Arthur N. Montanari, Ana Elisa D. Barioni, Chao Duan, Adilson E. Motter

The study of flocking in biological systems has identified conditions for self-organized collective behavior, inspiring the development of decentralized strategies to coordinate the dynamics of swarms of drones and other autonomous vehicles. Previous research has focused primarily on the role of the time-varying interaction network among agents while assuming that the agents themselves are identical or nearly identical. Here, we depart from this conventional assumption to investigate how inter-individual differences between agents affect the stability and convergence in flocking dynamics. We show that flocks of agents with optimally assigned heterogeneous parameters significantly outperform their homogeneous counterparts, achieving 20-40% faster convergence to desired formations across various control tasks. These tasks include target tracking, flock formation, and obstacle maneuvering. In systems with communication delays, heterogeneity can enable convergence even when flocking is unstable for identical agents. Our results challenge existing paradigms in multi-agent control and establish system disorder as an adaptive, distributed mechanism to promote collective behavior in flocking dynamics.

Read the full article at: arxiv.org

Claudia Westermann

Constructivist Foundations 20(2): 67–71

Context: In 2024, we celebrated the 60th-anniversary meeting of the American Society for Cybernetics (ASC. In more than eighty, mostly participatory, sessions, the conference stretched over five days. Under the overarching theme “Living Cybernetics Playing Language,” the conference encouraged participants to reflect on cybernetics in everyday contexts, ranging from academic research to the building of communities. This special issue of Constructivist Foundations contains four target articles that emerged from this conference, and their related discussions. Problem: Sixty years after the foundation of the ASC, defining cybernetics is still a challenge. Diversity, one could say, has haunted cybernetics since its inception. There are many practices that refer to cybernetics in many disciplinary fields and contexts, but do these different practices share anything or do they rely on different aspects of (historical) cybernetic practices? Method: I present the contributions to this special issue as case studies of cybernetic practice and diversity and expose them to the questions mentioned above. Results: Cybernetic practices are as diverse in their methods as the disciplines to which they relate. And yet, as the study of the four target articles and the related commentaries show, these practices all embrace uncertainty. This embrace is the foundation for a particular technicity in which formation and reflexivity become intertwined and co-evolve. In its engagement with contemporary challenges, cybernetic technicity introduces recursive links setting relations across boundaries. Contemporary cybernetic practice, through varied approaches, is a living tradition of enacting open futures. Implications: Cybernetic thinking does not necessarily become detectable through a common vocabulary or set of references, but rather through a particular inherent logic, which links thinking and doing in a recursive co-evolving relationship. Constructivist content: The editorial discusses second-order approaches to cybernetics.

Read the full article at: constructivist.info

Ori Livson, Mikhail Prokopenko

Incomputability results in formal logic and the Theory of Computation (i.e., incompleteness and undecidability) have deep implications for the foundations of mathematics and computer science. Likewise, Social Choice Theory, a branch of Welfare Economics, contains several impossibility results that place limits on the potential fairness, rationality and consistency of social decision-making processes. A formal relationship between Gödel's Incompleteness Theorems in formal logic, and Arrow's Impossibility Theorem in Social Choice Theory has long been conjectured. In this paper, we address this gap by bringing these two theories closer by introducing a general mathematical object called a Self-Reference System. Impossibility in Social Choice Theory is demonstrated to correspond to the impossibility of a Self-Reference System to interpret its own internal consistency. We also provide a proof of Gödel's First Incompleteness Theorem in the same terms. Together, this recasts Arrow's Impossibility Theorem as incomputability in the Gödelian sense. The incomputability results in both fields are shown to arise out of self-referential paradoxes. This is exemplified by a new proof of Arrow's Impossibility Theorem centred around Condorcet Paradoxes.

Read the full article at: arxiv.org

THEOPI RADOS, et al.

SCIENCE 3 Apr 2025 Vol 388, Issue 6742 pp. 109-115

The advent of clonal multicellularity is a critical evolutionary milestone, seen often in eukaryotes, rarely in bacteria, and only once in archaea. We show that uniaxial compression induces clonal multicellularity in haloarchaea, forming tissue-like structures. These archaeal tissues are mechanically and molecularly distinct from their unicellular lifestyle, mimicking several eukaryotic features. Archaeal tissues undergo a multinucleate stage followed by tubulin-independent cellularization, orchestrated by active membrane tension at a critical cell size. After cellularization, tissue junction elasticity becomes akin to that of animal tissues, giving rise to two cell types—peripheral (Per) and central scutoid (Scu) cells—with distinct actin and protein glycosylation polarity patterns. Our findings highlight the potential convergent evolution of a biophysical mechanism in the emergence of multicellular systems across domains of life.

Read the full article at: www.science.org

Tiago P. Peixoto

Network reconstruction is the task of inferring the unseen interactions between elements of a system, based only on their behavior or dynamics. This inverse problem is in general ill-posed, and admits many solutions for the same observation. Nevertheless, the vast majority of statistical methods proposed for this task -- formulated as the inference of a graphical generative model -- can only produce a ``point estimate,'' i.e. a single network considered the most likely. In general, this can give only a limited characterization of the reconstruction, since uncertainties and competing answers cannot be conveyed, even if their probabilities are comparable, while being structurally different. In this work we present an efficient MCMC algorithm for sampling from posterior distributions of reconstructed networks, which is able to reveal the full population of answers for a given reconstruction problem, weighted according to their plausibilities. Our algorithm is general, since it does not rely on specific properties of particular generative models, and is specially suited for the inference of large and sparse networks, since in this case an iteration can be performed in time O(Nlog2N) for a network of N nodes, instead of O(N2), as would be the case for a more naive approach. We demonstrate the suitability of our method in providing uncertainties and consensus of solutions (which provably increases the reconstruction accuracy) in a variety of synthetic and empirical cases.

Read the full article at: arxiv.org

Chengbin Sun, Alfonso de Miguel-Arribas, Chaoqian Wang, Haoxiang Xia, Yamir Moreno

In real-life complex systems, individuals often encounter multiple social dilemmas that cannot be effectively captured using a single-game model. Furthermore, the environment and limited resources both play a crucial role in shaping individuals' decision-making behaviors. In this study, we employ an adaptive control mechanism by which agents may benefit from their environment, thus redefining their individual fitness. Under this setting, a detailed examination of the co-evolution of individual strategies and resource allocation is carried. Through extensive simulations, we find that the advantageous environment mechanism not only significantly increases the proportion of cooperators in the system but also influences the resource distribution among individuals. Additionally, limited resources reinforce cooperative behaviors within the system while shaping the evolutionary dynamics and strategic interactions across different dilemmas. Once the system reaches equilibrium, resource distribution becomes highly imbalanced. To promote fairer resource allocation, we introduce a minimum resource guarantee mechanism. Our results show that this mechanism not only reduces disparities in resource distribution across the entire system and among individuals in different dilemmas but also significantly enhances cooperative behavior in higher resource intervals. Finally, to assess the robustness of our model, we further examine the influence of the advantageous environment on system-wide cooperation in small-world and random graph network models.

Read the full article at: arxiv.org

Robert L. Axtell, J. Doyne Farmer
JOURNAL OF ECONOMIC LITERATURE VOL. 63, NO. 1, MARCH 2025 (pp. 197–287)

Agent-based modeling (ABM) is a novel computational methodology for representing the behavior of individuals in order to study social phenomena. Its use is rapidly growing in many fields. We review ABM in economics and finance and highlight how it can be used to relax conventional assumptions in standard economic models. ABM has enriched our understanding of markets, industrial organization, labor, macro, development, public policy, and environmental economics. In financial markets, substantial accomplishments include understanding clustered volatility, market impact, systemic risk, and housing markets. We present a vision for how ABMs might be used in the future to build more realistic models of the economy and review some of the hurdles that must be overcome to achieve this.

Read the full article at: www.aeaweb.org

Malbor Asllani, Alex Arenas

Phys. Rev. E 111, 044306

Chimera states, marked by the coexistence of order and disorder in systems of coupled oscillators, have captivated researchers with their existence and intricate patterns. Despite ongoing advances, a full understanding of the genesis of chimera states remains challenging. This work formalizes a systematic method by evoking pattern formation theory to explain the emergence of chimera states in complex networks, in a similar way to how Turing patterns are produced. Employing linear stability analysis and the spectral properties of complex networks, we show that the randomness of network topology, as reflected in the localization of the graph Laplacian eigenvectors, determines the emergence of chimera patterns, underscoring the critical role of network structure. In particular, this approach explains how amplitude and phase chimeras arise separately and explores whether phase chimeras can be chaotic or not. Our findings suggest that chimeras result from the interplay between local and global dynamics at different timescales. Validated through simulations and empirical network analyses, our method enriches the understanding of coupled oscillator dynamics.

Read the full article at: link.aps.org

Enrique M. Muro, Fernando J. Ballesteros, Bartolo Luque, and Jordi Bascompte

PNAS 122 (13) e2422968122

The origin of eukaryotes represents one of the most significant events in evolution since it allowed the posterior emergence of multicellular organisms. Yet, it remains unclear how existing regulatory mechanisms of gene activity were transformed to allow this increase in complexity. Here, we address this question by analyzing the length distribution of proteins and their corresponding genes for 6,519 species across the tree of life. We find a scale-invariant relationship between gene mean length and variance maintained across the entire evolutionary history. Using a simple model, we show that this scale-invariant relationship naturally originates through a simple multiplicative process of gene growth. During the first phase of this process, corresponding to prokaryotes, protein length follows gene growth. At the onset of the eukaryotic cell, however, mean protein length stabilizes around 500 amino acids. While genes continued growing at the same rate as before, this growth primarily involved noncoding sequences that complemented proteins in regulating gene activity. Our analysis indicates that this shift at the origin of the eukaryotic cell was due to an algorithmic phase transition equivalent to that of certain search algorithms triggered by the constraints in finding increasingly larger proteins.

Read the full article at: www.pnas.org

Selin E. Nugen

AI & SOCIETY

This paper examines the application of evolutionary analogy in AI (artificial intelligence) research, focussing on narratives that perpetuate individuated and autonomous imaginaries of AI systems through biological diction. AI research has long drawn inspiration from evolution to design and predict algorithmic change. Occasionally, these narratives extend inspiration to reimagine AI as a non-human species subject to the same evolutionary pressures as biological organisms. As AI technologies embed more pervasively in public life and require critical perspectives on their social impacts, these comparisons in AI discourse raise critical questions about the limits of and responsibility in employing such analogies and their potential impact on how broader audiences consume and perceive AI systems. This paper examines the diverse ways and intentions behind how evolution is invoked in AI research narratives by analysing the adaptation of individuating evolutionary language and concepts across three fields of AI-related research: evolutionary computing, Artificial Life, and existential risk. It scrutinises the challenge of accurate scientific communication when drawing inspiration from biological evolution and assigning organismal attributes to digital technologies whilst decontextualising wider evolutionary scholarly discourses. I argue that the intertwined history between evolutionary theory and technological change paired with the potential risks to wider perceptions of AI and biological evolution, requires (1) strategic consideration about the limits of evolutionary analogies in categorising AI in relation to biological organisms, balancing creative inspiration with scientific caution and (2) active, collaborative multidisciplinary engagement with addressing potential misinformation, recognising that biological narratives have sociopolitical implications that influence human interaction with machines.

Read the full article at: link.springer.com

Debates about artificial intelligence (AI) tend to revolve around whether large models are intelligent, autonomous agents. Some AI researchers and commentators speculate that we are on the cusp of creating agents with artificial general intelligence (AGI), a prospect anticipated with both elation and anxiety. There have also been extensive conversations about cultural and social consequences of large models, orbiting around two foci: immediate effects of these systems as they are currently used, and hypothetical futures when these systems turn into AGI agents—perhaps even superintelligent AGI agents. But this discourse about large models as intelligent agents is fundamentally misconceived. Combining ideas from social and behavioral sciences with computer science can help us to understand AI systems more accurately. Large models should not be viewed primarily as intelligent agents but as a new kind of cultural and social technology, allowing humans to take advantage of information other humans have accumulated.

HENRY FARRELL, ALISON GOPNIK, COSMA SHALIZI, AND JAMES EVANS Authors Info & Affiliations
SCIENCE 13 Mar 2025 Vol 387, Issue 6739

Read the full article at: www.science.org

Ricard Solé, Manlio De Domenico

The path toward the emergence of life in our biosphere involved several key events allowing for the persistence, reproduction and evolution of molecular systems. All these processes took place in a given environmental context and required both molecular diversity and the right non-equilibrium conditions to sustain and favour complex self-sustaining molecular networks capable of evolving by natural selection. Life is a process that departs from non-life in several ways and cannot be reduced to standard chemical reactions. Moreover, achieving higher levels of complexity required the emergence of novelties. How did that happen? Here, we review different case studies associated with the early origins of life in terms of phase transitions and bifurcations, using symmetry breaking and percolation as two central components. We discuss simple models that allow for understanding key steps regarding life origins, such as molecular chirality, the transition to the first replicators and cooperators, the problem of error thresholds and information loss, and the potential for "order for free" as the basis for the emergence of life.

Read the full article at: arxiv.org

Thomas F. Varley, Pedro A. M. Mediano, Alice Patania, Josh Bongard

The study of irreducible higher-order interactions has become a core topic of study in complex systems. Two of the most well-developed frameworks, topological data analysis and multivariate information theory, aim to provide formal tools for identifying higher-order interactions in empirical data. Despite similar aims, however, these two approaches are built on markedly different mathematical foundations and have been developed largely in parallel. In this study, we present a head-to-head comparison of topological data analysis and information-theoretic approaches to describing higher-order interactions in multivariate data; with the aim of assessing the similarities and differences between how the frameworks define ``higher-order structures." We begin with toy examples with known topologies, before turning to naturalistic data: fMRI signals collected from the human brain. We find that intrinsic, higher-order synergistic information is associated with three-dimensional cavities in a point cloud: shapes such as spheres are synergy-dominated. In fMRI data, we find strong correlations between synergistic information and both the number and size of three-dimensional cavities. Furthermore, we find that dimensionality reduction techniques such as PCA preferentially represent higher-order redundancies, and largely fail to preserve both higher-order information and topological structure, suggesting that common manifold-based approaches to studying high-dimensional data are systematically failing to identify important features of the data. These results point towards the possibility of developing a rich theory of higher-order interactions that spans topological and information-theoretic approaches while simultaneously highlighting the profound limitations of more conventional methods.

Read the full article at: arxiv.org

Takaya Arita, Wenxian Zheng, Reiji Suzuki, Fuminori Akiba

This study explored how large language models (LLMs) perform in two areas related to art: writing critiques of artworks and reasoning about mental states (Theory of Mind, or ToM) in art-related situations. For the critique generation part, we built a system that combines Noel Carroll's evaluative framework with a broad selection of art criticism theories. The model was prompted to first write a full-length critique and then shorter, more coherent versions using a step-by-step prompting process. These AI-generated critiques were then compared with those written by human experts in a Turing test-style evaluation. In many cases, human subjects had difficulty telling which was which, and the results suggest that LLMs can produce critiques that are not only plausible in style but also rich in interpretation, as long as they are carefully guided. In the second part, we introduced new simple ToM tasks based on situations involving interpretation, emotion, and moral tension, which can appear in the context of art. These go beyond standard false-belief tests and allow for more complex, socially embedded forms of reasoning. We tested 41 recent LLMs and found that their performance varied across tasks and models. In particular, tasks that involved affective or ambiguous situations tended to reveal clearer differences. Taken together, these results help clarify how LLMs respond to complex interpretative challenges, revealing both their cognitive limitations and potential. While our findings do not directly contradict the so-called Generative AI Paradox--the idea that LLMs can produce expert-like output without genuine understanding--they suggest that, depending on how LLMs are instructed, such as through carefully designed prompts, these models may begin to show behaviors that resemble understanding more closely than we might assume.

Read the full article at: arxiv.org

Aaron Clauset, Barbara F. Walter, Lars-Erik Cederman, Kristian Skrede Gleditsch

Although very large wars remain an enduring threat in global politics, we lack a clear understanding of how some wars become large and costly, while most do not. There are three possibilities: large conflicts start with and maintain intense fighting, they persist over a long duration, or they escalate in intensity over time. Using detailed within-conflict data on civil and interstate wars 1946--2008, we show that escalation dynamics -- variations in fighting intensity within an armed conflict -- play a fundamental role in producing large conflicts and are a generic feature of both civil and interstate wars. However, civil wars tend to deescalate when they become very large, limiting their overall severity, while interstate wars exhibit a persistent risk of continual escalation. A non-parametric model demonstrates that this distinction in escalation dynamics can explain the differences in the historical sizes of civil vs. interstate wars, and explain Richardson's Law governing the frequency and severity of interstate conflicts over the past 200 years. Escalation dynamics also drive enormous uncertainty in forecasting the eventual sizes of both hypothetical and ongoing civil wars, indicating a need to better understand the causes of escalation and deescalation within conflicts. The close relationship between the size, and hence the cost, of an armed conflict and its potential for escalation has broad implications for theories of conflict onset or termination and for risk assessment in international relations.

Read the full article at: arxiv.org

Stephen Polasky, Marten Scheffer, and John M. Anderies

122 (14) e2320528122

A well-functioning society requires well-functioning institutions that ensure prosperity, fair distribution of wealth, social participation, security, and informative media. Such institutions are built on a foundation of trust. However, while trust is essential for economic success and good governance, interconnected mechanisms inherent in weakly governed market economies tend to undermine the very trust on which such success depends. These mechanisms include the intrinsic tendency for inequality to grow, media to boost perceived unfairness, and self-interest to gain rewards at the expense of others. These mechanisms, if left unchecked, allow wealth concentration to result in state capture where institutions facilitate further wealth concentration instead of the promoting the common good. As a result, people may become alienated and untrusting of fellow citizens and of institutions. Several democracies now experience such dynamics, the United States being a prime example. We discuss ways in which well-functioning democracies can design institutions to help avoid this social trap, and the much harder challenge of escaping the trap once in it. Successful cases such as the ability of Scandinavian democracies to maintain high-trust, and the US progressive era in the early 20th century provide instructive examples.

Read the full article at: www.pnas.org