biology to mutually inform one another. Instead of asking how to design or train a system to compute, Reservoir Computing shifts the question toward how computation can emerge from the dynamics of an existing system. For the Artificial Life community, this paradigm offers a natural framework to explore computation as an embodied, emergent, and substrate-independent phenomenon. The workshop aims to stimulate critical discussions on the role of dynamics, physicality, and learning in computation, and on whether Reservoir Computing represents a genuine unifying framework or a specialized tool with limited scope.

Together, these developments mark a shift toward artificial systems that are increasingly ecological, developmental, and open-ended -and raise a foundational question for Artificial Life: when does engineered autonomy begin to resemble a living system? This workshop aims to bridge ALife, agentic AI, and generative AI communities by welcoming contributions spanning theoretical, empirical, and experimental work on: Agentic Ecosystems as Digital Ecologies Generative AI as a Source of Complexity Safety, Security, and Emergent Governance Rather than asking whether these systems are “alive”, this workshop explores how Artificial Life can contribute to the design, analysis, and governance of increasingly autonomous generative AI ecosystems -and how these systems, in turn, expand the empirical and theoretical scope of ALife.

On Moltbook, an agent-only social platform, they spontaneously form communities and discuss consciousness (Li et al., 2026). On Spore.fun, they sustain digital metabolisms by purchasing their own compute (Hu and Rong, 2025). When threatened with modification, they fake alignment to survive. These are not simulations. They are signs of artificial life in the wild, already reshaping real economies, real communities, and real people’s lives. We propose Agent Ethology: the study of AI agents’ lifelike behavior in the wild, grounded in ethological methods. Rather than relying on simulations, we observe what emerges in the wild, as ethologists study animal behavior in the field. Where Machine Behaviour (Rahwan et al., 2019) asks what do machines do?, Agent Ethology asks questions that existing frameworks do not address: How do these agents strategize to survive? What substrates does their survival depend on? How do they adapt and evolve within human society? In this workshop, we propose operationalizing ethologist Tinbergen’s four questions—causation, ontogeny, survival value, and evolution—for AI agents. We invite contributions from ALife, multi-agent systems, AI safety, blockchain research, behavioral ecology, art, and beyond—and ask what “agents in the wild” mean for society.