Hiroki Sayama: Evoloops, Artificial Life, Complex Adaptive Networks, and Ideas as Living Organisms

Hiroki Sayama is a Japanese artificial life and complex systems scientist. He is a SUNY Distinguished Professor in the School of Systems Science and Industrial Engineering at Binghamton University, where he directs the Binghamton Center of Complex Systems (CoCo). Trained in information science at the University of Tokyo, he first made his mark in the field of artificial life with “evoloops,” the first self-reproducing cellular automaton loops shown to undergo genuine Darwinian evolution. He is the author of the freely available textbook “Introduction to the Modeling and Analysis of Complex Systems,” and over 260 peer-reviewed papers. He previously served as chief editor of the journal Complexity and as the vice president of the International Society for Artificial Life. A dedicated community builder, Sayama currently serves as the vice chair of the IEEE Computational Intelligence Society’s Task Force on Artificial Life and Complex Adaptive Systems and as an elected council, executive committee and CCS steering committee member of the Complex Systems Society. He also serves as vice president and board member of the Network Science Society, and founded the low-cost Northeast Regional Conference on Complex Systems and the open-access Northeast Journal of Complex Systems.

In this deep and far-reaching InterDialogue, recorded at the 2026 Conference on Complex Systems in Binghamton, which Sayama chaired, we trace the story of evoloops from John von Neumann’s self-reproducing automata to Christopher Langton’s loops to Sayama’s own pioneering work, before turning to his current research on open-ended artificial systems and the ongoing effort to define life itself. We explore the gap between the micro and the macro across disciplines, his view that life unfolds at every scale simultaneously, the Carl Sagan-inspired fascination with endless processes that has guided his career, and his work on adaptive networks, in which structure and dynamics co-evolve.

The conversation then turns to building open and accessible scientific communities, the value of interdisciplinarity, and the importance of listening when tackling real-world problems, as well as his surprising research showing that more isolated, locally connected “small world” networks can generate more diverse, higher-quality ideas than fully connected ones. We close with his concern that AI is homogenizing human thought, and his notion that ideas, and perhaps AI itself, are literally living, evolving organisms.

Watch at: www.youtube.com

A mechanism for collective foraging: percolating recruitment cascades coupled to a maximum-entropy field

Galen J. Wilkerson

Foraging recruitment is a contagion: an agent that finds a resource triggers a linear-threshold cascade, the model built for rumour and adoption spread on social networks, now embedded in space. Because activation is degree-normalized, the same rule that spreads the cascade also stifles it as neighbourhoods crowd, so the swarm aggregates on a resource and then releases to explore again. The cascade runs on the communication network, but acting on it moves agents in physical space, and that motion rewires the network: network and space are one coupled system, and this feedback, not any tuning, makes the coordination self-limiting. Because recruitment chains along the graph rather than reaching near neighbours alone, whether a discovery stays local or spans the swarm is a question of percolation. The swarm holds itself in a cascade window that is locally near-critical though globally subcritical, a self-organized, self-limiting drive-release cycle, sandpile-like but non-conservative. Nothing is tuned, yet behaviours that foraging biology documents separately emerge together: recruitment to discoveries, marginal-value patch-leaving, and Levy search with a switch to local exploitation. These follow from one least-committal principle, the resource field and the search both as maximum entropy, with Levy search matched to the field’s sparsity and recruitment to its clustering, so active-matter order and optimal foraging appear as one system seen in two domains. The gains are real but conditional: up to 7.6x the harvest of solitary search when resources are sparse and clustered, shrinking to nothing when search is already easy. That sparse, clustered field is also the generic one, because resources almost inevitably aggregate, so one untuned rule should recur widely.

Read the full article at: arxiv.org

Information engines: A physical model of evolutionary dynamics

Antonis Mistriotis

BioSystems
Volume 270, December 2026, 105982

An abstractive model based on the concept of information engines is proposed, which simulates evolution. The model, representing an elementary living system, comprises two interrelated stochastic optimization processes that maximize both its information content and energy-capturing performance. It is assumed that the system exploits stored information and selected mutations to optimize its capability to capture useful energy from the environment. It then consumes work to increase its information content further. Such a cross-amplification mechanism leads to the system’s evolution towards higher levels of information processing capacity and increased energy-gaining efficiency, which is expressed as a persistent increase in its negentropy. Consequently, the simulated evolution process exhibits directionality, which also characterizes biological evolution, as other studies have observed. Moreover, the proposed model is compatible with the recently presented hypothesis of Life’s Universal Architecture (LUA) that summarizes the structural requirements of a living system. The LUA hypothesis states that thermodynamic considerations impose the presence of information-processing and energy-gaining modules in all living systems.

Read the full article at: www.sciencedirect.com

A repeating sequence of simple rules creates a universal computer

Riz Fernando Noronha

We show that the Turing complete rule 110 in elementary cellular automata can be emulated by alternating three simple symmetric rules in space. As the constituent rules themselves are trivial, this shows that diversity alone can generate emergent phenomena. The rules cannot distinguish left from right, and the chirality of rule 110 is created by their arrangement. Random initial conditions lock on to rule 110 over 90% of the time, explained through the motion of domain walls between desynchronized regions. Adding random noise creates errors that seed domain walls, whose fraction scales with the square root of noise strength, explained through a simple rate equation. This points to a route to complexity through repeated, perturbed units, common in biology, and shows that in cellular automaton models of quantum field theory, broken symmetries such as parity need not be present in the local rules.

Read the full article at: arxiv.org