Foreword: Large Techno Social Systems. Emergence of Social Welfare Techno Systems=Hitzaurrea: Sistema teknosozial handiak. Gizarte-ongizateko teknosistemen sorrera

1. Introduction

Large Techno Social (LTS) systems are infrastructures composed of different technological layers interoperating within society in order to provide global, both public and private services of technological character. They deploy their operations in the “real” material world and in the virtual “digital” world alike. Canonical examples of LTS systems include, Global Transportation and Mobility Facilities, Power distribution Grids, Food and Water Supply Infrastructures, Gas and Oil Pipelines, the Internet, the World Wide Web, Stock and Financial Markets, and the like.

In essence, they are multifaceted and multiscale in nature, and more importantly, they feature complex interconnected networks operating in complex (societal) environments. All this poses substantial difficulties for their full modeling and subsequent understanding. However, recent availability of processable massive datasets and advances in the theory and modeling of physical complex systems provide tools for the construction of an integrated framework towards making reliable predictions of the behavior of both the techno-social systems at stake and the society making-use of them. Thus, it is worth mentioning that recently it has been established, beyond any reasonable doubt, the existence of some physical complex systems which never reach equilibrium [1]. Their dynamical equations of motion yield an unsolvable mathematical problem in a close form. Even more, letting the system to evolve by itself it will not get convergence in a finite period of time. Such systems appear to be extremely fragile, in the sense that they tend to be very sensitive to small both exo- and endo-genous constraints alike, and their dynamical evolution turns out to be irregularly patchy[2]. Remarkably, similar phenomena are also observed in LTSs.

The LTS systems are constrained by the applicable social regulations and market developments which operate in place, sharing their values, aims, public/private tensions, etc. However, the web of interactions between LTS systems and societal regulations and market developments is dynamic. Consequently, they must adjust to “internal” changes, e.g., the changes due to developments of the LTS systems themselves, and to the “external” changes, meant: the evolving “environment” as determined by the actual changes operating within the society which they serve. This dynamic character, carries a time lapse of events which, for the models aimed at describing the behavior of the LTS systems, pinpoints to the requisite of accounting accurately for the responses of LTS systems to changes of their environment in real time. In other words, models need to lay the basic time-ordered features leading to predicting trends, anticipating risks and providing clues for the managing of future events.

Weather forecasting represents a canonical example to illustrate further this point. For our purposes weather forecasting can be seen as the set of “operations” of collecting large real-time climate datasets[3], which assisted with a big set of (consistent) libraries of historical climate patterns, and on-purpose created computational algorithms, after carrying out large-scale computational simulations ends up with the weather forecast for the day[4]. A process that starts off from the established physical laws of fluid and gas flows, generates valuable and reliable societal information, which turns into recommendations which are normally positively, but often insufficiently taken by the individuals. Let us emphasize that knowledge resides at the beginning of the process, e.g. the physical laws. Technical competence, rather sophisticated in many instances, for the processing of large and often unstructured date-sets makes the remaining and, finally, experts issue the recommendations. The end societal goal being the recommendations to make an effect on human “behavior” this day. This, nonetheless, depends on external factors, like risk perception in different communities, tendencies to adopt certain positive individual attitudes towards recommendations, local and global mobility patterns and infrastructure, etc.

Here we can identify a number of “layers”. One layer concerns regulations. Thus, for the weather forecast operation to be successful, full consideration should be paid to (i) on-site regulations for the retrieval of physical data, (ii) the ownership of such data, (iii) the on-site existing regulations/patterns for the mass media organizations, whether they are publicly of privately owned/operated, for the dissemination of the recommendations, and/or (iv) the on-site regulations for the treatment of the recommendations issued. Another layer concerns the local patterns to put in place such regulations. Thus, one should also consider (i) whether the issued recommendations should be turned into mandatory actions to be adopted by the citizens in accordance with the gravity of the predictions, (ii) who is to be taken that decision, and (iii) who is to enforce the measures, let them be, for instance, mandatory fleeing of cities, deployment of police/military forces on the ground, etc. Furthermore, the latter ones are inextricably intertwined with local regulations, which needless to say, can vary enormously from place to place, and bring to the fore the complexity of their mutual interactions.

2. A Scientific Twist

The large majority of LTS systems along with the social environments into which they are deployed are customarily modeled as networks which consist of selected segments of the population interconnected by means of exchanges through the LTS system(s). A large body of earlier studies has reached the consensus that these exchanges are strongly dependent on the scale, both spatial and temporal[5]. Their short- and long-range features differ substantially, thus, requiring a differential treatment as a function of the scale chosen to be investigated. Even more, most real-world networks exhibit a strong dynamic self-organization over time due to the evolution of their internal connectedness, even without the intervention of external factors. In spite of these efforts, LTS systems are yet poorly understood, a fact that often is ignored, until the crash comes. We need a deeper understanding of both the foundations and the time evolution of the LTS systems operating in complex societal environments to get a better grip in order to respond efficiently and responsibly to their changes and failures.

Recent advancements in the science of collective phenomena can offer a fresh view of the dynamics of the evolution of complex systems[6]. Opposite to “primary laws” which put forward their dominant linear terms, complex phenomena science introduces non linearities and multiple causalities (multiscale approach) in the description of phenomena which do not evolve towards equilibrium (stability), but instead instability builds up in a recurrent manner following the newly discovered scaling laws[7]. These scaling laws are already familiar in epidemics and earthquakes, but they also appear in the growth of cities, stock market behavior, etc. The bottom-line here is that complex systems do not evolve randomly, buy in accordance with underlying forces which trigger their evolution yielding a critically self-organized structure which applies equally to socioeconomical and technological systems. The criticality of their self-organization leads to sudden breakdowns. This is what we see and most of the times suffer.

Fortunately, the multiscale approach as it stands today, has paved the way towards the understanding of the statistical and dynamic laws governing the functioning of LTS systems coupled to complex social “environments”. Certainly, there are a number of LTS systems which appear to fit better than others, within this multiscale framework. Thus, infrastructures for human, raw materials, and manufactured goods transportation, as well as large scale operations aimed at delivering large amounts of gases and fluids over long distances, do nicely fit for multiscale type approaches. The large techno systems for the transportation, distribution and assurance of enough power to supply electricity at every electric plug, constitutes one additional example.

Ephemeral, to say the least, matter is also, nowadays in particular, processed by large techno social systems distributed physically around the world, and in reality, in the so-called “cloud”, where the custody, either designated or open, the accessibility and delivery of data, big data, through the Internet is carried out by the administrators of such cloud. These include the network of microwave antennas for mobile phone communications, and the channels for the dissemination (or occultation) of “sensible” data for the operations of financial stock markets, to name a few.

3. The Web

Perhaps the canonical LTS system is the Internet and its associated World Wide Web, the web for short. Customarily, the Internet and the web are used to describe the same, but they refer to two very different things. The web refers to the means to get access to the pages that we browse on the computer’s monitor when it is online. We can browse pages, edit them, share their location, etc., by means of a standard on-purpose language, the HyperText Markup Language (HTML). The Internet, on the other hand, refers the network made of a myriad of interconnected computers, serves, and countless of other devices which support the web.

The web cannot be conceived without an explicit reference to human society. On the one side, the Internet is the most globally shared technological infrastructure of the modern world. On the other side, the web consists of the materializations of the activities of humans on such an infrastructure. Humans are here producers and viewers of the web’s content. This dual nature of human agents is the driving force of the hyper-evolving nature of the web, resulting in an emergent information dissemination, creation and sharing infrastructure on a global scale. The web is, therefore, a LTS of computer mediated cognition, communication and collaboration for humans[8]. It is the only LTS system where users contribute effectively to its technical development for they adapt the code to their social communicative needs. Neither the detailed structure nor the precise content of the web can be predicted at any given time. Pages appear and disappear in accordance with the users’ needs. But one thing can be predicted for sure, its complexity grows with the passing of time, as the number of web sites/pages, web search engines, linking patterns and their interconnections increase. Additionally, one more agent has made recently its debut in the web’s world, the artificial intelligence[9], which brings to the fore that the hyper-evolving nature of the web refers not only to its content’s diversification, but also to the appearance of new stakeholders. Naturally, yet a satisfactory and sufficient understanding of the reality of the field, just for not going all the time chasing its development’s tail so we could properly manage its implications without impeding advances, remains to be achieved. In particular, artificial intelligence powered with quantum computing, a combination that is sought to be within reach in the near future, deserves an attentive close watch because it has the potential to lift biotechnology, nanotechnology and robotics to a such level that could utterly reshape the world that we live in today.

4. Cities

However, there are other LTS systems which are less prone to fit into this scheme. Consider for instance the urbanism criteria which informs the planning for the rational development of the layouts of cities, homes, houses, along with roads, railroads, and additional means of connections among them like, for instance, rivers as transportation means, which implies the deployment of fluvial ports, locks for raising or lowering boats around cliffs, docks for the ships to be repaired, etc. In this case the demarcation line between the LTS and the services provided is rather blurry. Cities can be viewed as a complex system connecting individuals and communities - bearing their specific economic and cultural characteristics - with technical matters, including planning, (re)construction, maintenance, etc. Consequently, they can be viewed as LTS, an approach that can help enormously to set the problem into proper perspective. Furthermore, it does not escape to our attention that progress in the theoretical formulation of the problem can represent a significant advancement in terms of practical applications, for it will provide rational tools to design better city development planning strategies.

It is estimated that more than half of the world’s population lives in cities, a proportion that is increasing along with the problems associated with settlements of dense communities. This is one of the great paradoxes of our ages. Namely, new technologies offer individuals and companies alike greater freedom than ever to choose their location and mobility, yet more than ever people choose to live in close vicinity of each other. This irrational[10] behavior is, surprisingly, well described by the Shelling model[11]. The model assumes that every individual prefers to live in a “moderately” dense city, because the accessibility to communal services is easier than in highly dense cities, and they offer better opportunities for “socializing” than deserted cities. Also, it is assumed that each individual is free to move to another city if she considers that its density is closer to the optimum, i.e., moderate. This freedom of movement is expected to act as a density equilibration factor. But in the long run it does exactly the opposite to equilibration. Namely, some cities end up overpopulated (relative to the optimum) and some totally emptied. Instead of equilibration it yields segregation, and consequently accumulation of the population in the big cities.

The genius of the big cities resides on the fact that they are a physical space of (high) density of human interactions which facilitate the rapid dissemination of information and the surge of new ideas[12]. This nonetheless brings to the fore the claim that computerized information technologies, ICTs, reshape the concept of “space”, and this profoundly affects cities[13]. Indeed, ICTs mark the emergence of the so-called “space of flows”, which refers to the flows of computer bits of information which overlaps with the physical “space of places”. Both having nowadays similar impact on the growth of cities.

All in all, since cities are bound to grow, they have to find strategies to allow a rational growth without falling under their own growing impetus. The scaling law of accelerated production, “growth stimulates growth”, makes attractive to people to move to the city. But then growth jeopardizes the infrastructures and services of the city, resulting in increasing numbers of individuals struggling to make its space in increasingly overcrowding environments. This might occasionally result in a crash. Detroit is canonical example, but recent events like the one occurred in Birmingham, where its City Council has declared itself effectively bankrupt preventing all but essential spending to protect core services[14], remind us how little is needed to cause systemic failures. Construction of suburbs has been seen as a solution, but it puts pressure on transportation services to begin with, and in the long run seems to be non-sustainable. At what size does the “repulsion” exerted by the inhabitants causes a city to collapse? At the time being, cities avoid collapsing by relentless innovation. The larger the population the shorter the time for the next innovation. We are now beginning to set the foundations to understand the mechanisms of these non-linear saw-toothed patterns[15]. However, further analysis is foreseen on such sensible issues, including the validation of the proposed approaches.

5. Research and Technology

Global research, development and innovation initiatives where a large number of laboratories and companies dispersed around the globe do networking and cooperate to produce one final product, constitutes one example of LTS oriented towards social service. Consider for instance the European Organization for Nuclear Research (CERN) which operates the largest particle physics laboratory in the world. The components of the CERN’s large hadron collider are normally manufactured, and sometimes even partially assembled, in places very distant from the laboratory’s venue at Geneva, Switzerland. Another example is the International Thermonuclear Experimental Reactor (ITER). The ITER project is a well-known large techno social initiative with thousands of engineers and scientists of 35 nations collaborating since 1985 to build a magnetic fusion device designed to prove the feasibility of nuclear fusion as a large-scale and carbon-free source of energy, based on the same principle that powers our Sun and stars. This brings to the fore the scale of the complexity of the technological LTS systems.

Developing a new and useful technology is and has always been a laborious process. Implementing it is even more laborious, because it has to replace, partially or in full, an old but well tested and optimized one. Thus, efficient technological societies are resilient to change. Exceptions occur when they are driven by imperious outside stimuli. The development of synthetic nitrates for securing production of ammunition or the nuclear bomb project are examples of responses to military stimulus. The synthetic rubber and the mass production of penicillin and COVID vaccines were successfully implemented in a short period of time driven by economic stimulus, and perhaps a little bit of altruism. But in general, technological development is complex, hard to predict, and full of unexpected surprises around many corners.

Gone are the days in which research centers, universities and companies conducted their research and development (R&D) operations in their own laboratories. Today, R&D is conducted globally. Technology is often acquired from external suppliers, start-ups, other research centers around the world, etc. In fact, the overwhelming majority of knowledge is generated externally, within the global (planetary) technological LTS systems’ network. Thus, initiatives aimed at capitalizing R&D done elsewhere, are nowadays part of the system itself, and have become a critical innovation hot spot for many nations. The old schema of Government financed basic research which permeates the local public/private sector’s R&D centers to generate innovations which finally creates economic value is deeply flawed, instead products come into existence after a long process of many steps interconnected by intricate ways. The majority of these products do not yield radical changes in society, but some do. And when this occurs it leads to a complete overhaul, known as Schumpeterian waves[16]. This mimics phase transitions in self-organizing critical systems. A characteristic of them is that they are recursive and repetitive[17]. Understanding their evolution requires coming to grips with the overall complexity of the technological LTS system.

6. Failures in LTS Systems

Things, though, may go wrong due to failures of LTS systems. This may have various types of consequences depending on the nature and services provided by the LTS system(s) which have suffered the failure. Two types of failures can be foreseen, e.g., conventional and systemic. Conventional failures can be constrained in space and time, they follow linear cause-effect relationships and can be addressed with effective and pointed interventions into the cause-effect chain, and are often restricted to a given single societal domain. Systemic failures[18], however, are characterized by high complexity, far reaching effects, stochastic relationships, non-linear cause-effect patterns with tipping points, and often receive less public attention than required. Part of the problem for properly handling such failures resides in the poor fundamental understanding of their nature, which impedes a proper and fair perception of the risk of systemic type failures.

Although it is well known that practical tools for properly assessing the significance of systemic failure risk have recently been started to be put in place, systemic failure risks seem to be irrationally attenuated in public risk perception[19]. This has been named as the “systemic risk perception paradox”, namely, social perception of systemic failure risks and the empirical reality as reflected by publicly available data, counter-match[20].

Systemic failures constitute complex, trans-domain, non-linear phenomena often showing non-return points and causing cascade events which harm various societal domains at a time. LTS systems in general and specifically those engaged in economy and finance, naturally fall into systems prone to tough to detect in advance systemic failure episodes. Schumpeter argued[21] that part of the problem stems from the very nature of the social dynamics, which is a “process” with so many sources of error and variables, most of them not attainable to precise enough measurement, so that making the correct diagnosis of certain situations becomes a matter of chance.

In this vein, the new complexity science approach for social dynamics can give us a fresh view of systemic failures and their social impact from a different viewpoint largely routed in methods coming from complex systems studies and non-equilibrium statistical physics[22]. This can be seen by many as a non-solicited incursion of newcomers (physicists) into a field (social sciences) not of their expertise to which they can certainly provide some novelty, but at the same time, undoubtedly, it must be recognized that they arrive with an exaggerated self-confidence and naivety believing that they truly can overhaul old prejudices and contribute a fully new perspective. Under such circumstances, normally, the ones that have been in the field for many years do not expect great “revelations”. And yet fresh perspectives are welcomed for they allow us to spread doubts onto venerable ruling doctrines that have been establishing fond conceptions and dictating professional practices in many matured fields.

7. Economy and Finance

In view of the above, consider the venerable doctrine comprising market equilibrium economic models aimed at describing market interactions of rational agents. The resulting equilibrium balances prices, cost, demand and supply, and is supposed to be sustained and preserved forever without ever suffering from any sort of crisis[23]. However, this assumption is at odds with the reality of the statistics of price variations in financial markets which shows abrupt changes of all sizes, given rise to a truly non-equilibrium like evolution.

The crises of 1987 and 2008 constitute a convincing confirmation of the latter. Consequently, the “equilibrium assumption”, although it can be useful, seems to be false, stricto sensu. Namely, crisis do occur. Why do they occur? Three factors stand prominently in the explanation of crisis breakouts: (i) unknown tiny bits of information about relevant matters, (ii) the concurrent irrational behavior of the trades, and (iii) the fact that trader’s behavior is driven by their expectations concerning the future evolution of the economy, and in consequence the evolution of the economy is driven by those expectations. That is, the economy and peoples' expectations about it coevolve[24]. This brings to fore the idea that the economy nowadays is better seen as a rational exercise based on incomplete information, lack of knowledge, and uncertain expectations. The rationale behind this statement is that the only hope to deal with swings, bubbles and crashes lies in the competent rational behavior of the traders, based on a deeper knowledge of the underlying non-equilibrium mechanisms of the complex, hyper-connected financial large techno social system.

The fresh perspective provided by the advent of massive amounts of data along with new means for handling large data repositories in “unconventional” manners is opening new avenues for collaboration between experts of different disciplines[25]. In particular, given the enormous amount of available data on human behavior, econometricians are nowadays actively engaged in processing their data by taking into account non-linear, non-equilibrium considerations. These effects are crucial to understand and predict, minimally at least, seemingly “irrational” market behavior, which in turn allows us to study how markets either remain steady or change abruptly. As L. Hansen put it “little jiggles within the system today can have a big impact in the future”[26]. This could provide the means for dampening wild fluctuations to tame future crisis by properly quantifying the impact of uncertainty[27].

8. Health

Among the most pressing challenges that modern societies face nowadays, the provision of a universal health system for the population ranks high on the policy agendas of many countries. In developed societies, this means not only keeping citizens alive but, allied with medical technological advances, to provide individuals with good enough health, as to enjoy high quality living standards during an increasing period of their lives. Developed countries allocate some 10% of their gross domestic product for the health provision systems, which beyond the usual political-party fights, in most European democracies at least, enjoys a massive consensus of pride towards what it is considered to be one the foundation stones of the so-called “European social model”[28]. Certainly, this amount of funding represents a substantial portion of the public resources, which highlights the enormous size and complexity of the LTS health systems. Recall additionally, that the LTS health systems are constituted by a mélange of public institutions and private foundations and/or companies given rise to an overwhelming diversity of legal and regulatory structures operating simultaneously and concurrently. The first lesson to be learned is that organizational issues will be a standing out matter of discussion in this context, consisting of the deployment of increasingly sophisticated medical technologies for an increasingly “sophisticated” and diversified society.

Recent modeling of collective social behavior shows that emergent phenomena can sprout under special circumstances. Small variations of social conditions, in the edges of internal tension situations, are enough to push the whole system into a systemic crisis, which in fact propagates through the social network affecting the individuals severely[29]. Since public policies to manage such scenarios are a matter of social engineering, deep understanding of the mechanisms responsible for the outbreak and subsequent spreading of such phenomena are particularly important.

Thus, in addition to the unexpected and yet not fully understood long-term effects of the SARS-CoV-2 on humans’ physical health[30], the outbreak of mental illnesses supposedly linked to the social constraints imposed during the COVID-19 pandemic, has come as a surprise to many[31] and has raised red flags towards the need of changing old methods to treat mental health problems triggered by large periods of large stress on the population, under the additional pressure exerted by modern infodemic[32] scenarios. Under these circumstances, systemic failures -we have discovered it recently, can also be caused by the lack of fast enough adaptive response of the LTS health systems themselves to sudden changes of their social environments. The nowadays consensus on these matters seems to be that the outbreak of the COVID-19 pandemic created an environment where poor mental health found the conditions to become ubiquitous and, consequently, brought to the fore, in almost every country of the world, the need to put into question the fundamental assumptions that for many years past have been at the core of the balance between the physical and mental health care provision protocols in most LTS health systems. This leads to a scenario of a large (techno) social system adapting to a changing, disruptively in this case, environment, requiring among other actions to be taken, monitoring more closely than ever before early signs of the mental health deterioration within the population, in order to properly face along with physical health issues, also mental health problems too.

In this issue, a number of conventional LTS will be put side by side with some of a less conventional nature, and them all will be revised thoroughly. The intrinsic complexity of their structures will be emphasized with the aim of bringing to the fore the necessity of a multifaceted approach in order to get a clear glimpse of the connections with the societal environment where they operate. Noteworthy, efforts will made to highlight the common complex mechanisms that underlie the basic features of most LTS systems, which characterize their surprising similarity.

Acknowledgments

Authors are grateful to Prof. Palacios-Huerta for his time and energy in providing invigorating comments and helpful suggestions that have improved the manuscript.