Although artificial intelligence (AI) has become increasingly smart, its wisdom has not kept pace. In this opinion article, we examine what is known about human wisdom and sketch a vision of its AI counterpart. We introduce human wisdom as strategies for solving intractable problems-those outside the scope of analytic techniques-including both 'object-level' strategies, such as heuristics (for managing problems), and 'metacognitive' strategies, such as intellectual humility, perspective-taking, or context adaptability (for managing object-level task fit). We argue that AI systems particularly struggle with this type of metacognition. Wise metacognition would lead to AI that is more robust to novel environments, explainable to users, cooperative with others, and safer by risking fewer misaligned goals with human users. We discuss how wise AI might be benchmarked, trained, and implemented.

Online behavioural research faces a growing methodological and epistemic threat as participants increasingly rely on large language models: LLM Pollution. Amid accumulating empirical evidence of contamination, we introduce a conceptual framework that distinguishes three variants — Partial LLM Mediation, Full LLM Delegation, and LLM Spillover. Their interaction distorts samples, biases inferences, and fuels an escalating methodological arms race. We outline mitigation strategies spanning researcher practices, platform accountability, and community adaptation.

Content exposure online is increasingly shaped by socio-technical systems, where human and algorithmic factors interact. TikTok exemplifies strong algorithm-driven curation, tailoring the stream of content towards users’ interests based on their behavior on the platform. Despite growing concerns about TikTok’s societal impact, the dynamics of content amplification at the individual level remain largely unquantified. How quickly, and to what extent, does TikTok amplify content aligned with users’ interests? To address these questions, we conduct a sock-puppet audit, deploying bots with different interests to engage with TikTok’s “For You” feed. Our findings reveal that content aligned with the bots’ interests undergoes strong amplification, with rapid reinforcement typically occurring within the first 200 videos watched. While amplification is consistently observed across all interests, its intensity and dynamics vary, as revealed by time series analyses and Markov models. Although TikTok’s algorithm preserves some content diversity, we find a strong negative correlation between amplification and exploration: as the amplification of interest-aligned content increases, the diversity of encountered hashtags declines sharply. Our findings contribute to discussions on socio-algorithmic feedback loops in the digital age and the trade-offs between personalization and content diversity.

The substantial literature on charitable giving–a prototypical pro-social behavior–commonly faces constraints on sample size, diversity, or experimental design, resulting ina fragmented understanding of factors that influence charitable giving and theirinteractions. We developed a game that co-varies multiple factors that influencegiving—akin to running thousands of experiments. Over 2.7 million decisions, someincentivized, from 257,000 participants in 203 countries were collected. Helping at least 3or 6 strangers, on average, surpassed even the strongest effects of giving to oneself or arelative, respectively. This finding holds cross-culturally. Further, we found substantialheterogeneity from experimental stimuli in all main factor effects (e.g., identifiable victimeffect), often reversing direction. Our findings paint a more comprehensive and promisingpicture of human prosociality.

Abstract Intelligent machines have the potential to uncover problem-solving strategies beyond human discovery. Emerging evidence from competitive gameplay, such as Go and chess, demonstrates that AI systems are evolving from mere tools to sources of cultural innovation adopted by humans. However, the conditions under which intelligent machines transition from tools to drivers of persistent cultural change remain unclear. We identify three key dimensions that modulate machine influence on human problem-solving: the discovered strategies must be non-trivial, learnable, and offer a clear advantage. Using a cultural transmission experiment, we demonstrate that when these conditions are met, machine-discovered strategies can be transmitted, understood, and preserved by human populations, leading to enduring cultural shifts. Conversely, using agent-based simulations, we show how machine influence is constrained in the absence of these conditions. These findings provide a framework for understanding how machines can persistently expand human cognitive skills and underscore the need to consider their broader implications for human cognition and cultural evolution.

Artificial intelligence (AI) systems are rapidly becoming more capable, autonomous, and deeply embedded in social life. As humans increasingly interact, cooperate, and compete with AI, we move from purely human societies to hybrid human-AI societies whose collective dynamics cannot be captured by existing behavioural models alone. Drawing on evolutionary game theory, cultural evolution, and Large Language Models (LLMs) powered simulations, we argue that these developments open a new research agenda for social physics centred on the co-evolution of humans and machines. We outline six key research directions. First, modelling the evolutionary dynamics of social behaviours (e.g. cooperation, fairness, trust) in hybrid human-AI populations. Second, understanding machine culture: how AI systems generate, mediate, and select cultural traits. Third, analysing the co-evolution of language and behaviour when LLMs frame and participate in decisions. Fourth, studying the evolution of AI delegation: how responsibilities and control are negotiated between humans and machines. Fifth, formalising and comparing the distinct epistemic pipelines that generate human and AI behaviour. Sixth, modelling the co-evolution of AI development and regulation in a strategic ecosystem of firms, users, and institutions. Together, these directions define a programme for using social physics to anticipate and steer the societal impact of advanced AI.

Generative AI's novel capacities raise questions about the future role of human expertise: does it level the playing field between professional artists and laypeople, or does expertise enhance AI use? Do visual art skills transfer to these tools? This pre-registered study experimentally compares 50 active professional visual artists with a matched laypeople sample. Two tasks assessed copying accuracy and creative thinking, using a bespoke platform powered by text-to-image AI. Results show expertise transfer: artists produced more accurate copies and more divergent ideas. This advantage remained small but detectable even in this tightly controlled experimental setting. We also explored how well a vision-capable LLM (GPT-4o) would fare: on par in copying and slightly better on average than artists in creativity, although never above best humans. These findings highlight the importance of integrating artistic skills with AI, suggesting a potential for collaborative synergy that could reshape creative industries and arts education.

Current approaches to instill prosociality in large language model (LLM) agents often rely on humans specifying desired behaviors at the individual level, which does not guarantee cooperation within LLM populations. As frontier training shifts toward individual rewards for verifiable tasks, such as mathematics and coding, this outcome-based focus may further undermine cooperation in multi-agent settings. Large-scale cooperation in human populations emerged via unguided evolutionary mechanisms, not a central architect. Group selection, in which cooperative groups within a population outcompete less cooperative ones, has been argued to be essential. In this study, we explore whether group selection can promote cooperation in populations of LLM agents. We introduce a multi-agent simulation framework in which LLM agents play a repeated social dilemma game and transmit their natural-language prompts across generations under either individual- or group-level selection. Under group selection, prompts from high-performing groups are transmitted, thereby promoting prosociality and stabilizing cooperation. Under individual selection, self-interested prompts dominate, causing populations to collapse into collective defection. This gap is robust across prompt ablations, alternative game framings, and model swaps. We theoretically reproduce key results using a replicator-mutator model, whose empirical transmission kernel predicts a phase transition at a critical threshold. Preliminary findings show that, when informed about the selection mechanism, GPT-5.4 preemptively and gradually adjusts first-generation donations. This demonstrates strong anticipatory behavior that was not observed in the other tested models. These results demonstrate that prosocial prompts and cooperative behaviors evolve in LLM agent populations under group selection.

Reliance on generative AI can reduce cultural variance and diversity, especially in creative work. This reduction in variance has already led to problems in model performance, including model collapse and hallucination. In this paper, we examine the long-term consequences of AI use for human cultural evolution and the conditions under which widespread AI use may lead to "cultural collapse", a process in which reliance on AI-generated content reduces human variation and innovation and slows cumulative cultural evolution. Using an agent-based model and evolutionary game theory, we compare two types of AI use: complement and substitute. AI-complement users seek suggestions and guidance while remaining the main producers of the final output, whereas AI-substitute users provide minimal input, and rely on AI to produce most of the output. We then study how these use strategies compete and spread under evolutionary dynamics. We find that AI-substitute users prevail under individual-level selection despite the stronger reduction in cultural variance. By contrast, AI-complement users can benefit their groups by maintaining the variance needed for exploration, and can therefore be favored under cultural group selection when group boundaries are strong. Overall, our findings shed light on the long-term, population-level effects of AI adoption and inform policy and organizational strategies to mitigate these risks.

We propose a society-first theory of normative appropriateness where individuals, modeled as pre-trained actors with cognitive architectures analogous to Large Language Models (LLMs), generate behavior via predictive pattern completion. Our theory posits that individuals act by completing distributed symbolic patterns based on context, answering questions such as "What does a person such as I do in a situation such as this?". This sense-making mechanism provides a parsimonious account of the key features of human norms: their context-dependence, arbitrariness, automaticity, dynamism, and their support from social sanctioning. It challenges rational-choice theories of social norms by accounting for their key features without needing to exogenously posit scalar rewards or preference relations. By distinguishing between explicit norms, which we associate with in-context adaptation, and implicit norms, which we associate with long-term memory, the theory reconceptualizes several foundational ideas in cognitive science. In particular, it gives an alternative account to the data traditionally seen as supporting dual-process models, and it flips the role of rationality, allowing us to construe it as adherence to culturally-contingent justification standards.

Large language models offer new opportunities for behavioural science, but their rapid evolution poses challenges for research rigour. We introduce a consensus-based reporting checklist to improve transparency, reproducibility and ethical accountability of large-language-model-based research in the behavioural sciences.

Scientific discovery is constrained not only by what is true, but by what is cognitively available to the researchers currently exploring a field. Many directions are coherent in light of the literature yet unlikely to be proposed because no existing community occupies the right combination of concepts, methods, and intuitions. Modern language models inherit this bias, recombining high-density regions of the literature when prompted for novel ideas. We introduce a framework that targets the complementary region, which we call the alien space of science, where directions are plausible under the structure of existing knowledge but unlikely under the distribution of existing researchers. Our method first decomposes papers into granular conceptual units and clusters them into a shared vocabulary of idea atoms. It then learns two complementary models over this vocabulary. A coherence model scores whether a combination of atoms forms a viable research direction, and an availability model scores whether any existing author community is positioned to produce a given combination. Sampling alien directions then reduces to ranking atom combinations that maximize coherence while minimizing availability. On a corpus of 16,068 peer-reviewed LLM papers from NeurIPS, ICLR, ICML, and major NLP venues, the resulting sampler explores a 3.5 - 7 x broader effective atom vocabulary than frontier LLM ideation baselines without sacrificing coherence, and produces ideas that match or exceed those baselines under blind LLM, human, and downstream experimental evaluation. By separating scientific plausibility from community availability, our framework points toward AI ideation that complements rather than merely accelerates human science, expanding exploration into coherent directions that the current community may overlook.

The rapid development of advanced AI agents and the imminent deployment of many instances of these agents will give rise to multi-agent systems of unprecedented complexity. These systems pose novel and under-explored risks. In this report, we provide a structured taxonomy of these risks by identifying three key failure modes (miscoordination, conflict, and collusion) based on agents' incentives, as well as seven key risk factors (information asymmetries, network effects, selection pressures, destabilising dynamics, commitment problems, emergent agency, and multi-agent security) that can underpin them. We highlight several important instances of each risk, as well as promising directions to help mitigate them. By anchoring our analysis in a range of real-world examples and experimental evidence, we illustrate the distinct challenges posed by multi-agent systems and their implications for the safety, governance, and ethics of advanced AI.

Recent evidence suggests that delegating tasks to machines can facilitate unethical behavior, but the psychological mechanisms driving this effect are not yet well understood. This study investigates whether two interventions can mitigate cheating in an algorithmic honesty game: transparency (information about which user input causes which algorithm behavior) and framing (natural language cues about the moral valence of behavior). In a 2 × 2 experimental design, we find that transparency does not reduce dishonest behavior, despite participants actively engaging with and understanding the provided information. Conversely, framing — replacing neutral labels like ”maximize profit” with ethically charged terms like ”maximize cheating” — substantially reduces dishonesty. These findings suggest that curbing misuse of AI requires confronting users with its moral implications, not just explaining the mechanics.

A growing body of multi-agent studies with Large Language Models (LLMs) explores how norms and cooperation emerge in mixed-motive scenarios, where pursuing individual gain can undermine the collective good. While prior work has explored these dynamics in both richly contextualized simulations and simplified game-theoretic environments, most LLM systems featuring common-pool resource (CPR) games provide agents with explicit reward functions directly tied to their actions. In contrast, human cooperation often emerges without full visibility into payoffs and population, relying instead on heuristics, communication, and punishment. We introduce a CPR simulation framework that removes explicit reward signals and embeds cultural-evolutionary mechanisms: social learning (adopting strategies and beliefs from successful peers) and norm-based punishment, grounded in Ostrom's principles of resource governance. Agents also individually learn from the consequences of harvesting, monitoring, and punishing via environmental feedback, enabling norms to emerge endogenously. We establish the validity of our simulation by reproducing key findings from existing studies on human behavior. Building on this, we examine norm evolution across a 2×2 grid of environmental and social initialisations (resource-rich vs. resource-scarce; altruistic vs. selfish) and benchmark how agentic societies comprised of different LLMs perform under these conditions. Our results reveal systematic model differences in sustaining cooperation and norm formation, positioning the framework as a rigorous testbed for studying emergent norms in mixed-motive LLM societies. Such analysis can inform the design of AI systems deployed in social and organizational contexts, where alignment with cooperative norms is critical for stability, fairness, and effective governance of AI-mediated environments.

The remarkable ecological success of humans is often attributed to our ability to develop complex cultural artefacts that enable us to cope with environmental challenges. The evolution of complex culture (cumulative cultural evolution) is usually modelled as a collective process in which individuals invent new artefacts (innovation) and copy information from others (social learning). This classic picture overlooks the growing role of intelligent algorithms in the digital age (e.g. search engines, recommender systems and large language models) in mediating information between humans, with potential consequences for cumulative cultural evolution. Building on a previous model, we investigate the combined effects of network-based social learning and a simplistic version of algorithmic mediation on cultural accumulation. We find that algorithmic mediation significantly impacts cultural accumulation and that this impact grows as social networks become less densely connected. Cultural accumulation is most effective when social learning and algorithmic mediation are combined, and the optimal ratio depends on the network's density. This work is an initial step towards formalizing the impact of intelligent algorithms on cumulative cultural evolution within an established framework. Models like ours provide insights into mechanisms of human-machine interaction in cultural contexts, guiding hypotheses for future experimental testing.

Intelligent machines with superhuman capabilities have the potential to uncover problem-solving strategies beyond human discovery. Emerging evidence from competitive gameplay, such as Go and chess, demonstrates that AI systems are evolving from mere tools to sources of cultural innovation adopted by humans. However, the conditions under which intelligent machines transition from tools to drivers of persistent cultural change remain unclear. We identify three key conditions for machines to fundamentally influence human problem-solving: the discovered strategies must be non-trivial, learnable, and offer a clear advantage. Using a cultural transmission experiment and an agent-based simulation, we demonstrate that when these conditions are met, machine-discovered strategies can be transmitted, understood, and preserved by human populations, leading to enduring cultural shifts. These findings provide a framework for understanding how machines can persistently expand human cognitive skills and underscore the need to consider their broader implications for human cognition and cultural evolution.

In open-ended domains like art, autonomous agents must generate ideas that are both original and internally coherent, yet current Large Language Models (LLMs) either default to familiar cultural patterns or sacrifice coherence when pushed toward novelty. We address this by introducing the Cultural Alien Sampler (CAS), a concept-selection method that explicitly separates compositional fit from cultural typicality. CAS uses two GPT-2 models fine-tuned on WikiArt concepts: a Concept Coherence Model that scores whether concepts plausibly co-occur within artworks, and a Cultural Context Model that estimates how typical those combinations are within individual artists' bodies of work. CAS targets combinations that are high in coherence and low in typicality, yielding ideas that maintain internal consistency while deviating from learned conventions and embedded cultural context. In a human evaluation (N = 100), our approach outperforms random selection and GPT-4o baselines and achieves performance comparable to human art students in both perceived originality and harmony. Additionally, a quantitative study shows that our method produces more diverse outputs and explores a broader conceptual space than its GPT-4o counterpart, demonstrating that artificial cultural alienness can unlock creative potential in autonomous agents.

Research on online news consumption has long focused on improving people’s ability to select accurate news sources. However, people who consume accurate information do not necessarily make an accurate collective. Reliable information aggregation and collective decision-making also benefit from independence between judgments. We tested whether collective incentives can improve collective accuracy by reducing judgment correlation, even when average individual accuracy remains unchanged. We define collective incentives as reward structures that tie an individual’s payoff to the accuracy of their group’s aggregated decision rather than their own personal accuracy. In an online experiment (N = 232), participants made forecasts after viewing realistic tweets from correlated or independent news sources. Shared tweets among group members were more accurate but fully correlated across group members; local tweets were less accurate but independent between participants. When rewarded for collective accuracy, participants used more independent local tweets, improving collective but not individual accuracy. Generalizing our findings, an exploratory analysis of public news datasets found a greater content correlation among larger publishers, modulated by accuracy. These findings highlight the overlooked importance of judgment independence alongside accuracy in creating thriving information ecologies online. Collective incentives may complement accuracy-based interventions by nudging news consumption towards diverse, local, and independent sources. We discuss how future research should implement collective incentives to improve collective decisions in real-world settings.

Online behavioural research faces an emerging threat as participants increasingly turn to large language models (LLMs) for advice, translation, or task delegation: LLM Pollution. We identify three interacting variants through which LLM Pollution threatens the validity and integrity of online behavioural research. First, Partial LLM Mediation occurs when participants make selective use of LLMs for specific aspects of a task, such as translation or wording support, leading researchers to (mis)interpret LLM-shaped outputs as human ones. Second, Full LLM Delegation arises when agentic LLMs complete studies with little to no human oversight, undermining the central premise of human-subject research at a more foundational level. Third, LLM Spillover signifies human participants altering their behaviour as they begin to anticipate LLM presence in online studies, even when none are involved. While Partial Mediation and Full Delegation form a continuum of increasing automation, LLM Spillover reflects second-order reactivity effects. Together, these variants interact and generate cascading distortions that compromise sample authenticity, introduce biases that are difficult to detect post hoc, and ultimately undermine the epistemic grounding of online research on human cognition and behaviour. Crucially, the threat of LLM Pollution is already co-evolving with advances in generative AI, creating an escalating methodological arms race. To address this, we propose a multi-layered response spanning researcher practices, platform accountability, and community efforts. As the challenge evolves, coordinated adaptation will be essential to safeguard methodological integrity and preserve the validity of online behavioural research.