A recently published paper authored by an interdisciplinary group of 17 researchers representing 21 institutions in 5 countries investigates the ways in which recent technological developments have affected human social systems, and the effects that this has on society as a whole. In ‘Stewardship of global collective behavior’ Joseph Bak-Coleman et al. argue that the study of human collective behaviour – a combination of the methods of both computational social science and complex systems science – should be considered a ‘crisis discipline’ 1. The authors define the term crisis discipline as an area of evidence-based research in which there is urgency to study the dynamics of a complex system, potentially before a complete model of the system in question has been constructed.
The paper highlights four key ways in which our social systems have changed.
Firstly, it points out the drastic increase in scale of the social networks that we are a part of. Starting 12,000 years ago, with the start of the agricultural revolution, our networks have been expanding at a drastic rate, facilitated by the emergence of cities, empires, and nations, as well as the development of global trade routes. More recently, the development of social media platforms has taken the rate of growth up a further gear, to the extent that approximately 3.6 billion of the Earth’s almost 8 billion human inhabitants are users of social media. This incredible inflation of our social systems poses significant coordination problems with results in statistical physics, opinion dynamics, and collective intelligence suggesting that larger group sizes can erode cooperation, inhibiting the group’s ability to reach a consensual decision.
Secondly, social systems have undergone significant changes in their network structure. For much of human history the average individual would have made meaningful interpersonal connections to only her immediate clan or tribe – often fewer than 100 other humans. Similarly to how the scale of the network has ballooned since the agricultural revolution, the size of an average individuals network has also dramatically increased, with many of us having social connections to thousands of others throughout our lives. Furthermore, our social systems now exhibit a far greater number of ‘long ties’ – social connections between two individuals that span a large network distance, commonly as a result of large geographical distance between the pair. These two properties alone allow for the existence of a number of highly connected individuals with the capacity to reach millions, if not billions of others distributed all around the globe.
Speed & Fidelity
Thirdly, improved communication technologies have provided us with rapid, high-fidelity channels through which to communicate with each other. Compared to traditional word-of-mouth communication, which inherently contains ample opportunity for the introduction of noise and errors, as well as a high rate of information decay, modern social channels allow us to bypass these noisy methods, delivering our messages to our desired audiences immediately and accurately. While this has many clear advantages, it may also facilitate the rapid spread of false information through a population. Indeed, preliminary research has shown that introducing artificial ‘friction’ to online sharing mechanisms through shifting users’ attention to the accuracy of the information they share is a promising approach to controlling the spread of false information online 2.
Finally, Bak-Coleman et al. refer to the recent introduction of machine learning techniques to moderate and modify the functioning of the social channels that many of us use on a daily basis. Suggestion algorithms, designed to present us information that they think will elicit interaction, act as a mediator in many of our online interactions, subtly affecting which pieces of information we will be shown. In many cases these algorithms are designed with the sole intention of maximising profitability for the developers with negligible consideration with how they will affect users’ interpersonal interactions.
Bak-Coleman et al. conclude that these features may make current social and political systems ill-suited for dealing with many of the global and existential threats that humanity is currently facing. Issues such as responding to climate change and nuclear disarmament require multilateral cooperation from a wide range of social and political institutions. Knowledge of how best to promote societies, corporations, and states that are willing and capable of collaboration on such issues is thus key to averting catastrophic outcomes. In line with this, Bak-Coleman et al. argue that it is imperative that we prioritise developing our shared understanding of collective human behaviour in order to influence evidence-based policy making. In particular, by suggesting that collective behaviour should be classified as a crisis discipline, Bak-Coleman et al. argue that it should be used to ‘provide actionable insight without a complete understanding of the underlying dynamics’.
Bak-Coleman et al. end the paper with an emotive call to action to address these significant challenges in order to mitigate the risks that humanity currently faces;
‘Inaction on the part of scientists and regulators will hand the reins of our collective behavior over to a small number of individuals at for-profit companies. Despite the scientific and ethical challenges, the risks of inaction both in the present and for future generations necessitate stewardship of collective behavior.’