Social anthropology 4.0

1 Acting together – an introduction

This contribution proposes a new approach to human-computer interaction (HCI) based on the scientific findings of self-organisation and in particular on the social example of swarms. The aim is to initiate a discussion about the potential of self-organised collective intelligence in the connection between humans and machines on a horizon in 50 years. To this end, various topics relating to collective decision-making are discussed. A backward-looking approach is taken.

Since the early 1950s, science is said to have been preoccupied with the emerging phenomenon of complexity and has since attempted to fathom it with the help of artificial intelligence, cybernetics, mathematics and systems theory. The aim of cybernetics was to understand human and machine as elements of a self-controlling system. The aim was also to combine intelligence. An approach that comes close to the human being, as collective intelligence is precisely what has brought humanity forward. If technological progress is also taken into account, the idea of a hybrid collective intelligence, as an amalgamation of human intelligence and artificial intelligence from silicon chips and software, is inevitable. ¹ This collective intelligence in the form of socio-technical human-machine systems has long been the focus of scientific interest as collective intelligence systems (CIS) in order to optimise their design. ^{2 , 3}

Today, in 2074, we look back and cannot understand why a principle already given by nature, such as self-organisation, was not applied in order to give the management of complexity a regulated process through this desired amalgamation and thus make the bundled intelligence usable for us. We live in constant dialogue with artificial agents. They enrich both our professional and private lives. In all respects, from the wake-up call in the morning – set specifically between deep sleep phases – to the coordinated departure of one’s own avatar from the virtual world after death in the analogue world. A created symbiosis of human and machine, without which this complex world would no longer be conceivable.

A flock of birds glides across the gray winter sky. A sight that makes me smile and signals to my artificial colleague, who makes this view possible, that it was a good idea to open the roof hatch for this special moment. Such a dialog between human and machine increasingly became the focus of interest in the 1980s. ⁴ However, the term “human-computer symbiosis” ⁵ had already been coined in the early 1960s. The concept behind this was seen as a development that could be expected for future cooperation between human and machine. The background to this was the analyses already available at the time, which indicated that a symbiosis of human and machine would be more effective than could ever be achieved by humans alone. ⁵ The discovery of parallels between biological swarms and human societies, ⁶ as well as between natural swarm intelligence and logical computer science, also occurred in the second half of the 19th century. ⁷ Thus, numerous scientists were working on the overlapping topics of complexity science and the human-machine connection. And there were probably some of them who paved the way for our present-day life in this symbiosis with artificial agents.

2 Necessity of cooperation

Based on the assumption that the joint accomplishment of set tasks is the most effective, cooperation will be considered as a structuring component in the following.

Game theory has already crystallised that psychological influences and social factors, in particular norms or moral concepts, play an important role in humans. ⁴² Since people often behave cooperatively for strategic reasons, institutionalised processes can create incentives in favour of cooperation through artificial agents and their feedback or reputation systems. ⁴³ Even the regulation of artificial decision-makers, as envisaged in the Artificial Intelligence Act (AI Act) ⁴⁴ of the European Union (EU) is a regulation of humans. For example, those involved in the development of artificial agents are to be encouraged to act within the set guidelines by means of targeted incentives or penalties. ^{45 , 46}

However, as the decentralised and hybrid structures in HCT meant that cooperation was no longer the basis for a joint decision-making process, it had to be replaced by coordination. This was already predicted by consensus theory, which is based on constitutive principles to which the actors in a system are subject and which drive them to make a joint decision. Thus, the structure of communication within a system already creates a pull that drives the actors to make decisions and is necessary for a community. ⁴⁷ For our current joint decision-making process in a system of human and artificial actors, this communication structure had to be created specifically to generate this pull. ⁴⁸

If we look at the decision-making processes of natural persons or communities and compare them with the approach of artificial agents, such as machine learning (ML) algorithms, the fundamental differences become apparent. An artificial agent calculates a decision based on mathematical rules. A set of variables is used as input, which is compared with a target as a calculated prediction. Natural agents, on the other hand, rely on a mix of variants. Various heuristics are combined with static procedures and implicit knowledge. At the same time, the diversity and abundance of information has increased exorbitantly in recent decades. This is why an artificial colleague has been added to the human mix of variants as a data collector, processor and visualiser and has become indispensable for successful decision-making.

However, the artificial agent still has difficulties in assessing situations that the human decision-maker has not yet experienced. Although it is possible to exchange experience reports as a kind of verbal simulation within a joint decision-making process, this represents the phenomenal values of the communicating person, which does not allow any conclusions to be drawn about the epistemic experience of others. ³⁰ As a result, visualisation and simulation still have their limits as decision support for both human and artificial decision-makers.

On the other hand, an artificial agent offers support by means of simulations in the cooperation between human actors. By means of sociometric representations such as diagrams, which visualise the course of the decision-making process and its dimensions, or by means of sociomatrices, ^{27 , 49} which illustrate the relationship structure within a group. This helps to deal with differences in interpersonal relationships in order to develop an awareness of the opinion patterns and group-specific trends.

Artificial co-decision-makers are predestined for the creation of visualisations from a meta-perspective despite their involvement in the process. The background to this is their internal and, in some cases, external autonomy and the associated ability to be objective, provided they draw on data that is not evaluated by humans, such as pure sensor data. This makes artificial agents social due to their cooperation-promoting behaviour.

Reasons for assuming that an HCT could deliver viable results became apparent early on. The reason for this is the different decision-making tactics, which in turn result in different strengths and weaknesses. ⁵⁰ For example, human decision-making tactics have a weakness in the assessment of risks, ⁵¹ while algorithms in turn have a weakness in terms of robustness, which can be recognised in particular with increasing dynamics in the initial or data situation. At the same time, artificial agents benefit from the feedback of natural experts, as demonstrated by decision making in a clinical context using reinforcement learning as early as 2022. ⁵² The interactive machine learning (IML) approach already started with the integration of feedback during the modification of an ML model. ^{53 , 54}

The stringent approach of an artificial agent and the emergence and creativity potential of natural systems is another reason for the human-machine connection that exists today. This allows a decision-making process to be optimised even under complex conditions. Of course, this is only possible if there is a symbiosis of human groups and artificial agents using intelligent tactics. For example, the human ability to recognise and understand simple causalities must be revealed to the artificial decision-maker. However, if the correlation used as a basis cannot correctly depict the causal relationship, this inevitably leads to incorrect decisions. This insight was gained through the Generative Pre-trained Transformer 3 (GPT-3) language model, ⁵⁵ which reacted as a trained neural network to speech input from human users. Compared to humans, GPT-3 was almost as good at making rational decisions. However, there were glaring deficiencies in abilities such as causal reasoning. These were due, for example, to the way in which the training was conducted, in which information was passively extracted from data without actively interacting with the environment or its context, which would have been necessary for the development of fully complex human cognition. ⁵⁶

If, on the other hand, the interaction increases in several dimensions due to the complexity of the situation to be decided, the human decision-maker quickly reaches its limits. Due to the multiplying interlocking of humans and machines and the associated joint decision-making, the optimal cooperation between the two actors was used as a kind of social anthropology 4.0. The aim was to interweave causal awareness and creativity at the human micro level and collective intelligence at the macro level, while at the same time efficiently processing huge amounts of data from a linear process using an effective structure of evolutionarily proven and self-organised principles. This approach came a big step closer to the flock of birds permeated by cooperation as a prime example of decision-making as one organism than a mere consideration of the diverse forms of HCT.

Although natural and artificial actors differ in the way they make decisions at the micro level, they do not differ so much in the way they interrelate in favour of a joint decision at the macro level. Thus, communication and cooperation are essential for both actors to negotiate goals and develop the associated process. A hybrid collaboration is conducted through a sensory input and output of data in favour of communication and cooperation via behaviour as a dialogue. ⁵⁷

By drawing on nature’s strategies as a kind of bionic concept, a principled joint hybrid decision-making process could be generated. In addition to mere coordination, this process also offered protection against the mere adoption of calculated opinions from artificial colleagues, as was the case with ChatGPT or in clinical decision support systems (CDSS), for example. ^{58 , 59} In addition, the necessary pull to reach a consensus could be generated by the adapted processes, ⁶⁰ thus creating a system that could be described as a human-computer organism (HCO). Moreover, this concept of joint decision-making had already been tested for thousands of years in evolutionary terms: described as so-called self-organisation.

A self-organised socio-technical culture was already being promoted in 1994. At that time, the model was already living beings that act autonomously as a whole without a centre and also use the phenomenon of emergence for themselves. ⁶¹ This idea was already modelled on a swarm. Not a flock of birds, however, like the one that has just inspired me, but the superorganism of a swarm of bees.

Such a swarm of bees is an excellent analogy for distributed systems in which both the potential of the individual at the micro level and that of the community at the macro level can be optimally utilized. ⁶²

If we summarise all the aspects mentioned in this chapter, the tasks of pioneers become apparent for HCI science. Effective HCT is only possible if cooperation is practised between the entities. In turn, cooperation is only feasible if there is communication between the entities. The essential task is therefore to create structures in which communication can be cultivated and from which cooperation can simultaneously emerge. Possible tools for accomplishing these tasks are usually located at the macro level, as this is where most of the overlaps between the decision-makers occur. This is the way to reach a joint decision: using the tactics of communication and cooperation. The HCI community is supported by the bionic concept of self-organisation, which can be analysed in practice using the example of swarms. If this task of building structures for communication and cooperation in favour of HCT succeeds, in an optimal case a swarm-inspired HCO can emerge from an HCT and social anthropology 4.0 can emerge from the scientific field of HCI, which goes beyond the consideration of interaction.

3 The paradigm of self-organisation

The concept of self-organisation introduced in the previous chapter will now be examined in more detail.

Research into collective intelligence was successful at an early stage in computer science. For example, in the leader-follower problem, packet forwarding and variants of Arthur’s El Farol bar problem. Since numerous other sciences also make use of collective intelligence, such as sociology, ⁶³ or research areas such as behavioural economics, the advancement of research findings was and is essential. ⁶⁴ However, it is challenging that the bundle of shared experiences, intuitions and knowledge does not correspond to an addition of the contributed intelligence, but can exceed the actual sum many times over due to emergence. ^{65 , 66} This emergence in social systems becomes apparent through the reductive description as a social change due to the decision-making action of an individual, but without wilfully bringing about this change. ^{67 , 68} This process is also described as a single “invisible hand” ⁴⁸ that achieves a result in a social system completely unintentionally. In this way, something like objective reason can assert itself in the secrecy of a joint decision-making process, which the actors themselves were not even striving for at the time. The phenomenon of emergence in turn stems from the concept of self-organisation.

When it comes to researching the theory behind self-organisation, the classical models of physics such as deterministic or stochastic methods are not expedient. Even the terminology used in physics does not seem appropriate, which is why almost two hundred years after Friedrich Wilhelm Joseph Schelling’s (1775–1854) theories – which for the first time ran counter to the mechanical view of the world at the time – the search was on again for suitable terminology and his natural philosophy was used. Schelling’s theses were thus the impetus for numerous other hypotheses, research questions and cognitive interests, including those relating to process-based self-organisation. ⁶⁹

In the sky, I am now presented with the spectacle of a swarm that seems to dance as a whole. The swarm doesn’t seem mechanical to me, more like a perfectly choreographed sequence of changing directions, widening distances and then immediately narrowing them again. No one is left behind, no one falls off, no collisions, no runaways: everything seems perfectly harmonised.

The findings relating to such a concept of self-organisation can be traced back to Erwin Schrödinger (1887–1961). ⁷⁰ The development of a resulting theory of self-organisation, on the other hand, arose from Hermann Haken’s (*1927) so-called synergetics. ^{71 , 72} This theory should make it possible to analyse the conditions and processes of self-organisation as well as the resulting states. ⁷³ The fact that this endeavour was not unproblematic was shown by the existing different concepts of self-organisation, each of which also had a different definition. This led to a situation in which no universally accepted, comprehensive and generally valid theory for self-organising systems could be assumed. ^{74 , 75 , 76 , 77 , 78}

The fact that the phenomenon of collective intelligence has already found its way into numerous scientific disciplines has thus been sufficiently demonstrated. Likewise, the situation of a non-existent definition of self-organisation, to which the development of an HCO could refer. Both the conceptual culture and the process itself, as well as the possibilities for shaping this process, will be presented in the following chapters in order to create an opportunity to clarify the concept of self-organisation and the resulting tasks for the HCI community.

3.1 Conceptual culture of self-organisational processes

The transition from the mechanistic view of the world to the later so-called modern physics was based on numerous discoveries and insights. For example, the tunnelling effect showed that elementary particles can also be found beyond the potential barrier. At the same time, overall atomic systems are not sets of individual particles, but each individual electron already changes the overall wave function of the entire system. ⁷⁹ A common denominator in the definition of complex systems was provided by the self-organising dynamics at work. ⁸⁰

In this way, self-organising phenomena can overcome the challenges of forming complex structures due to thermodynamic conditions in living organisms with the development of non-linear physics or non-equilibrium thermodynamics. Together with models of kinetics, this provides a way to analyse and explain cooperative processes in a physico-chemical or mathematically quantifiable way. Both the dynamics and the genesis of such synergetic structures as well as macromolecular biochemical evolutionary mechanisms are comprehensible in their approaches. The background is dissipative and fluctuation-induced instabilities and non-linear phase transitions: self-organisation. ⁸¹ This concept of synergetics gained ground because it defines cooperation within a system by mathematically modelling the transitions of non-equilibrium phases. ⁷² Numerous examples from biology, chemistry, physics or ecology follow this approach to cooperation, such as cooperation in markets, patterns in liquids, spiral arms of galaxies or consensus building in superorganisms and neural networks. ⁸² Cloud formations also follow this approach. Unfortunately, no example that I could reproduce in the grey winter sky. However, as proof, the flock of black starlings continues to dance in front of the grey sky backdrop.

Last summer, I drove a superorganism out of the house, which could also serve as proof: Ants. For example, the use of ant colony optimisation (ACO) helped to find the shortest route for transport robots and to meet the challenges of designing supply chains in logistics. ⁸³ However, other swarms have also provided inspiration, such as the principle of synchronising a school of fish to optimally use the flow field of wind turbines in wind farms or autonomous NASA exploration swarms, ^{84 , 85} which are based on the behaviour of bee colonies as shown in Figure 1; Similarly, autonomous drone swarms in use as a fire-fighting unit in disaster control based on nature-analogue particle swarm optimisation (PSO) as well as medical interventions using nanorobots, ⁸⁶ which as a group can provide minimally invasive and precise treatment. ^{87 , 88}

Figure 1:

Swarm of honey bees (Apis mellifera) during a joint decision-making process (1.1) and during the joint implementation of such a democratically reached decision (1.2).

In addition, upheavals occurring in nature, culture or human society, such as new orders or structures, can also be seen as the result of self-organising processes. ⁸⁹ In physics, such equilibrium phase transitions were considered early on as a form of self-organisation. ¹⁴ The fact that self-organisation is not a linear development of different entities side by side is shown by the property of forming cooperative links that generate a homogeneous structure that enables harmonious integration of all system elements. ⁹⁰ If this were not the case, the acrobatics presented to me in the sky would resemble an air show with starlings flying in parallel.

In the context of research into self-organising processes, the focus was mostly on understanding the emergence and maintenance of order. ^{70 , 75 , 91} This led to the question: How must a complex system be organised so that it is able to organise itself? ⁹² Without being able to answer this question, it is obvious to want to coordinate systems by means of hierarchies or a centre. However, a dynamic nature with cooperating entities in complex structures, which organises itself and constantly reinvents itself via feedback and its synergy effects, is a complete contrast to a deterministic, completely predictable nature. ⁹³ Thus, a complex system is not an isomorphic, static structure, which means that it does not achieve thermal equilibrium. ⁹⁴ On the contrary, phase transitions generate interconnections and thus create new structures in favour of self-organisation. For example, the perception of a complex system can be understood as a learning process on a macroscopic level, which links to existing structures through input, restructures them or forms them entirely as new patterns. ^{95 , 96} Just like the billowing black cloud in front of my eyes.

If atomic systems do not form a mere aggregation of particles, it can be assumed that other systems also have this property. In the section above, self-organising dynamics in complex systems were demonstrated, which can provide an explanation for the formation of cooperative processes. For example, as mathematical models of cooperation through the concept of synergetics. Numerous disciplines follow this approach and have been able to implement cooperation in their dynamic systems without control via a centre using examples from nature. From this it can be concluded that an HCO could also be realised.

4 Human-computer organism

However, in order to create an awareness of the research subject of HCO and thus also for the theory of social anthropology 4.0, the hurdles of self-organization must also be known. We now want to address these and outline them using the natural example of swarms.

The validation of models for self-organisation was simplified or made possible in the first place by computer-aided simulation, for example. However, this was only possible at a time when the Brussels School ¹⁰² was endeavouring to justify self-organisation on the basis of examples. ^{69 , 103} The necessary machines or measuring devices for analysing self-organising processes were either not yet available or not yet in common use. The first descriptive models for demonstrating emergent capabilities offered precise concepts of non-linear non-equilibrium thermodynamics. This enabled a uniform description of organisational structures in systems for different disciplines. Despite the fact that these mathematical models were initially used in physical chemistry and physics, which led to accusations of physicalism, the models can be said to be universally valid in terms of their applicability. However, this assumes that both their semantics and syntax are projected for the application and can be assumed to be appropriate from an empirical point of view. ^{101 , 104 , 105 , 106}

So what stopped science from projecting self-organisation as a principle onto the human-machine system for so long? Why did it cling for decades to its view of HCI, which focuses on the human-computer pair and not the entire system?

Due to the multi-layered nature as well as the prevailing complexity, self-organisation only became a much sought-after model at a late stage in order to make navigating systems comprehensible and to be able to analyse the existing dynamics and their tendencies. ⁷³ However, the fact that even understanding the inherent processes of a self-organising system is a challenge becomes clear to me when I look at the flock of birds. I can hardly stop watching the harmonious structure and yet neither I nor my artificial colleague can predict which turn the birds will take next. When it comes to analysing flocks or complex systems in order to subsequently make use of the knowledge gained, for example by making predictions, modelling and simulation is helpful, but this is no trivial undertaking due to the multidimensionality of the spatially and time-dependent objects. ^{107 , 108} This enormous challenge is immediately obvious to me and perhaps this is also part of the answer as to why the interest in HCI research has focussed on one detail of the human-machine system.

The simulation of an HCO should enable the coordination of dynamics by means of self-organising processes as well as the resulting development of emergent capabilities. To achieve this, however, a space must first be created within the model in which the variable micro-state of each actor is taken into account. In the subsequent calculation of the model in favour of a successful simulation, this means a permanent dynamic, which also constantly correlates anew. But it is precisely this undertaking that enables the self-organising symbiosis of human and machine, which I can lead with my artificial colleague as a kind of HCO.

However, with regard to the challenges of modelling self-organising processes, a look at the modelling of the behaviour of swarms could have been informative. Swarms, for example, are not defined as independent objects, but by their self-organised behaviour. They are always associations or systems that develop through communication, swarm behaviour and the resulting emergence. Examples of this possible emergence include joint decision-making processes, but also coordinated exploration, self-organization or autopoiesis. ^{109 , 110} This definition of swarms via their behaviour and the associated exploration of collaborative processes forms the basis for an understanding of the communication flows and the resulting emergent capabilities of social self-organising systems. Colonies of honey bees (Apis mellifera) are social organism-forming insects, ¹¹¹ which together form a superorganism. ¹¹² Observing the behaviour of a single honeybee outside the hive while collecting pollen and nectar can give the impression of a rule-based process that is continuously repeated. On closer inspection, however, this potentially deterministic system of thousands of foraging honeybees reveals numerous random deviations and fluctuations. Nevertheless, a honey bee colony exists as a whole with a complex organisational structure. The entire system becomes physically visible as a cluster of bees outside the hive (see Figure 1). Just as a flock of birds appears in the sky as a large cloud of birds. ¹¹³

A modelling of consensus building in a swarm was only realised a few years ago in the form of a self-organising network. This made it possible to highlight aspects and mechanisms of self-organisation, which is essential in social systems for a joint decision-making process and thus for autonomous action and the solution of complex problems. ¹¹⁴

Numerous other capabilities, such as cooling or energy and building material supply, show that superorganisms only need minimal changes in the environment to create a new process on a microscopic level. At the same time, cooperative processes are also set in motion at the community macro level, in which individual honey bees distribute, position themselves and whirl their wings at the micro level in the beehive in such a way that a jointly organised flow is created, which draws fresh air from outside to the organism in the beehive. This creates solutions to complex problems by means of fluctuations at the micro level, which become emergent skills, structures and tools at the macro level. The ability to find solutions through emergence arises from the tension between the complex system, the thermal equilibrium and the mathematical non-linearity of the time-dependent evolution equation. ⁹³ The condition of a centre, such as a decisive and thus controlling unit in the system, is therefore no longer necessary. This results in the property of autonomy for a self-organising system.

I realise that I am gradually getting cold and at the same time my artificial colleague starts to close the skylight. Thanks to the understanding of the principles of decision-making in both humans and artificial systems, as well as the findings relating to the natural phenomenon of self-organisation, the focus of research has shifted from HCI to HCO. This means that I can rely on a unique hybrid community of humans and machines that has emerged as a result of technological developments, particularly in the field of artificial intelligence. The concept of the human is no longer at the centre of interest as a placeholder for all people, but rather the organism consisting of humans and machines as a whole and its bidirectionally influenced environment. The artificial actor can be customised to the respective individual, which corresponds to the perception of the individual human being with all its special effects. At the same time, the human actor adapts to the strengths and weaknesses of his artificial counterpart in a mutually enriching way. Thus, autonomous action is optimized through self-organising processes in which strengths and weaknesses are now balanced out in joint decision-making. This chapter aims to provide a brief description of the field of self-organisation with the hurdles that need to be known in order to prepare a path that can be followed by means of skilful manoeuvring to form self-organised HCOs.

5 Structure of a human-computer organism

Building on the previous chapters on the nature of self-organisation, its processes and influenceability, and the challenges arising from self-organising systems, the structure of an HCO will now also be addressed.

Based on the collective intelligence already mentioned at the beginning of this paper, but also due to the scientific disciplines involved and their respective intersections, ⁶³ one can already guess the complexity that must be mastered in order to bring human and machine together as an intelligent, self-organised organism.

As an autonomously acting HCO, it is essential to be able to decide and act together. Complexity science itself, which has developed various models for decision-making processes, has in turn been developed from the scientific strands since the 1940s: Mathematics of complexity, systems theory, theory of complex systems, cybernetics and artificial intelligence. ¹¹⁵ These strands are already interwoven within themselves through numerous ramifications. The theoretical foundations from the logic of joint decision-making are the building blocks on which a sustainable joint decision can grow.

Philosophy and sociology form the gap between orientational and disposable knowledge in order to connect the various building blocks of a collective intelligence. At the same time, the integration of these scientific disciplines expands the instrumental rationality ^{112 , 116 , 117} of mere natural sciences, which is perceived as truncated. Although an ends-means relationship ¹¹⁸ remains assignable within a collective intelligence as a predetermined pattern of action at the micro level, the associated goal achievement at the macro level fails to materialise due to the manifold links within the HCO. The background to this is, on the one hand, the complexity of an HCO and, on the other, the rules and processes extracted from decision theory, psychology and sociology.

Anthropology itself poses the question of the nature of human, but as an overarching discipline it draws its results from the interlinked findings of other, already mentioned sciences of collective intelligence. The existing knowledge and experience of the various disciplines can provide a viable structure as a kind of “fabric” ¹¹⁹ of the common decision-making culture.

All of the scientific disciplines that are relevant to anthropology are those that are closely related to humans. In the scientific disciplines of collective intelligence, these are biology, philosophy, psychology and sociology, and thus four out of six sciences (see the gray paths in Figure 2). ⁶³ None of these disciplines alone is able to define the human being or collective intelligence, although the material object in all of them is the human being. Anthropology, on the other hand, draws on all those sciences that use the human being as a material object in order to track down knowledge about the human being on a broad scientific basis through a variety of formal objects, i.e. from multiple perspectives. This is why anthropology includes the general perspective of philosophy, which asks what is human, the social cognitive interest of social science, the scientific attempt to fully understand humans by means of the structures of their bodies and the desire of psychology to understand humans on the basis of their actions and thought processes. While the term anthropology refers to the sciences relating to humans, the sub-concept of social anthropology focuses on the study of humans specifically as social beings and is therefore of particular interest to HCO. Social anthropology 4.0 now combines all these scientific findings of anthropology as well as the more focused social anthropology of the human actor.

Figure 2:

Roadmap of scientific disciplines in favor of social anthropology 4.0.

Figure 3:

Development phases from user-centered to process-oriented human-computer connection.

However, in order to analyse an HCO in which the opposing strengths and weaknesses of natural and artificial actors come into play, knowledge of the artificial agent is also essential. This can be achieved through the two other scientific disciplines of collective intelligence: computer science and mathematics (see the black paths in Figure 2). At the same time, this is similar to the claim of cybernetics to create a system of self-organisation which, as a hybrid, balances the opposing strengths and weaknesses of human and machine in a joint process.

Here, too, a purely anthropocentric mode is not desired, but rather the embedding of equally entitled actors in a common hybrid system that interacts as a whole and is thus jointly subject to the principles of its environment. Based on digital anthropology, ¹²⁰ a research discipline that emerged from social anthropology and analyses human-machine systems in digital space using a cybernetic approach, digital anthropology can be expanded to social anthropology 4.0 in the context of HCO. Anthropology 2.0 was already conceived at the beginning of the 21st century as a further development of the human body in the context of technological developments and as an upgrade of the human being through technical innovations that characterise the human environment. ^{121 , 122} Research into HCI would therefore represent a type of anthropology 1.0. Transhumanism, on the other hand, could be defined as a dualistic approach to anthropology 2.0 with its connection to the human body. The approach of anthropology 2.0 represents an optimisation of humans through artificial intelligence, whereas social anthropology 4.0 represents research into the joint action of human and artificial actors as a unit in the form of the HCO. Behind the term extension 4.0 lies the concept of web 3.0 with its focus on the aspect of decentralisation on the one hand and the processual logic of industry 4.0 on the other. Thus, the foundations of industry 4.0 are: networking of actors as well as intelligence in the form of communication and the resulting autonomous self-control. ¹²³ An HCO thus combines all these concepts as a decentrally networked intelligent system that is self-organised and therefore autonomous. As social anthropology 4.0, in which multi-optionality and transdisciplinarity characterise the environment and the framework conditions of joint hybrid decision-making actions, the principles of swarm intelligence can combine the tactical rationality of artificial intelligence with the strategic manoeuvres of intuition, creativity and the recognition of causality in human groups and thus form excellent self-organised and thus emergent systems that also act autonomously. For the research field of HCI, the turn towards HCO and thus towards social anthropology 4.0 means an expansion of disciplines towards transdisciplinarity and its multiple perspectives. In addition to the various scientific disciplines (connections in Figure 2), it is the factors (boxes in Figure 2) that influence or enable joint decision-making as the basis for an HCO. The flow from top right to bottom left via cooperation, communication and coordination as well as the paradigm of self-organization also points to the essential factors that must be present for a self-organized HCO.

With this realisation, I turn my thoughts away from the observed flock of birds for good and am grateful for my own flock-like connection, which my artificial colleague and I maintain as HCO.

6 On the shoulders of the HCI giants – process-driven human-computer connection

Based on the attribution that the future is not what will definitely come, but what we believe will come, this article was written on the future of the human-machine connection. This assumption of how such a future can be shaped is based on numerous transdisciplinary findings from scientific research and practice.

Based on numerous findings from HCI, in which the machine served as a tool, it was possible to develop an HCT that assigns cognitive tasks to humans and/or machines and thus switches from a user-tool connection to a human-machine hierarchy. From this hierarchy, the connection between human and machine can be further developed into a self-organized cooperation and thus from a user-centered to a process-oriented approach.

For research into the human-machine cosmos, this means a change in paradigms from the symbolic user as a shepherd over a herd of machines to a beekeeper of a self-organized human-machine swarm. The aim is to create a symbiosis of human and machine that can act self-organized and thus autonomously in the form of an HCO. The natural model for such a symbiosis is the swarm, which acts as a whole by means of its collective intelligence and its self-organized effects and processes. The fact that the concept of swarms works is shown in practice, where social insects have been making decisions and living together for millions of years. However, mathematical models can also prove that colonies of social insects can reach statistically optimal decisions as a unit, ¹²⁴ which are then implemented together (see Figure 1).

Similar to synchronization in the kuramoto model, the individual actors cooperate dynamically to form a coherent whole. Due to successful communication and cooperation, a coordinating center is just as unnecessary as interventions from outside the system. Similar to the primate brain, feedback processes are responsible for creating a coherent state. ^{125 , 126}

In order to help ensure that humans and machines cooperate optimally with each other as a swarm and balance their opposing strengths and weaknesses, numerous transdisciplinary findings are required, which must be discovered, collected, brought together and made available in the research field of HCI (Figure 3). The focus must be on the processes between the individual players. To this end, humans as social beings in a social context with machines must be researched further in order to be able to utilize the processes and structures that occur for the benefit of HCOs. The underlying theory of social anthropology 4.0 can thus be built up and supplemented piece by piece, providing a viable framework of knowledge for the further development of HCOs. For the scientific community in the research field of the human-machine connection, this means that it is both a producer of the product of social anthropology 4.0 theory and an HCO keeper for the functions and processes between humans and machines at the micro level and between HCOs and the environment at the macro level. It is essential to analyze and understand the individual actor of an HCO as well as the HCO itself as an actor (see Figure 4).

Figure 4:

HCO as a whole system at macro level as well as the individual players at micro level.

For academics, dealing with an HCO involves both empirical and theoretical research. Thus, the interest in knowledge revolves around the perception and behavior of the individual actors, as well as their development over time. At the same time, the social behavior of an HCO as a whole need to be researched. For both research interests, internal and external conditions and factors, sequences and consequences of processes and their changes must be recorded and investigated. To this end, the processes of HCOs can be captured as hybrid human-machine systems using the means and methods of systems engineering. Both actors, natural and artificial, must be consistently integrated with all their strengths and weaknesses. In this way, models can be created that truly reflect the necessary aspects of communication and cooperation as well as any necessary coordination and the processes of self-organization (see the factors at the flow in Figure 2). By means of such system designs, for example, dependencies, relationships, possibilities of influence or connections can be explored. In favor of self-organization, the focus should not be on the actors, but on the effects between them. At the same time, a theory should be developed based on these hybrid systems, which can be updated as social anthropology 4.0. This social anthropology 4.0 should also focus on the relationships between the actors, the processes taking place and the characteristics of the system as well as the influences from outside the system.

Numerous problem areas outlined in this article can help to understand human and machine as a system in order to develop innovative approaches that drive the system forward. For example, the inclusion of afferent signals or efferent reactions in feedback loops, the creation of transparency about the uncertainty of subjective factors such as phenomenal values or the creation of intransparency in the case of information that is relevant for decisions but nevertheless influences them, such as age or gender. Similarly, indications of cooperation blockages can be visualized, for example, as well as errors and consumption in calculations or thinking. At the same time, unfavourable dynamics both inside and outside the system can be made aware of and thus neutralized by means of visualizations such as sociometric representations.

As a result, many of the tasks and questions raised will probably have to wait a while for adequate answers, such as the nature of the communication processes between the hybrid actors. However, it remains important that an existing hybrid system, such as an HCO, should not be interfered with from the outside. Only by creating offers or surrogates can changes be made possible, but not guaranteed. This in turn requires science, which creates an understanding of the system, its structures and interrelationships as well as its environment in order to develop optimal offers and assess potential effects.