-
Machine Culture
Authors:
Levin Brinkmann,
Fabian Baumann,
Jean-François Bonnefon,
Maxime Derex,
Thomas F. Müller,
Anne-Marie Nussberger,
Agnieszka Czaplicka,
Alberto Acerbi,
Thomas L. Griffiths,
Joseph Henrich,
Joel Z. Leibo,
Richard McElreath,
Pierre-Yves Oudeyer,
Jonathan Stray,
Iyad Rahwan
Abstract:
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 algo…
▽ More
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.
△ Less
Submitted 22 November, 2023; v1 submitted 19 November, 2023;
originally announced November 2023.
-
Network Structure shapes the Impact of Diversity in Collective Learning
Authors:
Fabian Baumann,
Agnieszka Czaplicka,
Iyad Rahwan
Abstract:
It is widely believed that diversity arising from different skills enhances the performance of teams, and in particular, their ability to learn and innovate. However, diversity has also been associated with negative effects on the communication and coordination within collectives. Yet, despite the importance of diversity as a concept, we still lack a mechanistic understanding of how its impact is…
▽ More
It is widely believed that diversity arising from different skills enhances the performance of teams, and in particular, their ability to learn and innovate. However, diversity has also been associated with negative effects on the communication and coordination within collectives. Yet, despite the importance of diversity as a concept, we still lack a mechanistic understanding of how its impact is shaped by the underlying social network. To fill this gap, we model skill diversity within a simple model of collective learning and show that its effect on collective performance differs depending on the complexity of the task and the network density. In particular, we find that diversity consistently impairs performance in simple tasks. In contrast, in complex tasks, link density modifies the effect of diversity: while homogeneous populations outperform diverse ones in sparse networks, the opposite is true in dense networks, where diversity boosts collective performance. Our findings also provide insight on how to forge teams in an increasingly interconnected world: the more we are connected, the more we can benefit from diversity to solve complex problems.
△ Less
Submitted 30 June, 2023;
originally announced June 2023.
-
Biased-voter model: how persuasive a small group can be?
Authors:
Agnieszka Czaplicka,
Christos Charalambous,
Raul Toral,
Maxi San Miguel
Abstract:
We study the voter model dynamics in the presence of confidence and bias. We assume two types of voters. Unbiased voters whose confidence is indifferent to the state of the voter and biased voters whose confidence is biased towards a common fixed preferred state. We study the problem analytically on the complete graph using mean field theory and on an Erdős-Rényi random network topology using the…
▽ More
We study the voter model dynamics in the presence of confidence and bias. We assume two types of voters. Unbiased voters whose confidence is indifferent to the state of the voter and biased voters whose confidence is biased towards a common fixed preferred state. We study the problem analytically on the complete graph using mean field theory and on an Erdős-Rényi random network topology using the pair approximation, where we assume that the network of interactions topology is independent of the type of voters. We find that for the case of a random initial setup, and for sufficiently large number of voters $N$, the time to consensus increases proportionally to $\log(N)/γv$, with $γ$ the fraction of biased voters and $v$ the parameter quantifying the bias of the voters ($v=0$ no bias). We verify our analytical results through numerical simulations. We study this model on a biased-dependent topology of the network of interactions and examine two distinct, global average-degree preserving strategies (model I and model II) to obtain such biased-dependent random topologies starting from the biased-independent random topology case as the initial setup. Kee** all other parameters constant, in model I, $μ_{BU}$, the average number of links among biased (B) and unbiased (U) voters is varied at the expense of $μ_{UU}$ and $μ_{BB}$, i.e. the average number of links among only unbiased and biased voters respectively. In model II, $μ_{BU}$ is kept constant, while $μ_{BB}$ is varied at the expense of $μ_{UU}$. We find that if the agents follow the strategy described by model II, they can achieve a significant reduction in the time to reach consensus as well as an increment in the probability to reach consensus to the preferred state.
△ Less
Submitted 10 March, 2022;
originally announced March 2022.
-
From equality to diversity - bottom-up approach for hierarchy growth
Authors:
Agnieszka Czaplicka,
Janusz A. Hołyst
Abstract:
The hierarchical topology is a common property of many complex systems. Here we introduce a simple but generic model of hierarchy growth from the bottom to the top. Therein, two dynamical processes are accounted for: agent's promotions to next hierarchy levels when local speakers are elected and followed by other agents and agent's degradations to the lowest hierarchy. Following the initial stage…
▽ More
The hierarchical topology is a common property of many complex systems. Here we introduce a simple but generic model of hierarchy growth from the bottom to the top. Therein, two dynamical processes are accounted for: agent's promotions to next hierarchy levels when local speakers are elected and followed by other agents and agent's degradations to the lowest hierarchy. Following the initial stage when all agents are at the bottom level in the course of time the system approaches a stationary state where new hierarchies no longer emerge and the distribution of agents at different levels is exponential. In the stationary state the average hierarchy level and the fraction of agents at the lowest level are independent from the system size however the height of hierarchy, i.e. maximal number of observed hierarchy levels grows logarithmically along the total number of agents. The average number of followers of an agent in the stationary state is much smaller than the number of followers he possessed at the promotion moment. Results from numerical simulations are confirmed by an analytical treatment based on the rate equation.
△ Less
Submitted 4 July, 2017;
originally announced July 2017.
-
Competition of simple and complex adoption on interdependent networks
Authors:
Agnieszka Czaplicka,
Raul Toral,
Maxi San Miguel
Abstract:
We consider the competition of two mechanisms for adoption processes: a so-called complex threshold dynamics and a simple Susceptible-Infected-Susceptible (SIS) model. Separately, these mechanisms lead, respectively, to first order and continuous transitions between non-adoption and adoption phases. We consider two interconnected layers. While all nodes on the first layer follow the complex adopti…
▽ More
We consider the competition of two mechanisms for adoption processes: a so-called complex threshold dynamics and a simple Susceptible-Infected-Susceptible (SIS) model. Separately, these mechanisms lead, respectively, to first order and continuous transitions between non-adoption and adoption phases. We consider two interconnected layers. While all nodes on the first layer follow the complex adoption process, all nodes on the second layer follow the simple adoption process. Coupling between the two adoption processes occurs as a result of the inclusion of some additional interconnections between layers. We find that the transition points and also the nature of the transitions are modified in the coupled dynamics. In the complex adoption layer, the critical threshold required for extension of adoption increases with interlayer connectivity whereas in the case of an isolated single network it would decrease with average connectivity. In addition, the transition can become continuous depending on the detailed inter and intralayer connectivities. In the SIS layer, any interlayer connectivity leads to the extension of the adopter phase. Besides, a new transition appears as as sudden drop of the fraction of adopters in the SIS layer. The main numerical findings are described by a mean-field type analytical approach appropriately developed for the threshold-SIS coupled system.
△ Less
Submitted 18 November, 2016; v1 submitted 11 May, 2016;
originally announced May 2016.
-
Coevolution of Information Processing and Topology in Hierarchical Adaptive Random Boolean Networks
Authors:
Piotr J. Gorski,
Agnieszka Czaplicka,
Janusz A. Holyst
Abstract:
Random Boolean networks (RBNs) are frequently employed for modelling complex systems driven by information processing, e.g. for gene regulatory networks (GRNs). Here we propose a hierarchical adaptive RBN (HARBN) as a system consisting of distinct adaptive RBNs - subnetworks - connected by a set of permanent interlinks. Information measures and internal subnetworks topology of HARBN coevolve and r…
▽ More
Random Boolean networks (RBNs) are frequently employed for modelling complex systems driven by information processing, e.g. for gene regulatory networks (GRNs). Here we propose a hierarchical adaptive RBN (HARBN) as a system consisting of distinct adaptive RBNs - subnetworks - connected by a set of permanent interlinks. Information measures and internal subnetworks topology of HARBN coevolve and reach steady-states that are specific for a given network structure. We investigate mean node information, mean edge information as well as a mean node degree as functions of model parameters and demonstrate HARBN's ability to describe complex hierarchical systems.
△ Less
Submitted 9 February, 2015;
originally announced February 2015.
-
Information slows down hierarchy growth
Authors:
Agnieszka Czaplicka,
Krzysztof Suchecki,
Borja Minano,
Miquel Trias,
Janusz A. Holyst
Abstract:
We consider models of growing multi-level systems wherein the growth process is driven by rules of tournament selection. A system can be conceived as an evolving tree with a new node being attached to a contestant node at the best hierarchy level (a level nearest to the tree root). The proposed evolution reflects limited information on system properties available to new nodes. It can also be expre…
▽ More
We consider models of growing multi-level systems wherein the growth process is driven by rules of tournament selection. A system can be conceived as an evolving tree with a new node being attached to a contestant node at the best hierarchy level (a level nearest to the tree root). The proposed evolution reflects limited information on system properties available to new nodes. It can also be expressed in terms of population dynamics. Two models are considered: a constant tournament (CT) model wherein the number of tournament participants is constant throughout system evolution, and a proportional tournament (PT) model where this number increases proportionally to the growing size of the system itself. The results of analytical calculations based on a rate equation fit well to numerical simulations for both models. In the CT model all hierarchy levels emerge but the birth time of a consecutive hierarchy level increases exponentially or faster for each new level. The number of nodes at the first hierarchy level grows logarithmically in time, while the size of the last, "worst" hierarchy level oscillates quasi log-periodically. In the PT model the occupations of the first two hierarchy levels increase linearly but worse hierarchy levels either do not emerge at all or appear only by chance in early stage of system evolution to further stop growing at all. The results allow to conclude that information available to each new node in tournament dynamics restrains the emergence of new hierarchy levels and that it is the absolute amount of information, not relative, which governs such behavior.
△ Less
Submitted 12 June, 2014; v1 submitted 18 November, 2013;
originally announced November 2013.
-
Emotional agents at the square lattice
Authors:
Agnieszka Czaplicka,
Anna M. Chmiel,
Janusz A. Holyst
Abstract:
We introduce and investigate by numerical simulations a number of models of emotional agents at the square lattice. Our models describe the most general features of emotions such as the spontaneous emotional arousal, emotional relaxation, and transfers of emotions between different agents. Group emotions in the considered models are periodically fluctuating between two opposite valency levels an…
▽ More
We introduce and investigate by numerical simulations a number of models of emotional agents at the square lattice. Our models describe the most general features of emotions such as the spontaneous emotional arousal, emotional relaxation, and transfers of emotions between different agents. Group emotions in the considered models are periodically fluctuating between two opposite valency levels and as result the mean value of such group emotions is zero. The oscillations amplitude depends strongly on probability ps of the individual spontaneous arousal. For small values of relaxation times tau we observed a stochastic resonance, i.e. the signal to noise ratio SNR is maximal for a non-zero ps parameter. The amplitude increases with the probability p of local affective interactions while the mean oscillations period increases with the relaxation time tau and is only weakly dependent on other system parameters. Presence of emotional antenna can enhance positive or negative emotions and for the optimal transition probability the antenna can change agents emotions at longer distances. The stochastic resonance was also observed for the influence of emotions on task execution efficiency.
△ Less
Submitted 29 March, 2010;
originally announced March 2010.
