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, in
Political bias in media plays a critical role in shaping public opinion, voter behaviour, and broader democratic discourse. Subjective opinions and political bias can be found in media sources, such as newspapers, depending on their funding mechanisms and alliances with political parties. Automating the detection of political biases in media content can limit biases in elections. The impact of large language models (LLMs) in politics and media studies is becoming prominent. In this study, we utilise LLMs to compare the left-wing, right-wing, and neutral political opinions expressed in the Guardian and BBC. We review newspaper reporting that includes significant events such as the Russia-Ukraine war and the Hamas-Israel conflict. We analyse the proportion for each opinion to find the bias under different LLMs, including BERT, Gemini, and DeepSeek. Our results show that after the outbreak of the wars, the political bias of Western media shifts towards the left-wing and each LLM gives a different result. DeepSeek consistently showed a stable Left-leaning tendency, while BERT and Gemini remained closer to the Centre. The BBC and The Guardian showed distinct reporting behaviours across
We present Territory Paint Wars, a minimal competitive multi-agent reinforcement learning environment implemented in Unity, and use it to systematically investigate failure modes of Proximal Policy Optimisation (PPO) under self-play. A first agent trained for $84{,}000$ episodes achieves only $26.8\%$ win rate against a uniformly-random opponent in a symmetric zero-sum game. Through controlled ablations we identify five implementation-level failure modes -- reward-scale imbalance, missing terminal signal, ineffective long-horizon credit assignment, unnormalised observations, and incorrect win detection -- each of which contributes critically to this failure in this setting. After correcting these issues, we uncover a distinct emergent pathology: competitive overfitting, where co-adapting agents maintain stable self-play performance while generalisation win rate collapses from $73.5\%$ to $21.6\%$. Critically, this failure is undetectable via standard self-play metrics: both agents co-adapt equally, so the self-play win rate remains near $50\%$ throughout the collapse. We propose a minimal intervention -- opponent mixing, where $20\%$ of training episodes substitute a fixed uniforml
Europe's largest bank by assets as of 2025 started out in the 1860s in one of Europe's colonies: The Hongkong and Shanghai Banking Co (HSBC). Multiple wars forced Qing China and later the young Republic of China into a series of unequal treaties, one of which was the forced legalisation of the opium trade from parts of the British Empire into China, another was opening several cities, including Shanghai, for trade and granting extensive civil, property and business rights to non-residents and yet another was the annexation of Hong Kong by the United Kingdom. These are the conditions that created HSBC and in which it thrived, including from opium-related profits. During periods of relative calm, the bank grew geographically and made profits -- whether in moral or unethical, whether in legal or unlawful ways -- which helped HSBC weather the storms of civil and world wars. Other aspects contributed to HSBC's survival and success, such as its global nature, which allowed it to diversify and shift away from regions when danger emerges there and find shelter in safer havens. Yet the resilient survival abilities and the financial success of HSBC until 1950 should not distract from the fac
The Ricardian model of world trade based on comparative advantage is not sufficient to justify equal trade relations.The existing model of trade relations does not explain the distribution of income among trading countries. This paper presents a method for building equitable trade relations. Its essence is to present an algorithm for building such trade relations, based on the previously proposed model of world trade, that the trade balance of each country would be equal to zero. Under such conditions, tariff wars would become impossible. It is proved that, provided that the supply structure is consistent with the demand structure, it is always possible to build an equilibrium price vector for which the trade balance of each country is zero. This state of economic equilibrium is called ideal. The article presents an algorithm to build an export structure based on the structure of imports. This algorithm is quite simple and allows for a wide range of applications. Under fairly simple realistic assumptions about the behaviour of countries trading with each other that are subject to tariff restrictions, it is proved that this leads to an increase in the prices of the goods traded by t
Trade imbalances significantly alter the welfare implications of tariffs. Using an illustrative model, we show that trade deficits enhance a country's ability to alter its terms of trade, and thereby benefit from tariffs. Greater trade deficits imply higher optimal, or welfare maximizing, tariffs. We compute optimal unilateral and Nash equilibrium tariffs between the United States and China $\unicode{x2014}$ the countries with the largest bilateral trade imbalance $\unicode{x2014}$ using a multi-region, multi-sector applied general equilibrium model with service sectors and input-output linkages, a computationally complex task. Free trade benefits both countries compared to a trade war. Relative to existing tariff rates, however, the United States gains from a trade war with China $\unicode{x2014}$ a result that hinges on their bilateral trade imbalance.
This paper presents a new symbolic corpus of musical themes from the complete Star Wars trilogies (Episodes I-IX) by John Williams. The corpus files are made available in multiple formats (.krn, .sib, and .musicxml) and include melodic, harmonic, and formal information. The Star Wars Thematic Corpus (SWTC) contains a total of 64 distinctive, recurring, and symbolically meaningful themes and motifs, commonly referred to as leitmotifs. Through this corpus we also introduce a new humdrum standard for non-functional harmony encodings, **harte, based on Harte (2005, 2010). This report details the motivation, describes the transcription and encoding processes, and provides some brief summary statistics. While relatively small in scale, the SWTC represents a unified collection from one of the most prolific and influential composers of the 20th century, and the under-studied subset of film and multimedia musical material in general. We hope the SWTC will provide insights into John Williams' compositional style, as well as prove useful in comparisons against other thematic corpora from film and beyond.
According to physicist David Mermin, the science wars was a series of exchanges between scientists and "sociologists, historians, and literary critics" whom the scientists thought to be "ludicrously ignorant of science, making all kinds of nonsensical pronouncements." The science wars peaked in 1996 with the publication of a hoax article by physicist Alan Sokal in a cultural studies journal and then seemed mercifully to die away by the end of the 1990's. I recently noticed, however, that the kerfuffle has persisted. This motivated me to pen this essay and point out the silliness of the entire affair.
Even before 9/11 there were claims that the nature of war had changed fundamentally. The 9/11 attacks created an urgent need to understand contemporary wars and their relationship to older conventional and terrorist wars, both of which exhibit remarkable regularities. The frequency-intensity distribution of fatalities in "old wars", 1816-1980, is a power-law with exponent 1.80. Global terrorist attacks, 1968-present, also follow a power-law with exponent 1.71 for G7 countries and 2.5 for non-G7 countries. Here we analyze two ongoing, high-profile wars on opposite sides of the globe - Colombia and Iraq. Our analysis uses our own unique dataset for killings and injuries in Colombia, plus publicly available data for civilians killed in Iraq. We show strong evidence for power-law behavior within each war. Despite substantial differences in contexts and data coverage, the power-law coefficients for both wars are tending toward 2.5, which is a value characteristic of non-G7 terrorism as opposed to old wars. We propose a plausible yet analytically-solvable model of modern insurgent warfare, which can explain these observations.
Quark Wars is an all-new, adventure style game. We recommended playing it outdoors. Quark Wars is modeled upon the outdoor game called Hungarian Number War, with notable influence from Star Wars, the American epic space saga. The players form two opposing teams. Both teams elect their own leader. The team members and their leaders wear particle war bonnets on their foreheads. These headdresses consist of three or four cards indicating combinations of elementary particles. The two teams compete by identifying (reading out loudly) the elementary particle cards on the foreheads of their adversaries. Players are allowed to use the terrain to cover their particle identity on their foreheads and may try to hide, run, or band together in a group to win. Quark Wars was tested at a Summer Camp of Berze Science Club in Hungary. Students loved playing Quark Wars, as this game resulted in lots of hilarity and action. In addition, Quark Wars also solidified particle terminology and made the concept of particle identification and discovery more tangible to secondary/middle school students.
Can we avoid wars at the crossroads of history? This question has been pursued by individuals, scholars, policymakers, and organizations throughout human history. In this research, we attempt to answer the question based on the recent advances of Artificial Intelligence (AI) and Large Language Models (LLMs). We propose \textbf{WarAgent}, an LLM-powered multi-agent AI system, to simulate the participating countries, their decisions, and the consequences, in historical international conflicts, including the World War I (WWI), the World War II (WWII), and the Warring States Period (WSP) in Ancient China. By evaluating the simulation effectiveness, we examine the advancements and limitations of cutting-edge AI systems' abilities in studying complex collective human behaviors such as international conflicts under diverse settings. In these simulations, the emergent interactions among agents also offer a novel perspective for examining the triggers and conditions that lead to war. Our findings offer data-driven and AI-augmented insights that can redefine how we approach conflict resolution and peacekeeping strategies. The implications stretch beyond historical analysis, offering a bluepr
It is known that many Twitter users are bots, which are accounts controlled and sometimes created by computers. Twitter bots can send spam tweets, manipulate public opinion and be used for online fraud. Here we report the discovery, retrieval, and analysis of the `Star Wars' botnet in Twitter, which consists of more than 350,000 bots tweeting random quotations exclusively from Star Wars novels. The botnet contains a single type of bot, showing exactly the same properties throughout the botnet. It is unusually large, many times larger than other available datasets. It provides a valuable source of ground truth for research on Twitter bots. We analysed and revealed rich details on how the botnet was designed and created. As of this writing, the Star Wars bots are still alive in Twitter. They have survived since their creation in 2013, despite the increasing efforts in recent years to detect and remove Twitter bots.We also reflect on the `unconventional' way in which we discovered the Star Wars bots, and discuss the current problems and future challenges of Twitter bot detection.
We study several variants of the classical card game war. As anyone who played this game knows, the game can take some time to terminate, but it usually does. Here, we analyze a number of asymptotic variants of the game, where the number of cards is $n$, and show that all have expected termination time of order $n^2$. This is the same expected termination time as in the game where at each turn a fair coin toss decides which player wins a card, known as Gambler's Ruin and studied by Pascal, Fermat and others in the seventeenth century.
In this paper I propose a 'mechanism' for the explanation of power-law characteristics of casualty dynamics in inter-state wars, intra-state wars and terrorist attacks: the scale-free physical organization of social systems. Other explanations - self-organized criticality (Cederman, 2003) and the redistribution of total attack capabilities (Johnson et al. 2006) - do not provide a consistent framework for the power-law characteristics of casualty dynamics. The development in time of the power-law characteristics of casualty dynamics during wars and conflicts provides clues for the 'functioning' of social systems which are targeted, and/or for the (in)effectiveness and strategies of actors using force (violence) against these social systems.
We investigate the role of networks of alliances in preventing (multilateral) interstate wars. We first show that, in the absence of international trade, no network of alliances is peaceful and stable. We then show that international trade induces peaceful and stable networks: trade increases the density of alliances so that countries are less vulnerable to attack and also reduces countries' incentives to attack an ally. We present historical data on wars and trade, noting that the dramatic drop in interstate wars since 1950, and accompanying densification and stabilization of alliances, are consistent with the model but not other prominent theories.
Lewis Fry Richardson argued that the frequency and severity of deadly conflicts of all kinds, from homicides to interstate wars and everything in between, followed universal statistical patterns: their frequency followed a simple Poisson arrival process and their severity followed a simple power-law distribution. Although his methods and data in the mid-20th century were neither rigorous nor comprehensive, his insights about violent conflicts have endured. In this chapter, using modern statistical methods and data, we show that Richardson's original claims appear largely correct, with a few caveats. These facts place important constraints on our understanding of the underlying mechanisms that produce individual wars and periods of peace, and shed light on the persistent debate about trends in conflict.
We report a remarkable universality in the patterns of violence arising in three high-profile ongoing wars, and in global terrorism. Our results suggest that these quite different conflict arenas currently feature a common type of enemy, i.e. the various insurgent forces are beginning to operate in a similar way regardless of their underlying ideologies, motivations and the terrain in which they operate. We provide a microscopic theory to explain our main observations. This theory treats the insurgent force as a generic, self-organizing system which is dynamically evolving through the continual coalescence and fragmentation of its constituent groups.
Long-horizon rollout generation has become the dominant systems bottleneck in agentic reinforcement learning (RL). As agents interact with environments over many turns, trajectories rapidly grow to tens of thousands of tokens, making synchronous RL training increasingly constrained by rollout. We propose WAR, a workload-aware rollout system that substantially accelerates synchronous agentic RL by jointly optimizing decoding and scheduling. WAR is built on a key observation: the optimal rollout optimization strategy depends on runtime load: (1) Under low load, WAR enables model-free speculative decoding with SuffixDecoding, which reuses suffix patterns from previously completed trajectories as speculative drafts for future rollouts. Unlike model-based drafters, SuffixDecoding introduces no additional draft model and avoids GPU contention with rollout generation. (2) Under high load, where saturated batched decoding leaves limited room for speculative speedup, WAR shifts the optimization focus to cache-aware scheduling. A global scheduler places requests across rollout replicas based on cache locality, trajectory progress and server load, reducing redundant KV-cache recomputation and
The rise of generative AI (genAI) can transform the representation of different aspects of social reality, including modern wars. While scholarship has largely focused on the military applications of AI, the growing adoption of genAI technologies may have major implications for how wars are portrayed, remembered, and interpreted. A few initial scholarly inquiries highlight the risks of genAI in this context, specifically regarding its potential to distort the representation of mass violence, particularly by sanitising and homogenising it. However, little is known about how genAI representation practices vary between different episodes of violence portrayed by Western and non-Western genAI models. Using the Russian aggression against Ukraine as a case study, we audit how two image-generative models, the US-based Midjourney and the Russia-based Kandinsky, represent both fictional and factual episodes of the war. We then analyse the models' responsiveness to the war-related prompts, together with the aesthetic and content-based aspects of the resulting images. Our findings highlight that contextual factors lead to variation in the representation of war, both between models and within
Fog of War chess is a popular variant of classical chess, in which both players have only partial information about the position of the opponent's pieces. This study provides the first theoretical analysis of endgames in Fog of War chess. In particular, we analyze the setups king and queen versus king, king and rook versus king, and king and two rooks versus king. We show that a king and queen can always guarantee a win against a lone king. In contrast to classical chess, a king and a rook cannot guarantee a win against a lone king. However, adding one more rook guarantees a win.