Looking at Coronavirus Epidemic, Immigration, Brexit and Global Warming through the Lenses of Complexity Science

George Rzevski, Professor Emeritus, Complexity Science and Design,

The Open University

 

 

Introduction

 

At the time of writing, the UK was facing several important issues, which were quite clearly too complex for general public and politicians to understand and to reach consensus on how to resolve them. 

 

What was much more disturbing - scientists, advising government or simply expressing their views, could not agree among themselves. It is not surprising therefore that issues were politicised, and rational debate replaced by activism. That was true for each critical issue: coronavirus, immigration, Brexit and global warming.

Perhaps, if we looked at these issues through the lenses of complexity science, we could see much clearer which steps should be necessary to take to resolve them.

 

This is a collection of four essays which were written between 2015 and 2020 and updated just before collection was published. 

 

An appendix containing a short overview of complexity science fundamentals was added for the benefit of readers who are not familiar with the subject.

 

 

Coronavirus

Why was modelling of epidemic such a mess?

In the article entitled “In Bed with the Supermodellers”, published in the Sunday Times on 10 May 2020, Bryan Appleyard described a chaotic situation in which diverse and often conflicting models of coronavirus epidemic are proposed by academics, based on guesses and opinions rather than real data, and concluded “we follow the science, the politicians say, but there is no such thing as “the science””.

 

How appropriate, I thought at the time. We certainly don’t have the science – we have, in fact, two different sciences. 

 

We have a traditional science, with roots in physics, the Queen of Sciences, often referred to as Newtonian Science, which is based on the assumption that the world is deterministic, in other words, predictable. And if we don’t know how to predict an aspect of world’s behaviour, for example, how coronavirus epidemic works, it is because we do not know enough about this particular subject, but we could be sure that the further research will provide us with the answer. “God doesn’t play dice with the universe” is a famous way in which Einstein asserted determinism.

 

However, Newtonian science cannot deal with many issues in biology, social sciences and even in chemistry. Primarily because many key processes connected to life are irreversible. The tipping point was reached when Ilya Prigogine in his laboratory created organic elements from an inorganic mixture, modelled on the primordial soup. This astonishing discovery of autocatalytic properties of a mix of chemicals, could not fit into deterministic understanding of our world. It was, clearly, an extreme event, a paradigm shift.

 

From Prigogine’s work emerged the new science of complexity, the second, quite different, science to which I referred above. Santa Fe Institute in the US developed their own influential version of complex adaptive systems.

 

The philosophical foundation of the new science is expressed in the assertion that the world irreversibly evolves, and future is not predictable. Future is created by the accumulation of everyday actions and interactions of all living and non-living constituent components. Every infection, war, scientific discovery, trading transaction, financial crisis, erosion, earthquake, tsunami and procreation, changes the world in a small and unpredictable way. “Future is not given” wrote Prigogine, in sharp contrast to Einstein’ assertion of determinism.

 

The subject which complexity science investigates is exactly what is required to understand viruses, but also immigration, global warming and Brexit. Complexity science studies complex groups consisting of diverse participants exchanging messages and competing or cooperating with each other, within constraints imposed by the system to which they belong but without any central control. Participants are called agents. 

 

Such groups are labelled complex because their behaviour is inherently unpredictable, although not random. The behaviour emerges from the interaction of participating agents and is very smart – complex groups selforganise, rapidly adapt to disruptions and, under certain conditions, are known to exhibit collective intelligence and creativity. 

 

A good example is a group of viruses deciding if they should stay longer in the cells of a victim or switch to a new host. 

 

Butterfly effect is a colourful metaphor for consequences of nonlinear relations between members of a complex group or, more formally, a complex system. Under certain conditions, a disturbance, as small as motion of a butterfly wing, at one point on the planet may cause an extreme event, as dangerous as a heavy storm, at a faraway place or, indeed, on a truly global scale. 

 

The most striking example of the butterfly effect is the coronavirus pandemic – it began by one person eating infected bat (or, maybe, by one mistake made in a laboratory investigating coronavirus) and resulted in millions of people infected worldwide, huge number of deaths and a global economic crash. 

 

Here is how the article “The Secret Social Lives of Viruses”, published on the website of the prestigious journal, Nature, on the 18th of June 2019, depicts a group of viruses making decisions: “the viruses …. were chattering away, passing notes to each other in a molecular language only they could understand. They were deciding together when to lie low in the host cell and when to replicate and burst out, in search of new victims.” 

 

How could complexity science help to model epidemic?

This most remarkable discovery made by scientists working at the Weizmann Institute of Science in Rehovot, Israel, and published in Nature, should have signalled to all researchers modelling the spread of coronavirus that by neglecting group behaviour of viruses they are making a serious mistake.  

 

Complex groups are extremely good at adapting. A group of highly infectious but mild viruses successfully passing from one human body to another without killing the hosts, when faced with the lockdown, which drastically reduces the opportunities for finding new hosts, is likely to adapt to new circumstances, perhaps by mutating into a much more aggressive variety. The implication is that the lockdown, hailed as a panacea, may not be such a good idea unless is highly selective – locking in only those that are vulnerable. We should have paid more attention to the group behaviour.

 

The other serious failure of modelling epidemic was to rely on the average transmission coefficient R, when it was obvious that virus transmission was quite different under different conditions – for example, strong in hospitals and care homes and weak outdoors. 

 

We should have used digital simulators designed to investigate behaviour of complex groups under a variety of conditions, instead.

 

We should have certainly made an attempt to simulate the behaviour of a group of viruses under conditions of different protection policies. Simulating wars between a group of viruses and a team of protection strategists is an excellent way of learning how to manage an epidemic 

 

How could complexity science help to prepare for the “next normal”?

Locking us all in was a drastic action undertaken, apparently, on advice of an inadequate model. Let’s consider now how best to exit from the lockdown. To do this effectively, we should have a good idea how the “next normal” would look like. And here the complexity science can certainly help.

 

Remember the key philosophical premise of complexity science: the future is created by actions and interactions of all living and non-living components of the world. And one of the most powerful man-made components of our world is technology. Society invests into creative minds to invent new tools, which are expected to help society to achieve prosperity, but inevitably by using new tools, society changes. New tools create new jobs and new jobs upset the old order. This process is known as the coevolution of society and technology. 

 

We are currently in the middle of a rapid shift from industrial society, relying on the mass-production technology to generate prosperity, to information society, whose prosperity depends on the widespread use of digital technology - digital computing, digital communications and, above all, digital intelligence or, as it is better known, artificial intelligence. 

 

Socio-technological evolution is unstoppable. Neither the lockdown nor the opposition to change by those who prefer the status quo can derail its flow. In the near future, we can expect the transition to digital age to accelerate. Here is a small sample of changes which will happen.

 

We shall have more lectures at all levels delivered online, supported by extensive tutorials, discussions and team-based projects face-to-face. By replacing at least some face-to-face lectures by expertly produced multi-media presentations, we shall increase time available for active, independent learning and for important face-to-face contacts between academics and students and between students.

 

Shopping online will grow.

 

About 40% of all current jobs will be taken over by artificial intelligence, without reducing employment opportunities but requiring retraining or import of special skills.

 

The work will be moved closer to workers. We shall not have to travel long distances to our offices. Businesses and administrations will arrange for us spaces near our homes, where we can work both online and face-to-face with our colleagues, when not working from home. 

 

Manufacturing will be moved closer to consumption. To avoid global supply chains, we shall convert huge mass-production manufacturing plants to smaller, flexible and versatile manufacturing units served by robots and artificial intelligence, and located near the demand.

 

The UK will be moving further into the digital era, that is inevitable, but will we move fast enough? It all depends on how quickly we can switch our mindsets from the “economy of scale” to “small is beautiful, if connected”. We should have a V shaped recovery if we focus on digital transformation of the ways we practice business and administration.

 

Mass Migration

Is mass migration out of control?

Mass migration from Asia and Africa to Europe and, particularly to the UK, is likely to accelerate and therefore represents a threat. Nail Ferguson (Sunday Times, 6 January 2019) cites a Gallup survey in 2017, according to which more than 700 million adults around the world would like to move permanently to another country. 23% would prefer to move to a country within Europe and 21% to the USA. 

 

It is not surprising that in the target countries there is a strong opposition to the increase in the flow of immigrants. Recent polling by the Pew Research Centre shows that only 16% of the UK population would welcome more immigrants. 

 

In the same article, Ferguson writes “if the choice is between open borders and defensive walls, history suggests walls – and those who build them – will win”. 

 

This statement can be considered as sensible only if we give it a charitable interpretation that “defensive wall” is a metaphor covering all interventions, which could prevent undesirable visitors crossing open borders, rather than literally building physical walls. 

 

I shall attempt here to answer the following questions on migration:

  • What are the factors that trigger and drive mass migration?
  • Which, if any, intervention could affect current mass migration from Asia and Africa to Europe and, particularly, to the UK?
  • Does UK need immigrants?
  • What should be the elements of a comprehensive UK immigration strategy?

Let’s first establish that mass migration is indeed a complex system. 

 

Mass migration as a complex system

We already know from the first essay that a complex system, or a complex group, is open (interacts with its environment), consists of a large number of diverse components, called agents, which are richly connected with each other and engaged in intensive interaction, has no central control and behaves unpredictably, but not random – it follows recognisable patterns.

 

Terrorist networks, military conflicts, natural ecosystems (forests, grasslands, rivers, oceans), human brains, the Internet-based global market, nations and communes are all complex systems.

 

Mass migration has characteristics of a complex system, as defined above. 

  • It is open – it feeds on information on where is best to migrate and on availability of clandestine support for illegal entry into target countries, including hiring of transport. 
  • It consists of wide variety of constituent migrants connected using latest communication technology, like smart phones and streaming, and engaged in continuous interaction – sharing latest news, gossip and opportunities.
  • It has organisers, leaders but no central control.
  • It behaves opportunistically, adapting to ever-changing, hostile environment and its overall behaviour is therefore unpredictable but not random.

In the 21st century we have built a complex, connected world in which individuals and communities have access to advanced communication technologies and can exchange news and gossip and conduct business among themselves with (almost) the speed of light. In such a world, all global issues are interconnected - geopolitical conflicts, trading alliances, tariffs, migration, poverty, crime, terrorism – and therefore any unilateral political, military or economic intervention is likely to cause unpredictable consequences. Highly connected individuals, communities, movements (such as mass migration) and nations are capable of rapidly reacting to any disruption or attack with a view to eliminating, or at least reducing, consequences. 

 

The best evidence is the fact that no nation, or military block that recently initiated a military conflict, achieved its objective – nations on the receiving end managed to rapidly selforganise into guerrilla resistance, capable of prolonging a conflict, waiting for the attacker to give up. 

 

Here is an example closer to home. When efforts were intensified to prevent migrants to be smuggled from France to the UK in lorries, they rapidly discovered a new route across the Channel – using stolen French fishing boats.  

 

We live and work in a true global village. And in a global village, very much like in any village, positive results are easier to achieve by negotiations and conflict resolution than by aggressive posturing and threats. We should remember this when planning how to control mass migration.

 

Factors that trigger and drive mass migration

Current Migration from Asia and Africa to Europe and, particularly, to the UK appears to be triggered and driven by many related factors. Here are some:

 

  • Local military conflicts often caused or made worse by military interventions by big Western and/or Eastern Powers.
  • Very low living standards.
  • Very high unemployment.
  • Information reaching potential migrants that in Europe living standards are much higher and that there are no military conflicts (potential migrants have access to modern communication technology).
  • Information that the UK offers the best chance to find employment; British employment laws and regulations are the most liberal in the West, which is why migrants rarely stay in France – they try to get across the Channel by all possible and impossible means.
  • Availability of a clandestine network of resources supporting practical aspects of mass migration such as transport and support for illegal entry into target countries.
  • Sense of adventure and curiosity, the very same that drove early British explorers to discover wonders of the world and early British empire builders to conquer new territories.

 

All factors identified here are long term. Local military conflicts, low living standards and high unemployment are likely to persist, unless concerted efforts are made to help, and illegal networks of smugglers will make sure that flow of migrants does not dry up, unless destroyed. 

 

After Brexit, presumably, free flow of immigrants from European Union countries into the UK will not continue, and the same, or similar, entry criteria will be applied to all immigrants.

 

Interventions that could affect current mass migration

If factors that trigger and drive flow of immigrants from Asia and Africa to the UK have been identified correctly, we can reduce the propensity to migrate by

 

  • Stopping our military interventions, which are fundamentally cruel and useless; remember that no nation that recently started a war, won it. 
  • Seeking cooperation between big Western and Eastern Powers. It is transparently clear that big Eastern Powers will not go away anytime soon, and that the new cold war between the West and East cannot be resolved by aggression; it is much more likely that deals on particular issues would be agreed by negotiations and trading in favours. It is much more effective to make sure that our house is in order and that the population of communist countries see that life is better in a democracy – they will find a way to erode power of their dictators and help to end cold war.  
  • Investing into the infrastructure and businesses in regions that represent major sources of migration with a view to improving living standards and increasing employment opportunities. An additional benefit would be the creation of new markets.
  • Changing our employment policies and regulations to prevent undesirable migrants gaining employment in the UK, without closing door to skills that are essential for our economy.
  • Ensuring that information about changed employment opportunities reach potential migrants before they join the flow.
  • Acting to prevent operation of clandestine migration support networks, employing methods similar to those used in antiterrorist defence.

 

Does UK need immigrants?


Yes, we do

 

  • The UK economy is in transition from industrial to knowledge-services economy. Knowledge services skills and, in particular, Information Technology (IT) skills will be increasingly in demand and this demand is easy to meet by immigrants with appropriate profiles.
  • Our healthcare industry needs additional doctors, nurses and supporting staff.
  • The UK economy requires a steady supply of workers with simple manual skills for agriculture, catering and hospitality industries (farm workers, waiters, cleaners) and for support of households (domestic help, gardeners, plumbers, electricians).
  • British universities are among the very best in the world and they attract international talent - talented staff and students are required to maintain high quality teaching and research.
  • After Brexit, it will be essential to maintain image of an open society, ready to work together with the world. 

 

No, we don’t

 

  • One of the key problems with immigration is the conflict between cultures – the host country culture and the immigrant’s culture. Using parlance of complexity science, culture limits autonomy of its members; it imposes norms on how people dress, marry, worship, eat, and therefore emphasises differences between the hosts and immigrants. 
  • In all complex groups, there is a propensity for members with similar features to cluster. The UK immigrants are not exceptions, for example, in London we have prominent immigrant clusters - Polish in Ealing, Indian in Southall, French in South Kensington and Irish in Kilburn. Clusters of foreign culture, once formed, is almost impossible to disperse.  

 

Well, perhaps we do

 

Industry needs are perhaps best satisfied by allowing certain number of immigrants to settle permanently or temporarily in the UK, subject to provision of resources for housing, schooling and healthcare. 

 

All above considerations must take into account that within next 10 years Artificial Intelligence (AI) is likely to take over about 40% of current full-time jobs.

 

Elements of a strategy for managing immigration to the UK

To control the flow of immigrant into the UK, it will be necessary to devise a comprehensive immigration strategy, which would ensure that 

 

  • Inflow of migrants is demand driven. In practical terms this means that vacancies for work or education dictate who can come and for how long.
  • Protection against illegal entry considers all interventions listed above, not just patrolling boarders. 
  • The immigration strategy will have to coevolve with the global geopolitical and economic environment within which the UK will operate after Brexit. 

 

Brexit

What do we gain by looking at Brexit through the lenses of complexity science?

To fully appreciate the business opportunities that Brexit offers to the UK, and to make an informed judgement whether the UK is capable of taking advantage of these opportunities, it is necessary to realise that we live and work in the exceedingly complex social, political, economic and technological landscape and therefore we can expect to gain new insights if we review Brexit through the lenses of complexity science.

 

The word complex in complexity science derives meaning from the word plex (interwoven or interconnected) and should not be confused with words like “complicated” (as a jet engine), “cumbersome” (as bureaucracy), “unwieldy” (as an aged empire), “chaotic” (as a disorderly administration) or “difficult to understand” (as a verbose document). It has a precise meaning – it is a property of a group, or a system, which is open and whose members intensely interact with each other without being centrally controlled.

 

It is well understood that natural ecology is a complex system – it is not centrally controlled; it self-organises to adapt to disruptive events such as climate change or a hit by a meteor, and perpetually evolves through the co-evolution of species. It is less well known that social, economic and technological environments in which we live and work are also complex and they, indeed, continuously change through the mechanisms of self-organisation and co-evolution. The current rate of change is very high.

 

This is in complete contrast to rigidly structured systems, such as planned-economy, command-and-control corporations, or large bureaucracies, which are centrally controlled and are not capable of self-organising or evolving and are therefore a poor match to the currently prevailing complex economic environment. 

As a consequence of fast evolutionary changes, the global economy at the time of Brexit, and immediately after Brexit, is going to be quite different from the economy, which was dominant in the last century.

 

How different?

 

The Internet-based global economy at the time of Brexit

Trading in knowledge is replacing trading in goods as the main business activity. Whilst after the Second World War manufacturing generated around 50% of the UK GDP, its current contribution is below 15%. At present, the prime wealth generating engine is the service sector, which contributes over 75% to the UK GDP (Wikipedia). 

The US is leading the West into a brave new world of knowledge economy, as is evidenced by the new business elite - Apple, Google, Amazon, Facebook, YouTube and Twitter, which have pushed the car and oil giants (remnants from the industrial era) onto the B list and have propelled IT pioneers, like Jeff Bezos and Bill Gates, to the top of the rich list.

 

EU is trailing behind. There are no Apples in Europe, only subsidised apples.

 

The world is currently engaged in building a giant global digital network, which will profoundly affect employment opportunities and change social structures. Four main subnets of the new global digital network can be identified as (1) the Internet of Documents, (2) the Internet of People, (3) the Internet of Things and (4) the Internet of Value.

 

The Internet of Documents is the oldest and well used. It contains practically all the documents, newspapers, magazines, journals, images, photos, videos, films, TV shows and works of art, which have been digitised and are available to everyone through powerful search engines such as Google. 

 

The Internet of People is now almost completed, having experienced the exponential growth in the 21st century. There are currently (May 2020) over 4.33 billion active internet users worldwide. 57 percent of the entire world's population has internet access. There are 3.9 billion unique mobile internet users worldwide, which makes up 51 percent of the global population.

 

Over 50% of the world population can potentially connect with each other to share personal or professional news and experiences through social websites such as Facebook, YouTube, LinkedIn and Twitter; to purchase books, clothes or technology online through Amazon and to access their current accounts, investments or savings, at any time and from any place. Practically all businesses have their own websites which serve as shop-windows but also as a means of trading and processing business transactions with unprecedented speed.

   

The Internet of Things (IoT) is at an early stage of development. The idea is to connect 50 billion physical objects to the global network to enable them to compete and/or co-operate with each other without involving their users. For example, IoT will enable driverless cars to negotiate with each other road priority, machine-tools in a factory to agree with conveyers how to improve production schedules, etc.

 

The Internet of Value is a new global project aimed at enabling rapid electronic transfer of value among users, bypassing banks and therefore reducing cost of the transactions. The plan is to store international digital currencies on secure servers, enabling owners to access their assets online and engage in rapid processing of investments and loans.

 

The global digital network, outlined above, connects individual citizen, ethnic and religious groups, nations and unions of nations into a vast and complex global village. 

 

The second important trend is related to the recent advances in Artificial Intelligence (AI). We are now capable of extracting knowledge from big data and building digital ecosystems which, just like natural ecosystems, self-organise and co-evolve with their environment and therefore are perfectly suited to operate in the current complex economic environment. 

 

Advances in AI will lead to a steep increase in the number of autonomous robots and driverless vehicles capable of performing tasks currently performed by human operators and intelligent software capable of performing clerical, professional and managerial tasks more effectively than humans.

 

As already stated, the co-evolution of society and technology is unstoppable. New technologies create new employment opportunities and they in turn change social, economic and political power structures, with new technology leaders (and those who manage to jump on the bandwagon) moving to the top.

 

Sceptics should be reminded of fundamental social structural changes, which followed the industrial revolution caused by the invention of mass-production technology. Where are now landowners who dominated the establishment only 200 years ago?

 

What are the consequences? 

 

  • The Internet makes geographical distances less important; every global villager is a potential customer, supplier or a political partner for other global villagers.
  • Information economy provides abundant new business opportunities in knowledge-based products and services, which do not depend on international trade agreements or common markets, as exemplified by the worldwide success of products and services developed by the leading new technology companies such as Apple, Amazon and Samsung.
  • Employment opportunities are changing drastically, a large market for developers of the Internet of Things, the Internet of Values, Big Data and Digital Ecosystems is being created, a market which is open to all who are sufficiently well-informed to notice it.
  • New types of highly skilled knowledge workers are in demand, occasionally with unusual names, like, knowledge engineer, drone photographer, gaming shoutcaster, social media manager, YouTube influencer, data analyst or machine learning developer, to mention just a few.
  • Geopolitical system is becoming a highly volatile network of religious groups, nations, unions of nations, military alliances and exceptional individuals competing or cooperating with each other and sometime changing sides and it is not surprising that under such complex conditions every military intervention initiated during this period failed to achieve its political objectives and managed only to create prolonged misery and distraction. 
  • Under conditions of market complexity, it is not so important to be big; it is important to be adaptive and resilient and open to new ideas.

 

Is UK better placed to take advantage of new market opportunities in or out of EU?

As outlined above, the transition from industrial to information economy offers a variety of opportunities, some of which would be difficult to realise if the UK stayed in the EU because European Union is moving, against the global trend, towards an increasingly rigid, centralised political structure at the time when complex global market favours distribution of decision making and flexibility. As a rule, smaller units are better at adapting to changes than large simply because decision makers are closer to issues that require decisions and very close to those that implement decisions.

 

The UK has an appropriate mix of strengths for succeeding in the new world of complexity.

 

What are the UK strengths?

 

  • English language — English people are the masters of the world-favourite means of communication. They can deliver knowledge-based services to international clients in their mother tong; UK universities, private schools and tutors are preferred precisely because they teach in English.
  • Knowledge and creativity — UK universities are highly ranked for research; UK scientists are among the best in advanced IT, genetic engineering, bioengineering, nanotechnology; British designers design German cars and American computers; BBC is the best known creative media brand around the globe; UK architects design spectacular buildings world-wide; British music industry outputs are in huge demand.
  • Knowledge-based services — UK sells knowledge as a service in many forms: as education, training, advice, management services, design services, media outputs and as advanced engineering products, such as Rolls Royce aircraft engines, in which knowledge created by research and innovation is wrapped in physical matter; similarly, in software services knowledge is wrapped in code; UK has skills in automation and protection from cyberattacks, which is another knowledge-based service in high demand.
  • Entrepreneurial culture — UK has many small to medium information technology companies collectively representing quite a formidable knowledge-based services sector. Start-ups and freelance entrepreneurs are everywhere; networks of small enterprises are new economic giants; London is one of the world leading centres of advanced IT startups.
  • Flexible employment law — UK has a growing, modern flexible gig economy and widespread zero contracts, which are essential ingredients of the fast-changing economy.
  • Minimal red tape — It took me 15 minutes and £20 to open and register a company in the UK; and it required several months to accomplish the same task in Germany, which also involved lawyers at a considerable expense.
  • Connectivity — UK is among the 5 most connected economies in the world, just behind, US, Singapore, Sweden and Switzerland, with Germany and France trailing. In 2019 the investments in the UK information technology businesses were higher than combined investments in French and German IT businesses.
  • City of London is one of the most important global financial centres.

 

The main challenge

The main challenge will be changing the mind-set of political and business decision-makers and their advisers who still see the world as it was in the last two centuries - the world of superpowers and big corporations where stability was maintained by political, military and trading alliances and treaties, privileges acquired by lobbying and political power was securely held by political parties and professional politicians. 

 

Remnants of this world order are of course everywhere, but not for long. Adapting to the new order as soon as possible can bring considerable advantages.

 

A strategy for Brexit

 

  • To convert its assets into money at the time of Brexit, UK needs to 
  • Foster all aspects of its strengths, as described above, and stop trying desperately to retain old industrial working conditions and practices.
  • Promote, at home and abroad, English language and culture, creativity, innovation, knowledge-based services, adaptability, flexibility, diversity and connectivity.
  • Support businesses developing high-quality knowledge-based services - future UK Apples - remembering that the American Apple did not need any custom union to conquer the world. For the foreseeable future, high-quality knowledge-based products and services will be in great demand.
  • Attract knowledge-based businesses to Britain; possibly by competitive corporate tax and favourable investment opportunities.
  • Attract skilled knowledge workers and students to Britain; UK needs talents from diverse backgrounds; and of course, those who receive education in Britain, or enjoy working in UK, will be the best UK ambassadors and promoters.
  • Convert enemies into trading partners — there is a room on this planet for every political dogma and every religion. New trading partners will help to earn money much needed to improve quality of life at home and, most importantly, strong trading links reduce probability of wars.
  • Help individuals that cannot keep pace with the harsh transition to knowledge economy. 

 

 

Global Warming

How can we contribute to the reduction of CO2 emission?

To reduce CO2 emission, it is necessary to drastically reduce transportation of people and goods across the globe using internal combustion engines. Switch to electric and chemical drives will certainly help. 

 

This essay will focus primarily on how to eliminate global supply chains and thus save energy and reduce the emission of CO2 by transforming traditional mass-production manufacturing plants into versatile manufacturing systems capable of cost-effectively producing a variety of goods in small or large batches and, most importantly, which could be located close to the points of consumption. Like “corner shops”.

What is versatile manufacturing?

 

The concept of versatile manufacturing system is new. It basically means that the manufacturing system can be rapidly reconfigured and rescheduled to switch from producing one type of products to another.

 

The production is in batches, which may be small (even a batch of one), medium or large. The goal of a versatile manufacturing system is to meet demand for a range of manufactured goods in the proximity to the manufacturing plant. 

 

A manufacturing system is versatile if it is capable of selforganising in response to a request for changing its product type, by instantly detecting whether the requested product type is within its portfolio, rapidly identifying which manufacturing resources are required, and configuring and scheduling required resources. In other words, behaving like a complex system.

 

Production profile is a list of product types, which a versatile manufacturing system is capable of manufacturing. It defines the range of products of a versatile manufacturing system.

 

Product type is a type of the product that a versatile manufacturing system is configured and scheduled to produce (e.g., type of electric motors or batteries for electric cars).

 

Selforganising means autonomously (without human intervention) changing system resource configuration and/or schedule, e.g., switching from producing batteries type A (for a car brand X) to producing batteries type B (for a car brand Y).

 

It goes without saying, to build versatile manufacturing plants, as defined above, is only possible using digital intelligence – intelligent robots managed by artificial intelligence systems.

 

Background

In developed countries the standard practice has been to outsource manufacturing to developing regions, where low wages and favourable investment environments ensured competitive prices for manufactured products. Additional costs of global supply chains and long-distance transportation of manufactured goods were, and still are, relatively small. Such an arrangement seemed to be a win-win situation. Both the developed and developing countries have gained.

 

However, recent widespread concerns about manmade pollution and climate change, have made the outsourcing of manufacturing to faraway parts of the planet untenable - transporting huge volumes of goods from one end of the world to another is wasteful of energy and harmful to the environment.

 

Manufacturing will have to be located as close as possible to consumption. As it happens, we have technology, which could enable this transformation. The problem is with attitudes of decision makers.

 

Generations of economists and engineers have been brought up to accept without questioning the notion that the critical success factor for manufacturing is the economy of scale. This notion was correct under conditions of stable markets and predictable demands, the situation that prevailed in the 20th century. The dynamics of markets in the 21stcentury, however, has drastically changed.

 

The Internet has become the backbone of the new digital economy. The Internet-based global market spans the whole planet and is populated by billions of suppliers, consumers, middlemen, dealers, brokers, consultants, investors, bankers, insurers and retailers, engaged in making, breaking or modifying business transactions with unprecedented speed and frequency. As a consequence, the market is so complex that it is no longer possible to forecast demands and supplies with any certainty, whilst human errors, failures of resources, delays, fraud and cyberattacks are on the increase.

 

As dynamics and volatility of the market increases, we have to accept that the critical success factors for businesses cannot stay the same. Big, hierarchical corporations, designed to take full advantage of the economy of scale, have rigidity, caused by long delays between the detection of a change in demand and the decision on how to reschedule resources to meet the identified change, which prevents them to rapidly adapt to unpredictable changes in demand and supply. 

 

It is quite obvious that under conditions of frequent, unpredictable changes in demand and supply and frequent occurrence of various disruptive events, the critical success factor must be adaptability, the ability to detect any disruptive event, as it occurs, to rapidly identify the part of the business that will be affected by the disruption, and to eliminate, or at least reduce consequences of disruption by rescheduling the affected resources. Since only complex systems can be adaptive, the clear conclusion is that manufacturing systems must be designed to be complex (for adaptability), rather than big and rigid (for the economy of scale).

 

Once the decision makers accept this premise, it will be possible to start designing small, adaptive and versatile manufacturing units and locate them close to the demand for their products. 

 

The key to achieve results is changing old-fashioned mindsets of decision makers. 

 

Transforming a conventional manufacturing plant into a versatile manufacturing system

The basic premise of complexity science is that only complex systems can operate effectively within a complex environment. Rigid structures of deterministic systems tend to crumble when placed under conditions of complexity. In contrast, complex structures easily adapt to changes and are resilient to attacks.

 

It follows that manufacturing businesses which operate in a complex environment should be designed to behave as complex adaptive systems.

 

Since frequency of unpredictable disruptive events in the global market is high, the detection, impact analysis and rescheduling must be done rapidly – in real time – which can be achieved only by employing intelligent and fast decision-making technology, such as AMAT (Adaptive Multi-Agent Technology), which is briefly described in the Appendix. 

 

Manufacturing business processes are, in general, connected, therefore the adaptation in one process will cause ripple effects in other processes, which will have to be dealt with. The implication is that business processes in adaptive manufacturing systems will be continuously selforganising in order to adapt to disruptions.

 

Multi-agent real-time schedulers are capable of cooperating, or competing with each other if necessary, ensuring that manufacturing processes are interacting with each other.

 

The second step is to turn an adaptive manufacturing system into a versatile. For this, it is necessary for all manufacturing resources, e.g., robots, machine tools, conveyers, transporters, etc. to be reconfigurable which, for a portfolio of similar product types, is not that difficult.  

 

On top of that, it is necessary to build a second layer of selforganisation in response to a request for switching from one product type to another, which amounts to instantly detecting whether the requested product type is within system portfolio, rapidly identifying which manufacturing resources are required, and then configuring and scheduling required resources. The author’s team has a considerable experience in designing various types of selforganisation using in-house developed Complex Adaptive Technology.

 

Reinventing manufacturing, as described in this paper, on a world-wide scale, would have a considerable positive impact on the environment. However, the probability of changing mindsets of decision makers, which would enable such a project to take off, is minimal.

 

Appendix: An Overview of Complexity Science and Engineering Fundamentals

Introduction

The father of complexity science is the winner of the 1977 Nobel prize for Chemistry, Ilya Prigogine, from the Free University of Brussels. Two of his books are particularly relevant, “The End of Certainty: Time, Chaos and the new Laws of Nature” and “Is Future Given?”. 

 

By the end of the last century, observing sharp rise in volatility and dynamics of the Internet-based global market, making forecasts of demand and supply next to useless, I was relieved to find in Prigogine’s publications a coherent set of new concepts and principles, “new laws of nature” as he called them, explaining behaviour of complex irreversible processes, which later acquired the name of complex systems.

 

In the USA, the Santa Fe Institute developed their own variety of complexity science. Books from this school, which I have found particularly valuable are “At Home in the Universe: The Search for the Laws of Self-Organization and Complexity” by Stuart Kauffman, and “Hidden Order: How Adaptation Builds Complexity” by John Holland.

 

20 years ago, I have teamed up with a well-known Russian scientist and entrepreneur, Petr Skobelev. Together we have designed, built and commissioned many large-scale complex systems for commercial clients in the UK, USA, Germany, Denmark and Russia. Most importantly, we have used these unique complex systems for research, observing their behaviour and, based on observations, refining our knowledge of concepts and principles of complexity science. Results of our research were published in the book “Managing Complexity”, WIT Press, Southampton and Boston, 2014. Fundamentals of complex systems outlined below are an updated version of research presented in the book.

 

Complexity science fundamentals

A complex system is an open system that consists of large numbers of diverse, partially autonomous, richly interconnected components, called agents, which cooperate or compete with each other without central control, and whose behaviour emerges from the interaction of agents and is uncertain without being random – it follows distinct patterns.

 

Using uncertainty as a criterion, we can divide all systems into 3 classes: 

 

Deterministic, whose behaviour is predictable (uncertainty = 0). Examples include man-made systems such as cars, aircraft, bridges and clocks, as well as closed models of physical system such as pendulums and planetary movement. 

 

Complex, whose behaviour emerges from the interaction of agents, it is not predictable in detail and yet follows discernible patterns (1 > uncertainty > 0). Examples include natural ecology, social, political and economic systems, businesses and markets.

 

Random, whose behaviour is completely unpredictable (uncertainty = 1). A good example is movement of molecules.

 

Seven key properties of complex systems differentiate them from rigidly structured, man-made deterministic systemssuch as buildings, bridges, aircraft and conventional algorithms. 

These properties are 1) connectivity, 2) autonomy, 3) emergence, 4) nonequilibrium, 5) selforganisation, 6) coevolution and 7) nonlinearity.

 

Connectivity

Connectivity denotes the number of connections an agent has with other agents. Complexity of the system increases with connectivity. 

 

For example, in the human brain connectivity of a neuron is typically over 1,000, whilst in a natural ecosystem, say, a small forest, the estimated connectivity may be closer to 100. 

 

The strength of connections between agents also plays an important role. Complexity of the system is inversely related to the strength of agent connections. When connections are weak, they can be easily broken, and new ones created, thus increasing complexity. Systems characterised by strong connections have a lower complexity.

 

Autonomy

Agents have partial autonomy (freedom of behaviour), limited by constrains imposed by the system to which they belong. Constrains are specific to the domain in which system operate, and include norms of behaviour, traditions, habits, rules, regulations, laws of the land and, for physical systems, natural laws. 

 

For example, a leader of the pack of wolfs, when challenged by a younger pretender, has only two options, to fight or retire. Humans on the other hand have normally much greater autonomy of behaviour, which makes human communities more complex, unless their behaviour is strictly controlled (dictatorship, control-and-command management).

 

Emergence

Emergent properties of a complex system are properties that cannot be found in any of the constituent components – they emerge from the interaction of agents.

 

Emergent Intelligence

The most potent example is natural intelligence, a property that cannot be found in any brain neuron – it emerges from the interaction of neurons.

 

It follows that if we can build a sufficiently complex digital system, artificial intelligence will emerge, when required. This hypothesis was corroborated in all complex digital systems that the author’s team designed for commercial clients. Digital agents, when acting in groups, exhibited capability to make decisions under conditions of limited uncertainty as good as decisions made by humans. 

 

Currently popular version of artificial intelligence is rather different – It is based on artificial multi layered nets, which must be trained using large quantity of data and expensive training algorithms. Its performance critically depends on the quality of training data and it cannot self-improve, adapt or evolve. Nevertheless, when training data is of really good quality, it is very powerful.

 

Nonequilibrium

Complex systems operate in an unstable mode far from equilibrium. The system, when disturbed, has no propensity to return to the earlier operating point - it moves from one operating point to another with ease.

The nonequilibrium behaviour is particularly useful for systems operating in unpredictable hostile environments. With this in mind, military aircraft are designed to be unstable under normal operating conditions to enable them to rapidly change the course when attacked. In contrast, civil airliners, whose environment is highly regulated and managed by air traffic controllers, are designed to operate in a stable equilibrium state.

 

Selforganisation

Selforganisation is the capacity of a system to autonomously (without external intervention) change its own configuration and thus reach a new order. 

 

Selforganisation plays an important role in adaptation, conflict resolution, spontaneous self-improvement, creative destruction and clustering, as described below.

 

Adaptation

Adaptation is the capacity of a complex systems to selforganise with the aim of eliminating, or at least, reducing undesirable consequences of a disruptive event.

 

A complex system adapts by (a) detecting an unpredictable disruptive event (internal or external change, fraud or attack), (b) identifying parts of the system that are affected by disruption, and (c) rescheduling available resources of the affected parts of the system and, if necessary, deploying new resources without disturbing the unaffected parts of the system.

 

For example, in a multi-agent system for managing 2,000 London minicabs, designed by the author’s team, as street congestion changes, the adaptive real-time scheduler re-examines every few seconds which vehicle is best suited to collect a customer, maximising values for the minicab company, client and the driver.

Adaptation increases value of a system by increasing its resilience to disruptions, mistakes, fraud and electronic attacks.

 

Conflict Resolution

Conflict resolution is the capacity of a complex system to arrive at a mutually agreed resolution of a situation in which two or more agents representing demands request access to the same set of resources at the same time.

 

A complex system resolves a conflict by trial-and-error, (a) assuming that certain adjustment in a demand will resolve the conflict, (b) proposing the adjustment, (c) if necessary, modifying the previous assumption, and (d) repeating steps b and c until  a mutually agreed resolution of the conflict is achieved or resources available for conflict resolution run out.

 

Spontaneous Self-Improvement

Spontaneous self-improvement is the capacity of a complex system to selforganise with the aim of improving its own performance whenever the appropriate resources are available. In other words, complex systems have propensity to improve own performance.

 

A complex system self-improves by trial-and-error, (a) assuming that certain activity will improve system performance, (b) performing that activity, (c) if necessary, modifying previous assumption, and (d) repeating steps b and c until the desired performance improvement is achieved or resources available for self-improvement run out.

 

Creative Destruction

Creative destruction is the capacity of a complex system to destroy a part of itself and then rebuild itself from the beginning, when it realises that its performance cannot be improved piecemeal.

 

In adaptive multi-agent systems, designed by the author’s team, agents constructing a schedule, if stranded in a local optimum, may decide autonomously to destroy all what they accomplished and start from the beginning, following an alternative approach.

 

Clustering

Clustering is the propensity of a complex system to selforganise itself into clusters (communities, swarms) of agents with high mutual connectivity and with sparse connections between the clusters.

 

The most obvious example is the tendency of animals, birds and insects to selforganise into herds, pucks, flock, colonies or swarms. Humans behave likewise by selforganising into families, communities and nations.

 

Coevolution

Coevolution is the capacity of a system to irreversibly adapts to changes in its environment and, in turn, by changing through adaptation, affect its environment. As a result, both the system and its environment change in perpetuity. System environment is defined here as the totality of all systems interacting with the system.

 

Coevolution ensures sustainability of an ecosystem. For example, consider a smart city equipped with pollution sensors and adaptive transport schedulers. When sensors detect an increase in pollution, the schedulers autonomously increase the congestion charge for diesel vehicles, which decreases the number of these vehicles travelling through the city and, in turn, reduces pollution levels.

 

Perhaps the most striking example of coevolution is the interplay of technology and society, leading to drastic changes in both – agricultural technology pushing the transformation of society of hunters and gatherers into an agricultural society, massproduction technology forcing the agricultural society to change into industrial society and, recently, digital technology driving us into information society. 

 

With every step complexity of both society and technology has increased substantially, A clear indication that evolution favours complexity.

 

Nonlinearity

Connections among agents are, as a rule, nonlinear and may include amplification, acceleration, the auto-catalytic property, as well as positive and negative feedback loops. The nonlinearity of agent connections may produce unforeseeable and truly remarkable extreme events.

 

Butterfly Effect

Under certain conditions, a disturbance, as small as motion of a butterfly wing, occurring at one point on the planet may cause an extreme event, as dangerous as a tornado, at a faraway place or, indeed, on a truly global scale. 

 

The most striking example of the butterfly effect is the current pandemic – it began by one person eating infected bat (or, maybe, by one mistake made in a laboratory investigating Corona virus) and resulted in millions of people infected worldwide, huge number of deaths and a global economic crash. 

 

A Drift into Failure

A drift into failure is the accumulation of small, insignificant mistakes, omissions or fraud which, unless detected and stopped before reaching the tipping point, cause an extreme event – a failure. 

 

The best example of a drift into failure is the recent global financial crash, caused by a very large number of small toxic loans issued by banks over a number of years to clients known to be unable to repay them. It is important to note that a drift into failure can be controlled by deploying adaptive multi-agent systems to detect and prevent repeated mistakes, omissions or fraud. 

 

A Paradigm Shift

A paradigm shift is a stepwise process by which evolution advances. As world evolves, new problems emerge which cannot be solved by the well-established paradigm. New candidate paradigms compete with each other until one wins and becomes dominant. 

 

Two Worldviews: Two Mindsets

Many of us were brought up to believe in deterministic world in which everything is predictable. 

A complex world in which future is uncertain is, however, a much more realistic proposition when we look carefully at the life around us.

 

DETERMINISTIC WORLD

  • World is created according to a “grand design”
  • Future is predictable
  • There is universal law which predicts everything. We are waiting for it to be discovered
  • Uncertainty = 0
  • If we feel uncertainty tis is due to our lack of knowledge about the world
  • As we learn our uncertainty decreases
  • Time and space invariance
  • Cause-effect relationship once established is always valid

 

COMPLEX WORLD

  • The planet Earth irreversibly and unpredictably evolves from primordial soup to global village
  • FUTURE IS NOT GIVEN
  • Future emerges from billions of interactions among constituent agents
  • 0 < uncertainty < 1
  • Uncertainty is an inherent property of complexity
  • Uncertainty cannot be reduced by learning
  • “Becoming” is as important as “being” – to understand present you must know how we got here.
  • Cause-effect relationship cannot be established – there are too many small events that over time contribute to the “effect”.

 

You will find the most impressive thinkers, philosophers and scientists on both sides of the debate.

 

DETERMINISTIC WORLD

  • 384 BC Aristotle - Deductive logic; Classical physics; the world consists of matter and energy (he ignored information)
  • 1643 Newton - Natural laws are valid at any time and at any location
  • 1879 Einstein - “God does’nt play dice with universe”; “Time is an illusion”

 

COMPLEX WORLD

  • 600 BC Buddha - “The mind is everything. What you think you become”
  • 535 BC Heraclitus - “You could not step twice into the same river”
  • 1831 James Clerk Maxwell wrote about “a new kind of knowledge” that would overcome the prejudice of determinism
  • 1902 Karl Popper-In a deterministic world it would be impossible to choose or create
  • 1917 Prigogine – father of Complexity - “The end of certainty”; “Future is not given

 

Fundamentals of complexity engineering: digital ecosystems


What are digital ecosystems?

According to Wikipedia, “A digital ecosystem is a distributed, adaptive, open socio-technical system with properties of self-organization, sustainability and scalability inspired from natural ecosystems. Digital ecosystem models are informed by knowledge of natural ecosystems, especially for aspects related to competition and collaboration among diverse entities”.

 

The Wikipedia definition of digital ecosystems is good because it captures the key element of the new concept. According to this definition, a digital ecosystem is:

 

  • Socio-technological system (rather than purely technological or social) - constituent agents have either human intelligence (HI) or artificial intelligence (AI)
  • Characterised by distributed decision making (rather than centralised) - constituent agents are empowered to make certain decisions and may cooperate or compete with each other
  • Adaptive (rather than rigid, hierarchical, command and control) – capable of autonomously eliminating, or at least reducing, consequences of unpredictable disruptive events 
  • Selforganising (rather than organised and, from time to time, reorganised by managers)
  • Sustainable – capable of coevolving with its environment over a long period of time
  • Scalable – capable to grow in size and/or performance

Clearly, this definition places digital ecosystems into category of complex systems.

 

Why do we need digital ecosystems?

Natural ecosystems are exceptionally good at selforganising to adapt to unpredictable events and change to accommodate changes in their environments. The ability to selforganise and coevolve helped rivers, oceans and some forests and grasslands to continue existing for millions of years.

 

We now have digital technology, which could be used to design artificial ecosystems capable of selforganising and coevolving. Why not use this powerful tool to transform traditional businesses and administrations into organisations capable of rapidly adapting to unpredictable disruptive events and irreversibly evolving with their environments? 

 

Architecture of a digital ecosystem

The architecture of a digital ecosystem whose human, physical, financial and knowledge resources are allocated to demands by a network of adaptive real-time schedulers consists of 

 

  • Real-world, where human, physical and financial resources are engaged in business, including all human decision makers.
  • Virtua-world, where digital agents are engaged in real-time scheduling of real-world resources.
  • Knowledge base, where knowledge resources are stored (knowledge how to run the business). 
  • Interfaces between the two worlds, which are primarily concerned with the transmission of information about disruptive events from the real to virtual world and instruction how to reschedule human, physical, financial and knowledge resources, from virtual to real word.

The process of transformation of a conventional business to a digital ecosystem could be done with a minimum of disruptions by following an evolutionary transformation methodology.

 

Technology for building digital ecosystems


Selecting technology

Clearly, we need technology capable of building complex adaptive systems in software, which means, capable of supporting:

  • A large number of digital agents (short algorithms) capable of exchanging messages with each other and with their environment, competing or cooperating among themselves, without being centrally controlled. 
  • Organisation of agents into networks capable of competing or cooperating with other agent networks and/or networks of real agents (networks of human, physical and financial resources of the domain of the real world with which digital agent networks interact).
  • Domain knowledge, which defines the content of messages and imposes constraints on agent autonomy, as well as data on the domain of the real world with which digital agent networks interact.
  • Interfaces between digital agent networks and real agent networks.

 

Technology that matches the above criteria is advanced multi-agent technology backed by extended ontology, which was named complex adaptive technology or CAT.

 

 

Print Print | Sitemap
Copyright George Rzevski