Machine culture.

The ability of humans to create and disseminate culture is often credited as the single most important factor of our success as a species. In this Perspective, we explore the notion of 'machine culture', culture mediated or generated by machines. We argue that intelligent machines simultaneously transform the cultural evolutionary processes of variation, transmission and selection. Recommender algorithms are altering social learning dynamics. Chatbots are forming a new mode of cultural transmission, serving as cultural models. Furthermore, intelligent machines are evolving as contributors in generating cultural traits-from game strategies and visual art to scientific results. We provide a conceptual framework for studying the present and anticipated future impact of machines on cultural evolution, and present a research agenda for the study of machine culture.

Understanding hunter-gatherer cultural evolution needs network thinking.

Hunter-gatherers past and present live in complex societies, and the structure of these can be assessed using social networks. We outline how the integration of new evidence from cultural evolution experiments, computer simulations, ethnography, and archaeology open new research horizons to understand the role of social networks in cultural evolution.

Human cumulative culture and the exploitation of natural phenomena.

Cumulative cultural evolution (CCE)-defined as the process by which beneficial modifications are culturally transmitted and progressively accumulated over time-has long been argued to underlie the unparalleled diversity and complexity of human culture. In this paper, I argue that not just any kind of cultural accumulation will give rise to human-like culture. Rather, I suggest that human CCE depends on the gradual exploitation of natural phenomena, which are features of our environment that, through the laws of physics, chemistry or biology, generate reliable effects which can be exploited for a purpose. I argue that CCE comprises two distinct processes: optimizing cultural traits that exploit a given set of natural phenomena (Type I CCE) and expanding the set of natural phenomena we exploit (Type II CCE). I argue that the most critical features of human CCE, including its open-ended dynamic, stems from Type II CCE. Throughout the paper, I contrast the two processes and discuss their respective socio-cognitive requirements. This article is part of a discussion meeting issue 'The emergence of collective knowledge and cumulative culture in animals, humans and machines'.

Beyond social learning.

Cultural evolution requires the social transmission of information. For this reason, scholars have emphasized social learning when explaining how and why culture evolves. Yet cultural evolution results from many mechanisms operating in concert. Here, we argue that the emphasis on social learning has distracted scholars from appreciating both the full range of mechanisms contributing to cultural evolution and how interactions among those mechanisms and other factors affect the output of cultural evolution. We examine understudied mechanisms and other factors and call for a more inclusive programme of investigation that probes multiple levels of the organization, spanning the neural, cognitive-behavioural and populational levels. To guide our discussion, we focus on factors involved in three core topics of cultural evolution: the emergence of culture, the emergence of cumulative cultural evolution and the design of cultural traits. Studying mechanisms across levels can add explanatory power while revealing gaps and misconceptions in our knowledge. This article is part of the theme issue 'Foundations of cultural evolution'.

Technical reasoning alone does not take humans this far.

Although we see much utility in Osiurak and Reynaud's in-depth discussion on the role of what they term technical reasoning in cumulative culture, we argue that they neglect the time and energy costs that individuals would have to face to acquire skills in the absence of specific socio-cognitive abilities.

Cumulative Cultural Evolution within Evolving Population Structures.

Our species has the peculiar ability to accumulate cultural innovations over multiple generations, a phenomenon termed 'cumulative cultural evolution' (CCE). Recent years have seen a proliferation of empirical and theoretical work exploring the interplay between demography and CCE. This has generated intense discussion about whether demographic models can help explain historical patterns of cultural changes. Here, we synthesize empirical and theoretical studies from multiple fields to highlight how both population size and structure can shape the pool of cultural information that individuals can build upon to innovate, present the potential pathways through which humans' unique social structure might promote CCE, and discuss whether humans' social networks might partly result from selection pressures linked to our extensive reliance on culturally accumulated knowledge.

Competition for novelty reduces information sampling in a research game-a registered report.

Incentive structures shape scientists' research practices. One incentive in particular, rewarding priority of publication, is hypothesized to harm scientific reliability by promoting rushed, low-quality research. Here, we develop a laboratory experiment to test whether competition affects information sampling and guessing accuracy in a game that mirrors aspects of scientific investigation. In our experiment, individuals gather data in order to guess true states of the world and face a tradeoff between guessing quickly and increasing accuracy by acquiring more information. To test whether competition affects accuracy, we compare a treatment in which individuals are rewarded for each correct guess to a treatment where individuals face the possibility of being 'scooped' by a competitor. In a second set of conditions, we make information acquisition contingent on solving arithmetic problems to test whether competition increases individual effort (i.e. arithmetic-problem solving speed). We find that competition causes individuals to make guesses using less information, thereby reducing their accuracy (H1a and H1b confirmed). We find no evidence that competition increases individual effort (H2, inconclusive evidence). Our experiment provides proof of concept that rewarding priority of publication can incentivize individuals to acquire less information, producing lower-quality research as a consequence.

Causal understanding is not necessary for the improvement of culturally evolving technology.

Bows and arrows, houses and kayaks are just a few examples of the highly optimized tools that humans have produced and used to colonize new environments1,2. Because there is much evidence that humans' cognitive abilities are unparalleled3,4, many believe that such technologies resulted from our superior causal reasoning abilities5-7. However, others have stressed that the high dimensionality of human technologies makes them very difficult to understand causally8. Instead, they argue that optimized technologies emerge through the retention of small improvements across generations without requiring understanding of how these technologies work1,9. Here we show that a physical artefact becomes progressively optimized across generations of social learners in the absence of explicit causal understanding. Moreover, we find that the transmission of causal models across generations has no noticeable effect on the pace of cultural evolution. The reason is that participants do not spontaneously create multidimensional causal theories but, instead, mainly produce simplistic models related to a salient dimension. Finally, we show that the transmission of these inaccurate theories constrains learners' exploration and has downstream effects on their understanding. These results indicate that complex technologies need not result from enhanced causal reasoning but, instead, can emerge from the accumulation of improvements made across generations.

Social information can potentiate understanding despite inhibiting cognitive effort.

Both reasoning ability and social learning play a crucial role in human adaptation. Cognitive abilities like enhanced reasoning skills have combined with cumulative cultural adaptation to allow our species to dominate the world like no other. Thus, understanding how social learning interacts with individual reasoning ability is crucial for unravelling our evolutionary history. Here we describe a laboratory experiment designed to investigate the effect of social learning on individuals' ability to infer a general rule about unfamiliar problems. In this experiment, social information had both positive and negative effects on individuals' likelihood of inferring the rule. Social learners required more evidence to infer the rule than did individual learners, suggesting that social learning inhibits cognitive effort but social learning provided individuals with information that individual learners were unlikely to gather on their own, especially as the task became more difficult. When individuals are unlikely to discover useful information by themselves, social learning can potentiate understanding even though it reduces individual cognitive effort.

Divide and conquer: intermediate levels of population fragmentation maximize cultural accumulation.

Identifying the determinants of cumulative cultural evolution is a key issue in the interdisciplinary field of cultural evolution. A widely held view is that large and well-connected social networks facilitate cumulative cultural evolution because they promote the spread of useful cultural traits and prevent the loss of cultural knowledge through factors such as drift. This view stems from models that focus on the transmission of cultural information, without considering how new cultural traits actually arise. In this paper, we review the literature from various fields that suggest that, under some circumstances, increased connectedness can decrease cultural diversity and reduce innovation rates. Incorporating this idea into an agent-based model, we explore the effect of population fragmentation on cumulative culture and show that, for a given population size, there exists an intermediate level of population fragmentation that maximizes the rate of cumulative cultural evolution. This result is explained by the fact that fully connected, non-fragmented populations are able to maintain complex cultural traits but produce insufficient variation and so lack the cultural diversity required to produce highly complex cultural traits. Conversely, highly fragmented populations produce a variety of cultural traits but cannot maintain complex ones. In populations with intermediate levels of fragmentation, cultural loss and cultural diversity are balanced in a way that maximizes cultural complexity. Our results suggest that population structure needs to be taken into account when investigating the relationship between demography and cumulative culture.This article is part of the theme issue 'Bridging cultural gaps: interdisciplinary studies in human cultural evolution'.

Partial connectivity increases cultural accumulation within groups.

Complex technologies used in most human societies are beyond the inventive capacities of individuals. Instead, they result from a cumulative process in which innovations are gradually added to existing cultural traits across many generations. Recent work suggests that a population's ability to develop complex technologies is positively affected by its size and connectedness. Here, we present a simple computer-based experiment that compares the accumulation of innovations by fully and partially connected groups of the same size in a complex fitness landscape. We find that the propensity to learn from successful individuals drastically reduces cultural diversity within fully connected groups. In comparison, partially connected groups produce more diverse solutions, and this diversity allows them to develop complex solutions that are never produced in fully connected groups. These results suggest that explanations of ancestral patterns of cultural complexity may need to consider levels of population fragmentation and interaction patterns between partially isolated groups.

The foundations of the human cultural niche.

Technological innovations have allowed humans to settle in habitats for which they are poorly suited biologically. However, our understanding of how humans produce complex technologies is limited. We used a computer-based experiment, involving humans and learning bots, to investigate how reasoning abilities, social learning mechanisms and population structure affect the production of virtual artefacts. We found that humans' reasoning abilities play an important role in the production of innovations, but that groups of individuals are able to produce artefacts that are more complex than any isolated individual can produce during the same amount of time. We show that this group-level ability to produce complex innovations is maximized when social information is easy to acquire and when individuals are organized into large and partially connected populations. These results suggest that the transition to behavioural modernity could have been triggered by a change in ancestral between-group interaction patterns.

Social learning and the replication process: an experimental investigation.

Human cultural traits typically result from a gradual process that has been described as analogous to biological evolution. This observation has led pioneering scholars to draw inspiration from population genetics to develop a rigorous and successful theoretical framework of cultural evolution. Social learning, the mechanism allowing information to be transmitted between individuals, has thus been described as a simple replication mechanism. Although useful, the extent to which this idealization appropriately describes the actual social learning events has not been carefully assessed. Here, we used a specifically developed computer task to evaluate (i) the extent to which social learning leads to the replication of an observed behaviour and (ii) the consequences it has for fitness landscape exploration. Our results show that social learning does not lead to a dichotomous choice between disregarding and replicating social information. Rather, it appeared that individuals combine and transform information coming from multiple sources to produce new solutions. As a consequence, landscape exploration was promoted by the use of social information. These results invite us to rethink the way social learning is commonly modelled and could question the validity of predictions coming from models considering this process as replicative.

Experimental evidence for the influence of group size on cultural complexity.

The remarkable ecological and demographic success of humanity is largely attributed to our capacity for cumulative culture. The accumulation of beneficial cultural innovations across generations is puzzling because transmission events are generally imperfect, although there is large variance in fidelity. Events of perfect cultural transmission and innovations should be more frequent in a large population. As a consequence, a large population size may be a prerequisite for the evolution of cultural complexity, although anthropological studies have produced mixed results and empirical evidence is lacking. Here we use a dual-task computer game to show that cultural evolution strongly depends on population size, as players in larger groups maintained higher cultural complexity. We found that when group size increases, cultural knowledge is less deteriorated, improvements to existing cultural traits are more frequent, and cultural trait diversity is maintained more often. Our results demonstrate how changes in group size can generate both adaptive cultural evolution and maladaptive losses of culturally acquired skills. As humans live in habitats for which they are ill-suited without specific cultural adaptations, it suggests that, in our evolutionary past, group-size reduction may have exposed human societies to significant risks, including societal collapse.

Social learners require process information to outperform individual learners.

Humans exhibit a rich and complex material culture with no equivalent in animals. Also, social learning, a crucial requirement for culture, is particularly developed in humans and provides a means to accumulate knowledge over time and to develop advanced technologies. However, the type of social learning required for the evolution of this complex material culture is still debated. Here, using a complex and opaque virtual task, the efficiency of individual learning and two types of social learning (product-copying and process-copying) were compared. We found that (1) individuals from process-copying groups outperformed individuals from product-copying groups or individual learners, whereas access to product information was not a sufficient condition for providing an advantage to social learners compared to individual learners; (2) social learning did not seem to affect the exploration of the fitness landscape; (3) social learning led to strong within-group convergence and also to between-group convergence, and (4) individuals used widely variable social learning strategies. The implications of these results for cumulative culture evolution are discussed.