Global Society as a Problem Solving System: How We Lost Control and How We Might Regain It   By Tim Gooding   This document and the NetLogo simulations related to this paper copyright Tim Gooding 2009 to 2011. (tim@humansos.org)

 

Contents

Part I 3

Introduction: The Problem and the Promise. 3

Comparing Two Assumptions: The First Assumption. 3

The Second Assumption. 4

The Argument: Global Society as a Problem-Solving System... 7

The Evidence: Examples of Unprogrammed Realistic Trends That Have Emerged.. 9

·       The shape of global population growth. 9

·       In a high growth technological society, wealthier agents tend to have fewer children. 9

·       As the population industrialises and grows, fecundity tends to drop even if agent fertility rises. 9

·       Once industrialisation takes over, ‘farmers’ tend to be reduced to a single digit 10

·       Population changes broadly match the economic events. 10

·       During the technological boom, agents tend to become homogeneous. 11

·       During the technological population boom, one ‘culture’ comes to dominate globally. 11

·       Trading societies tend to organise themselves so that sharing is more a part of their social structure. 12

·       In thriving economies, individual agents tend to lose their ability to find resources on their own. 12

·       Previously stable local structures are broken apart 12

·       Particular populations of agents die as a result of technological societal disruptions. 13

·       During the technological boom, starvation is high. 13

·       Same peaked top before the decline as known historical human population collapses. 14

An Analysis: Why is Starvation So High When There is So Much Food?. 15

Contrasting Theories: Free Market Economics and the Wealth Creation Imperative. 17

Part II 19

Detailed Technical Description of the Simulation.. 20

The Basic Assumptions of the Societal Simulator: 20

The Patches (Environment). 20

The Agents (people). 20

Miscellaneous. 20

Programme Overview... 21

User Control of the Four States. 22

Simulation Controls. 23

User Control Over Agents. 23

User Economic Control 24

User Population Control 24

User Control Over the Environment 25

Miscellaneous Sliders. 26

Miscellaneous Buttons. 27

Evolutionary Variables. 27

Programme Graphing and Monitors. 28

Standard Graphs and Monitors. 28

The Evolutionary Graphs. 30

Some Details on How the Simulation Works. 31

How Trading Works. 31

How Reproduction Works. 31

How the Agents Move. 32

Some Last Notes. 32

Part III 33

Conclusion.. 33

Further Research: What’s Next?. 34

Free Will Within an Evolutionary System.. 34

Preliminary Work in Deriving Evolutionary Imperatives From First Principles. 35

Changing the Evolutionary Imperative. 36

Evolutionary System Comparative Chart 38

Part I

Introduction: The Problem and the Promise

Theories on how society works can evoke strong feelings in even the most logical and knowledgeable people; suggesting that human nature and intention might not be dictating society’s direction seems to be especially provocative. However, the possibility that the shape of global society is more influenced by self-organisational processes rather than the aggregate nature and endeavour of human beings not only comfortably explains global reality it also offers us powerful new tools with which to tackle some of the most serious problems facing the world today.

Comparing Two Assumptions: The First Assumption

First, let us consider the facts of a world as a creation of human nature and endeavour:

• High rates of starvation in the presence of large food supplies (58% of all yearly deaths[1]).
• Rapid world-wide eco-system decline. (31% decline between 1970 and 2003[2] which continues with edible fish predicted to disappear from the oceans by 2050[3])
• Exponential population growth that necessarily lowers the per capita availability of non-renewable resources[4] and destroys renewable resources[5].
• Building society on a foundation of non-renewable resources (non-renewable means they will predictably run out) while simultaneously pressuring our only long-term support, the renewable system, out of existence.
• Possible catastrophic climate change.
• Low prices on a vast range of goods and services. This, of course, is what we get in exchange for the above.

We cannot plead ignorance because these issues were foreseen centuries ago – the most recent, the possibility of human driven climate change, was accurately foreseen in scientific literature in 1896[6]. This begs a question: if today’s problems are the result of human nature and endeavour in the presence of accurate foresight, what has changed that we can now trust it to act in our collective self-interest?

On the other hand, many people feel that the nature of global society does not reflect their own nature or desires[7]. If the society’s character has been disconnected from individual character, then from where does the nature of global society emerge?

The Second Assumption

Artificial evolution is used by scientists and engineers to solve problems. They ‘aim’ artificial evolutionary systems by designing a fitness test that guides the forces of evolution to solve a specific problem. The evolutionary solutions for even simple problems, such as sorting numbers, are extremely efficient in relation to the problem and can be beyond human comprehension[8]. Artificial evolution is also used to programme elements, such as realistic human action and response, that would be far too complex to programme through conventional means. Again, the solutions are very efficient and beyond the understanding of the people who designed the evolutionary system[9].

Today, successfully setting up problem-solving evolutionary systems has become somewhat routine. This experience allows us to see that modern human society is set up perfectly to initiate an evolutionary problem-solving process aimed at wealth creation[10]. Unless it can be demonstrated that a system set up to evolve can avoid entering an evolutionary process, it is reasonable to assume that global society has possessed the qualities of an artificial evolutionary system aimed at wealth creation for some years now.

If global society is acting as a wealth creating evolutionary system, our knowledge of evolutionary systems suggest these characteristics would apply:

• Evolutionary solutions would emerge that are beyond human comprehension[11].
• These solutions would be applied towards increasing wealth creation and overcoming any blocks to wealth creation.
• Random factors are important to the successful function of evolutionary systems. When programming an evolutionary system, there is no functional difference between free-will and random action. Free-will would aid the evolutionary process.
• The nature of what eventually emerges from an artificial problem-solving evolutionary system is certain. If the problem is to sort numbers, the result will be excellence in number sorting. If the problem is creating wealth, the result will be excellence in wealth creation.

When we look at the world by assuming that global society is functioning as an artificial evolutionary system with a wealth creation imperative, world facts make a great deal of sense:

• Wealth creation will not regard human welfare except as it relates to making more wealth. As such, people not creating wealth will become irrelevant. Therefore, it is predictable that starvation will rise as global society progressively optimises itself to better create wealth. At the peak of modernity, approximately 3 out of every 5 human deaths comes as a result of a lack of food while the richest countries in the world (the best wealth creators) struggle with obesity.

• The eco-system supports human beings. A healthy eco-system is more resourceful than an unhealthy one. However, more wealth can be created today by converting the eco-system into wealth. Where profit cannot be found, the wealth imperative sweeps aside the ecosystem in order to make room for successful wealth creating activities. It becomes inevitable that the earth’s eco-system will decline. Any agreements to slow this optimisation process will go against the evolutionary imperative and will therefore be disempowered. For example, in 2002, an international agreement was signed to stop the decline of the eco-system. In 2010, a scientific study concluded the agreement made no impact at all[12].

• No matter how influential the group, going against the wealth creation imperative will activate evolutionary solutions. With this in mind, consider the climate change debate. Scientific evidence and debate has been ‘politicised’ and thus ‘climate-gate’ became international news even though the actual events were trivial[13]. Regardless of how they feel personally, those in power continue to resist any wealth diminishing solutions because circumstances continually pressure them to avoid any official action that goes against wealth creation. Meanwhile, large sectors of the public[14] disagree with the majority climate experts. In other words, much of the public feel confident they understand the world climate system better than the majority of climate experts wherever it affects their ability to create or maintain wealth. In the decades since the alarm was first raised in mainstream science, nothing meaningful has been done against the possibility of climate change.

• As long as we have no access to other resource rich planets, we live in a closed energy system called Earth. Closed systems have predictable restrictions on resources and energy. This creates two realities: 1) Once the resource acquisition passes the capacity of renewable sources, the overall per capital wealth potential will decline directly in relation to increases in the world population. However, because more people can make more wealth today, the wealth imperative creates pressures to increase the world population regardless of the inevitable consequences[15]. 2) Accessing any resources at an unsustainable rate will exceed hard lines that will lead to excessive growth in relation to resources available over the long term. The resulting growth bubble will predictably burst once the damage of exceeding those hard lines is made manifest (the financial crisis foreshadows something far bigger). However, accessing resources at a rate that exceed these hard lines leads to more wealth today and thus this behaviour is empowered by societal circumstances that are solutions created by the evolutionary system.

• Low prices on a vast range of goods and services would be a by-product of the optimisation of wealth creation and would therefore emerge automatically. No human planning is required. All other concerns will be secondary.

In this case, human potential is being constrained by circumstances generated by a wealth imperative. Individuals are empowered wherever they create wealth and progressively disempowered in all other places. Even brilliantly designed social or environmental initiatives will be countered by evolutionary solutions if they interfere with wealth creation.

If artificial evolutionary forces are present in global society, the current state of the world is a reflection of an evolutionary imperative, not human nature or human endeavour. More importantly, because we know a great deal about how to construct and manipulate artificial evolutionary systems, changing the nature of global society is a matter of applying practical and proven knowledge.

The Argument: Global Society as a Problem-Solving System

If we were to define ‘agent’ to be any societal entity to which a balance sheet and income statement (or profit and loss statement) could be applied, then agents would comprise governments, charities, businesses, universities, organised crime, households, government departments, individuals, etc. Few people would dispute that global society is populated by such agents or that these agents must respond to financial realities if they are to economically survive. Furthermore, an agent’s power in society changes in proportion to the wealth[16] it can wield. Accordingly, any practice or tool that advances wealth creation rapidly propagates throughout the society. This, in turn, raises the bar to which all agents must aspire if they do not wish to experience economic decline or extinction. This is an ongoing cycle.

If these elements were placed in a computer environment, it would create an artificial evolutionary system designed to solve the problem of wealth creation. Unless it can be demonstrated that a system possessing the conditions necessary to initiate evolution can avoid entering an evolutionary process, it is probable that human society is a problem-solving evolutionary system.

Simple artificial evolutionary systems empower agents according to their relative performance in regards to a single fitness test. The test is applied to all agents throughout the system creating a system wide force that leads to a predictable outcome. For example, if the fitness test empowers better number sorters, over time an ever increasing ability to sort numbers will emerge. Even these simple evolutionary systems create solutions beyond human comprehension[17].

In human society, agents who are relatively strong in wealth are relatively strong in power. This applies throughout the world. Relatively successful wealth-creating characteristics rapidly propagate throughout society through the adoption of the successful behaviour such as ‘best practice’ or efficiency. High value is placed on practices leading to economic success whereas caring is worth little. For example, assuming salaries are related to societal value then stock brokers and lawyers are far more valuable to society than care workers[18]. One group services the health of wealth creation while the other services the health of human beings. Regardless of how people feel about this, in today’s society, wealth creation is far more important to society than health creation and accordingly commands a great deal more power. This is exactly what would happen if society is being shaped by an evolutionary system optimising for wealth creation.

An evolutionary imperative does not only work on individual agents, it also shapes the overall organisation of the agents. James Lovelock explored how this can work when he described how life comes together to optimise the world’s climate in order to create environmental conditions best for life[19]. In a similar way, societal agents automatically work together to create the social circumstances necessary to best promote wealth creation. It is possible that such a process can so powerfully influence circumstance that any wealth creation activities can quickly be elevated to a necessity, regardless of any potential cost to the health of society or life.

For example, the use of oil creates great wealth but at the risk of possibly destabilising the earth’s climate[20]. If we assume society has entered a wealth creating evolutionary system, the former directly serves the evolutionary imperative while the latter is irrelevant[21]. People and organisations trying to significantly slow or stop the use of oil will be seen as a problem to which evolutionary solutions will be brought to bear. Solutions will include circumstances that make real opposition very difficult. In fact, even with a universal consensus to stop burning fossil fuels today, it would be impossible to actually do so without society collapsing. That is a very powerful circumstantial pressure to continue burning fossil fuels today no matter what the beliefs are about tomorrow. The majority will comply of their own free will and wealth creation will continue.

In this paper, two possibilities are being compared. If society is being shaped by forces of human nature and endeavour, then it will be impossible to accurately simulate even a small portion of society without realistically modelling all the rich complexity of human beings. On the other hand, if an evolutionary system is dominating the shape of human society, then building an accurate societal simulation becomes a matter of determining the few simple rules from which societal complexity emerges. The majority of human detail can be omitted because human character, human nature, and human intent will be largely irrelevant. In this case, success will be evident by the extent to which realistic societal behaviour and historical accuracy emerge.

The section called Examples of Unprogrammed Realistic Trends That Have Emerged summarises some results from such a simulator and concludes by examining some of the reasons why the global percentage of deaths by lack of food is reported by the UN to be 58%. Technical Simulation Description offers detail of how the programme is designed and operates. Conclusions briefly outlines some other areas of development that will follow as time and money allow.

The Evidence: Examples of Unprogrammed Realistic Trends That Have Emerged

None of the trends seen in these graphs were the result of programming. The following detail emerged from a few simple rules designed to capture society’s evolutionary forces. All the simulation graphs presented here are screen shots from actual simulation runs. See Detailed Technical Description of the Simulator for more information concerning the graphs and the simulator.

• The shape of global population growth tends to match historical graphs.

 
Graph courtesy Wikipedia, retrieved from:
 http://en.wikipedia.org/wiki/File:Population_curve.svg

• In a high growth technological society, wealthier agents tend to have fewer children. The red line graphs how many children poorer agents have had while the turquoise line graphs the richer agents’ children, each as a percentage of their respective populations.

• As the population industrialises and grows, fecundity tends to drop even if agent fertility rises. Notice how dropping fecundity does not influence global population growth.

• Once industrialisation takes over, ‘farmers’ tend to be reduced to a single digit percentage of the population. The number is the percentage of farmers (i.e. 5%) at this point indicated by the end of the population graph. The simulation usually starts with a farmer population of around 50% and then fluctuates in a totally random fashion until the technological development takes hold. Once the non-renewable resources become scarce, the ‘farmers’ reassert themselves.

·        Population changes broadly match the economic events associated with historical global population changes. The purple line in the # of Trades / capita chart indicates the number of trades and stability of ‘technology’ in society. The green line indicates trades in needs such as food and water. Just as in reality, as the technological presence and stability exerts itself, the population responds. There are many other technological indicators in the simulation but they are not graphed and can only be seen while watching the simulation run.

     (Surplus Labour)

• During the technological boom, agents tend to become homogeneous and specialised to technological economic success. The Trading Culture 1 graph indicates ‘cultures’ that seek non-renewables over renewables. This graph shows the 12 ‘cultures’ as a percentage of 100% (‘1’ in the graph) such that the vertical values of the 12 lines added together will always add to 1. Notice the steep decline in the ‘cultures’ that prioritise renewable needs over non-renewable technology (Trading Culture 2).

• During the technological population boom, one ‘culture’ comes to dominate globally. As seen in the above graph, only a single ‘culture’ dominates. On the other hand, before the technological boom, cultural diversity is much more mixed, as seen below.

• Trading societies tend to organise themselves so that sharing is more a part of their social structure than more advanced economies. Societal giving tends to be reduced once the technological boom takes hold, even if individual agent generosity rises. If the simulation is run far into the future, the habit of giving can remain low even though the need increases dramatically once non-renewable resource scarcity asserts itself. It depends on what is lost during the population collapse. The Charity Activity chart tracks global charity activity while the Charity chart tracks the averages and extremes of individual agent choice of how much to give in charity.

• In thriving economies, individual agents tend to lose their ability to find resources on their own. The lower horizontal line in the Nature Vision indicates the point where individual agents no longer possess environmental vision. The ability to ‘find’ resources becomes a function solely of social organisation (also not programmed) or ‘luck’.

• Previously stable local structures are broken apart in favour of system-wide production efficiencies during the technological boom. There are four distinct societies and each is established on different rules. Each kind of society stabilises in a slightly different way and when a new system creates disruption frequently, the old ‘communities’ are either destroyed or the old agents are killed off in favour of the agents aligned with a new way of doing things, much as the indigenous people of the Americas were disrupted and destroyed to make way for wealth creation. In my simulation, the transition to a non-renewable society is by far the most disruptive. This can be observed by watching the simulation running.

• Particular populations of agents die as a result of technological societal disruptions. The dip in population occurs when more technologically advanced agents impose their organisation on areas already populated by other cultures. These cultures tend to die out, much like the First Nations (Native Americans) in the hundred years after Europeans first encountered them. Approximately 100 million First Nations people died[22] which is just under 1/5 of the world’s estimated population at the time (see chart on the right). However, it seems many official historical world population estimates do not reflect this decline. Notice the sharp dip in the simulated chart just before the boom whereas only the population decline due to the black plague is recorded on the official chart.

Chart retrieved 29/01/11from
http://www.globalchange.umich.edu/globalchange2/
current/lectures/human_pop/human_pop.html

• During the technological boom, starvation is high even in the presence of a rapidly increasing food supply. According to a 2001 UN report, approximately 58% of all human deaths each year are caused either directly from hunger or indirectly through becoming vulnerable due to critical nutritional deficiencies[23]. The chart called Per Capita Excess Needs shows how much ‘food’ is available for each agent in the simulation. Even though the food supply dramatically rises, starvation rises sharply. Since there are no politics, social classes, laws, or conspiracies in the simulation, a system-wide force must be arising that involves none of these things. This is examined further in the next section.

• When non-renewable scarcity hits a critical point, the shape of the simulation population collapse demonstrates the same peaked top before the decline as known historical human population collapses. However, it should be stressed that local behaviour will differ in a number of crucial ways from global behaviour. For example, if local resources become scarce, a community’s population may drop because people are leaving to go to areas of better prospects whereas this is not an option during a global resource squeeze. As such, the global shapes may be different from local shapes. However, history does provide evidence that actual population collapses occur in much the same way as the simulated population collapses after the availability of non-renewable resources hits a critical point.

      [24]

An Analysis: Why is Starvation So High When There is So Much Food?

Cause of Death and Per Capita Excess Needs are two of the charts on the next page. The green line in the Cause of Death chart is starvation while Per Capita Excess Needs tracks how much ‘food’ has been harvested for each agent in the system. When this trend appeared (rising starvation in the presence of a rapidly increasing food supply), the next thing that happened was a frantic search for faulty programming. When nothing was found, the next step was to determine the actual actual starvation statistics and compare them with the simulation. It became evident that starvation in the real world was also very high: 58% of all deaths are caused either directly from hunger or indirectly through becoming vulnerable due to significant nutritional deficiencies[25].

Especially since ‘LiveAid’, billions have been spent to stop starvation. This leaves us with two possibilities. On the one hand, if the considerable and sustained world hunger efforts have been effective then the previous percentages must have been much larger. On the other hand, global policy and charity efforts may have made no impact at all at the global level.

According to Sahlins[26] aboriginals in the desert of Australia worked only 3 to 5 hours a day. In other words, they had a considerable food surplus which they converted into more leisure than we enjoy today. While his work was controversial, no one suggested that he was so wrong that he was missing 3 in 5 people dying from starvation. Additionally, after spending 5 years living with First Nations people in the Yukon, I found no evidence of such high starvation rates previous to encountering Europeans. The evidence suggests that communities around the world did not routinely experience such high starvation casualties before globalisation.

If this is true, the not only are the ‘feed the world’ efforts failing, the starvation rates are much higher now than they were before modernity.

But why would starvation rates go up in the presence of so much food? Evidence from the simulator offers intriguing possibilities. The simulator gives us insight into aspects of society that are normally impossible to explore.

Going back to the Cause of Death chart, the three main causes of death are infant mortality, disease and starvation. Near the end of the run the pink line (infant mortality) and the brown line (disease) become depressed as a result of a rise in technology. However, the green line (starvation) rises significantly even though there was more food available than at any other time. This is apparent in the Per Capita Excess Needs chart (‘needs’ is possible food in the simulator). At first, this seems to violate common sense, especially as the simulator contains none of the societal detail that is normally blamed on societal inequities such as privileged education, politics, law or capitalistic greed. However, by examining several charts at once, something entirely new becomes apparent. 

On the Birth and Death chart, the top black line is the births (as a percentage of the population) and the bottom one is the deaths. Notice how they tend to mirror one another even during rapid population changes. Only small deviations within a narrow corridor are required to accomplish huge population swings[27].

In thousands of simulation runs, the Birth and Death charts never significantly deviate from mirroring one another. If either birth or death spikes, it rapidly returns to its normal corridor.  Only when they both change together can long term changes become stable.

This hypothesis was tested in the simulator by manually increasing other deaths. When this was done, the starvation rate dropped. Another way to lower starvation in the simulator was to manually reduce fecundity. As long is it remained above the subsistence line, the link between starvation and the system’s food supply seemed weak.

If this simulation accurately reflects the reality of global society, while the real global birth rate remains at its current level and medical technology continues to suppress death due to other causes, starvation must release the pressure[28].

From a global point of view, local successes can be misleading. It is analogous to trying to get bubbles out of something that has been glued to a flat surface. Locally, one can succeed by applying enough pressure directly to the nearest bubble. However, when viewed from a global perspective that action accomplishes nothing because the pressure simply moves somewhere else and another bubble rises. This is how we can exert an enormous effort and enjoy several significant newsworthy successes while achieving nothing from a global point of view.

More ominously, when the non-renewable society emerges, the simulation seems to be utilising surplus labour for the purpose of optimising production rather than agent welfare. If large parts of the population are not properly fed, surplus labour immediately increases which then further frees up the productive agents to produce more. Production would be negatively affected in a number of ways if surplus labour were to be shifted to feeding people who would otherwise starve to death.

Contrasting Theories: Free Market Economics and the Wealth Creation Imperative

Accepted economic theory views the free-market as a tool for wealth creation that society wields for its own benefit. On the other hand, a society shaped by a wealth creation imperative is itself wielded by an evolutionary system optimising for wealth creation over all other things.

In the accepted economic view, society has the power to mitigate undesirable free-market consequences through policy and regulation. For example, pollution resulting from wealth creation can be countered with law and regulation. This way we get all the benefits of the free-market while maintaining control of the disadvantages. This picture suggests that any unfortunate events, such as the Deepwater Horizon oil spill[29], simply reflect a failure of regulation and law to properly control society’s wealth creating tool.

However, there is a great deal of evidence to suggest that this is not the reality. For example, consider the mainstream debate on regulating bank remuneration. In the UK, both sides of the debate and the reporting media accepted the evidence that society was being adversely affected by wealth disparity, but then proceded with the assumption that wealth creation was far more important. As such, the only point of debate left was whether banks could be regulated without impacting wealth creation[30].

If the public sector was actually free to mitigate disadvantages of the free-market, this would not be a debate at all: social well-being would trump wealth creation in a society where people are important and we have so much wealth that obesity as a problem. The reality is circumstances force societies to permit wealth creation activities even though it creates known social degradation. This suggests that wealth creation wields more power than social concerns in society. If this is the case, then how can society mitigate the bad effects of wealth creation?

This is just one example of how circumstances force any societal entity to progressively become more monetarily efficient regardless of the social impact. Wealth dominates everything in every sector and every walk of life. The only way to temporarily escape it is to become independently wealthy.

In a society being shaped by a wealth creation imperative, mitigating forces that impair wealth creation will be progressively disempowered. If laws and regulation have to be ‘competitive’ against a constantly rising bar then, over time, they will increasingly come to serve wealth creation. Those living in the wealthy countries will become wealth creating entities of the highest order, meaning that any expression not involved in wealth creation will become increasingly diminished. Those not creating wealth will be left alone. However, as the wealth creating forces come to control more of the earth’s resources, those left alone will increasingly die like every other form of life that does not create wealth[31].

Standard economics suggests we control the free-market and society through regulation, law and policy. The theory of societal evolution says we have never had control in the way we thought we did. Standard economics says big global problems can be fixed through laws, regulations and policy. The theory of societal evolution says all this activity is a waste of time and energy unless it ends up changing the global societal wealth creation imperative. As long as the wealth creation imperative remains in place, the future nature of global society is assured

Part II

Detailed Technical Description of the Simulation

This section is for those wishing more information about the actual simulator in order to determine the validity of the model and for those interested in its detailed workings and programming. While effort was taken to write this in simple terms, it does not to attempt to explain how multi-agent simulations work or how to programme a simulation. Some knowledge of simulations and/or programming will be helpful.

The Basic Assumptions of the Societal Simulator:

The Patches (Environment)

• The environment consists of renewable and non-renewable resources.
• Renewable resources are renewed at each iteration. The non-renewable resources are not renewed.
• Each patch contains different levels of resources.
• Each patch requires a different effort in order to access the resources.
• Some patches cannot produce ‘food’.

The Agents (people)

• All the agents need a same basic amount of renewable resources every turn in order to stay alive.
• Agents start with the same effort with which they can harvest renewables and non-renewables.
• Agents can trade excess food and goods with other agents.
• Money is the medium of trade.
• Agents die of various causes and, if nothing else gets them, old age.
• Agents breed. The children may or may not be like their parents.
• Agents can move.
• Agents possess ‘evolutionary variables’ that allow agents to utilise their resources and behave in different ways.

Miscellaneous

• Increasing non-renewable assets in society leads to energy, travel and medical technology.
• Renewables can be converted into non-renewable technology.
• Non-renewables wear.
• Harvesting non-renewable resources too rapidly can lead to the degradation of renewable resources. Recovery takes place at varying rates.
• A child will inherit the parent’s wealth. ‘Food’ is not inherited.

The key to getting realistic output from the simulator is in correctly configuring the environment in terms of available resources and the required harvesting in relation to the ability of the agents’ to harvest and the need of the agents’ for food.

Programme Overview

This simulation was built inside the academically accredited NetLogo 4.0.4 and 4.1.2 multi-agent simulation environment. At the core of this particular simulation are three simple environmental elements and one evolutionary element:

needs – a single environmental variable representing the renewable forces created in human society, especially food. This gives us a measurement of surplus labour.

wants – a single environmental variable representing the non-renewable forces in global society.

money – a fixed resource of no intrinsic value which can be traded for either ‘wants’ or ‘needs’. In this version, the money supply does not change.

agent – an evolutionary element meant to capture a few key forces generated by human beings in the environment of the world today. In contrast to models such as Uri Wilenski’s Wealth Distribution[32], all the agents in this simulation have identical potential (their ‘labour’ and ‘needs’ start out evenly across all agents).

The agents in the simulation can exist in four different states which, in turn, create four different possible agent-generated simulation environments. Somewhat arbitrarily, these four states are called Hunter/Gatherer, Trading, Innovating and Industrial.

The key characteristics of these four states in their simplest forms are:

Hunter Gatherer – (Consisting of Agents and Needs)

Agents look for and harvest food, reproduce (asexually), and die from either lack of food or old age. The patches contain the single renewable food source ‘needs’ that has a fixed random resistance to harvesting. Using their ‘labour’ (user defined using the ‘starting-labour’ slider and the same for all agents), the agents overcome the resistance and harvest the renewable resources. Once the resistance is overcome (whether by a single agent or a group of agents), there is no further cost to harvesting the food, but since this usually happens only once per iteration this resistance can be seen as a fixed cost, a variable cost or a combination of both[33]. If more than one agent is present on the patch, the ‘needs’ is divided up between them in accordance to the labour still available after they overcome the resistance (‘harshness’).

On each patch, ‘needs’ are available in a set amount that is randomly assigned to each patch at the beginning of a simulation run. ‘Renewables’ are renewed once per iteration. The patch resources and resistance are randomly set up (according to user defined parameters) at the beginning of each simulation run.

Trading – (Consisting of Agents, Needs and Money)

Agents do all of the above and, in addition, they can now buy and sell ‘needs’ using money. At the beginning of the later versions of the simulation, all the agents start with the same amount of money and the money supply is fixed thereafter. Each agent initiates buying (if they have money) and selling (if they have ‘food’ in excess of their chosen savings level) once per iteration. They seek to purchase below their chosen price and sell above their chosen price. If no trades can be found within their immediate vicinity (within 1 patch beyond them) they will not initiate a trade. This state change from Hunter/Gatherer to Trading is done to all agents simultaneously.

Innovating – (Consisting of Agents, Needs, Wants and Money)

Agents do all of the above but they now have the choice of converting their ‘needs’ into ‘wants’ simulating the force of turning renewable resources into fixed assets that deteriorate without human intervention. This only occurs in the ‘innovating’ state. These ‘wants’ (technology) increase the agent’s ability to overcome patch resistance and harvest more food, as well as medicate themselves and increase their movement. They can buy and sell ‘wants’.

The user can cause this state change to initially occur to as few as one agent up to the entire population. If the user chooses a small number, the new state will propagate if a ‘Trading’ agent purchases a ‘want’. ‘Hunter/Gatherers’ will not convert because they do not trade. Whenever they disappear it will be through extinction.

Industrial – (Consisting of Agents, Needs, Wants and Money)

Agents can now harvest non-renewable ‘wants’ (non-renewable technology) directly from the patches. Agents have identical rules to that of the previous level except they no longer convert ‘needs’ into ‘wants’. The rules of harvesting ‘wants’ are the same as harvesting ‘needs’ except that ‘wants’ is not renewed. The amount of ‘wants’ is randomly assigned to each patch within a user constraint upon setup.

The user can cause this state change to ‘Industrial’ to occur to as few as one agent up to the entire population. If the user chooses to convert a fraction of the population, the new state will propagate whenever a non-‘innovating’ agent purchases a ‘want’ from ‘industrialists’ or if an agent harvests a patch that has been altered by the harvesting of ‘wants’. The latter simulates industrialising powers such as land ownership and taxation or the enclosure laws in England that forced the mobilisation of an industrial labour force.

User Control of the Four States

The cluster of buttons and sliders to the right of the simulation window control these four states. On setup, the agents default to the Hunter/Gatherer state. If the user sets the Auto-Advance switch to ‘on’, the simulation will automatically advance individual agents through the different stages by occasionally randomly assigning an agent a new level, randomly generated from one level above the highest level of any agent already in the simulation. The slider Auto-Turnon lets the user run the simulation at a certain level before the Auto-Advance is automatically switched on. Auto-Shutoff allows the user to set how many Industrialists there must be in the system before the Auto-advance function automatically shuts off. Strength controls how strongly the Auto-Advance affects its environment.

If the user wishes to watch the process in the simulation, they can turn on the Colour switch which will cause the agent’s colour to reflect their level: red for ‘Hunter/Gatherer’, yellow for ‘Traders’, green for ‘Innovators’, and blue for ‘Industrialists’.

The user can also manually control this process. The appropriate Initial-Wave sliders determines the number of agents that will be converted: Initial-Wave1 determines the number of ‘Traders’ to be converted, Initial-Wave2 determines the number of ‘Innovators’ and Initial-Wave3 determines the number of ‘Industrialists’. Once set, the user can press the appropriate button to manually convert agents at will (Manually Trade for ‘Traders’, Manually Innovate for ‘Innovators’ and Manually Industrialise for ‘Industrialists’. The number converted will occur each time the button is pressed and the buttons work whether the simulation is running or has been paused. The user can also use sliders to have the simulation automatically convert agents at a particular time (tick) in the simulation by setting the appropriate slider: Start-Trading to introduce ‘Traders’, Start-Renewable-Economy to introduce ‘Innovators’ and Start-Industrialisation to introduce ‘Industrialists’.

The above buttons and sliders are grouped according to the agent state they control.

The user can start the simulation at any stage. Equally, the user can jump stages during simulation (i.e. introduce industrialists into a hunter/gatherer economy). If the sliders are set at the same ‘tick’ the simulation will convert the agents to the most ‘advanced’ state. Using the manual buttons, the user can also reverse the process (i.e. introduce traders into an industrialist society).

The simulation will occasionally[34] randomly convert agents to a new state in a range defined by the minimum (Hunter/Gatherer) to the maximum present in the simulation environment.

Simulation Controls

User Control Over Agents

The user has the ability to introduce and experiment with the effect of several sliders either for an entire simulation run or as an interruption during a simulation run. These sliders are (in alphabetical order):

Accidents – This is death by causes unmitigated by choices or wealth.

Charity-Enabled – This is the switch that turns on charity and thus activates the Giving-Threshold control.

Child-Benefit – This taxes all the adult agents in order to give a benefit to all the dependent children.

Child-Dependency – This determines the age at which an agent retains child attributes.

Child-Needs – This determines what proportion of Living-Cost a dependent child requires to live.

Giving-Threshold – This determines at what point an agent will start ‘begging’ for charity.

Disease  – This simulates individual disease, dangerous activities and other causes of death that stem from individual or collective choices.

Epidemics – This simulates deaths en-mass.

Fertility-Max – This determines the initial fertility level of the agents at the beginning of a simulation run. [fertile = random fertility-max]

InfantMort – the likelihood of an agent dying before the age of 5.

Life-Expectancy-Max – the maximum age attainable without the help of ‘technology’ (or ‘wants’).

Living-Cost – how much ‘needs’ is subtracted from the agent during each iteration. This is the cost in terms of renewable resources an agent requires at each iteration in order to remain alive. If ‘needs’ ever goes to zero or below, the agent dies.

Medical-Strength – This increases/decreases the effectiveness of the agents evolved use of ‘wants’ in combating death by ‘Untimely-Deaths’, ‘InfantMort’ and ‘Epidemics’.

Travel-Cost – ‘Wants’ (technology) increases an agent’s ability to travel (if they choose to use it this way). This is the cost of that travel in terms of the proportion of ‘wants’ (technology) that was used for that travel.

User Economic Control

Recycling – This determines how quickly the societal stock of non-renewable resources is worn away by use. The higher the number, the lower the wear (because of more recycling).

Renewable-Economy – This determines how efficient the agents are at converting their renewable resources into technology. This is only effective with agents working as ‘Innovators’ (the third state).

User Population Control

Birth-Control – This simulates the population choosing to have fewer children than the agents naturally would in that environment.

Child-Limit – This limits the number of children an agent can have. It simulates the effects of the ‘Chinese Policy’.

Fecundity-Policy – If we were able to, this would prevent fecundity from increasing beyond a certain level. It is a fantasy population control policy programmed just to see what would happen.

User Control Over the Environment

The most powerful influence the user has on the environment is during setup where the user can set the various amounts of resources available and their resistance to harvesting. However, once the simulation is running the following sliders have no influence. The sliders involved are:

Eco-Recovery – This determines how quickly patches recover from the environmental disruption caused by the harvesting of ‘wants’. (This needs further fine tuning for control. It’s a bit sluggish at the moment.)

Ex-Harshness-Amount – This determines by how much the extraordinarily harsh patches are resistant to harvesting.

Extraordinary-Harshness – This determines what part of the environment is inaccessible to the agents in terms of resources by increasing the resistance to extraordinarily high levels. It creates environmental boundaries and ‘deserts’. Sometimes, very powerful agents can harvest those areas that have relatively lower resistance.

Fixed-Harshness – On setup, the programme uses this number from which to randomly generate the resistance covering the fixed ‘wants’ on each patch.

Fixed-Soft-Patch – This is the number of patches that will have its resistance covering ‘wants’ forced to become lower than a single agent’s . It is used to allow a technological society to seed itself in a very harsh environment. Without it, sometimes it can be difficult to get a technological society to ‘ignite’.

Num-people – this determines the population of randomly generated agents that begins the simulation. Keep in mind that many of these initial agents may have an evolutionary setup that will not allow them to survive.

Max-Fixed-Resources – On setup, the programme uses this number to randomly generate the number of ‘wants’ available on each patch.

Max-Renewable-Resources – On setup, the programme uses this number to randomly generate the number of ‘needs’ available on each patch.

Renewable-Harshness – On setup, the programme uses this number from which to randomly generate the resistance covering the ‘needs’ on each patch.

Renewable-Soft-Patch – This is the number of patches that will have the resistance (harshness) covering ‘needs’ forced to become lower than a single agent’s efforts. It is used to allow the agent society to seed itself in a very harsh environment. Using it, the user can create very harsh environments indeed.

Starting-Labour – This slider determines how much force an agent can exert in its environment (before technology). The number set on Starting-Labour directly equates to how the maximum amount of ‘needs’ an agent can potentially harvest before ‘technology’ is introduced. The balance between this slider and Living-Cost is important in tuning the simulation as it will determine how much excess an agent can harvest (i.e. if Living-Cost is set to 5 and Starting-Labour 4 then all the agents will starve to death. On the other hand, if Living-Cost is set to 1 and Starting-Labour 4 the agents will easily create surpluses … up to 4 cycles of food requirement in one just one cycle’s harvest.

There are some environmental sliders that can be used effectively during the simulation. These exceptions are:

Contamination – This determines the extent to which ‘pollution’ leaks to the neighbouring patches.

Environmental-Degradation – The number set by this slider is the percentage reduction of the maximum resources available. For example, setting the slider to .1 will decrease the maximum possible on a patch by 10% and increase the maximum harshness possible on a patch by 10%. This allows the user to simulate various degrees of environmental disruption.

Environmental Disruption – A button that regenerates the renewable resource environmental setup using the original figures set even if the user moves the environment setup slider after the simulation has started. This simulates an environmental shift.

Miscellaneous Sliders

Auto-Naturalisation

The number on the slider relates to the amount of ‘wants’ in the possession of an agent. If that level of ‘wants’ falls below a certain level, depending on the agent’s propensity to mutate, that agent may become an ‘innovator’.

Middle-Wealth-Cutoff-% and Bottom-Wealth-Cutoff-%

These two sliders define the agent’s wealth class used by the graphs. The programme takes the richest agent and then defines the rest of the agents by what percentage of that wealth the other agents possess. These sliders influence the ‘X (Lower Wealth) = Higher Wealth’ graph and the ‘Children by Wealth’ graphs. The results are reported by the ‘High Wealth’, ‘Middle Wealth’ and ‘Low Wealth’ monitors. These sliders do not influence the behaviour of the simulation in any way.

Needs-Balancing, Wants-Balancing and Money-Balancing

These sliders balance out the value of each item. When the simulation is allowed to determine value by price, then these sliders become superfluous but otherwise they allow the user to determine the relative value of each of the three tradable items for the purposes of wealth calculations. These sliders influence the graphs only and have no impact on the simulation, as is the case with the V128 of the simulation (the latest at the time of writing).

Stop-Simulation – This slider allows the user have the simulation stop at a particular ‘tick’. It is useful when quickly comparing the shape of the graphs from different runs because the graphs will have an identical compression. Also, this allows the user to ‘set and forget’ the simulation and come back later on to see what happened.

Money-Supply – Determines how much money is given to each agent when the simulation is started. (earlier versions have variable money supply functions but it didn’t influence the shape of the ‘society’ in any way that I could determine).

Miscellaneous Buttons

Go – Starts and stops the simulation.

Reset Pop Peak– Resets the peak population monitor to the current population.

Setup – Sets up the simulation environment and places the agents in the environment.

Evolutionary Variables

Over time, the individual agents have the ability to make numerous choices and adaptations about the way they behave in their environment. The way they do so is through these evolved variables.

charity – this determines how much an agent gives to ‘beggars’.

fertile – this determines an agent’s likelihood of having children.

industrial-efficiency – determines how intensely an agent will use the ‘wants’ they have at their disposal. The trade off is that the more ‘wants’ are used, the more they wear.

movement – if an agent’s income begins to drop, this determines their likelihood of moving off that patch before it becomes a life or death decision.

mutation – This determines the likelihood an agent’s child will mutate away from the parent’s characteristics on birth. If ‘mutation’ goes to zero, the children will be clones of their parents.

nurture-needs – this determines the percentage of ‘needs’ a parent will give its child on birth.

nurture-wants – this determines the percentage of ‘wants’ a parent will give its child on birth.

nurture-money – this determines the percentage of ‘money’ a parent will give its child on birth.

order-of-trade – an agent can harvest ‘needs’ or ‘wants’, buy ‘needs’ or ‘wants’ and sell ‘needs’ or ‘wants’. This variable determines the order that this is done. In industrial mode, it allows the agent to choose from 24 different combinations, 12 of which emphasise obtaining ‘wants’ over ‘needs’ while the other 12 emphasises ‘needs’ over ‘wants’ (these 12 I call ‘farmers’). In the ‘innovate’ mode, it allows the agents to choose the order at which they buy and sell their wears. Graphing this variable is fascinating as different ‘cultures’ can rapidly rise and fall but it has become too complicated to do so anymore. In the future I intend to add more ‘cultures’.

price-up – whenever the agent changes its price up, this is the amount it raises the price. 

price-down – whenever the agent changes its price down, this is the amount it lowers the price.

propensity-health – this determines what percentage of ‘wants’ the agent uses to fight disease. 

propensity-labour – this determines what percentage of ‘wants’ is used to increase its labour.

propensity-leisure – this determines what percentage of ‘wants’ does nothing. 

propensity-use-needs – this determines what percentage of excess ‘needs’ an ‘innovator’ will use to convert into ‘wants’.

savings-rate-needs – This is the amount of ‘needs’ an agent will hoard before selling. All sales occur on amounts surpassing this amount. It is in terms of ‘living-cost’ and so savings will change if the ‘living-cost’ slider is changed.

savings-wants – the amount of ‘wants’ hoarded before selling. All sales occur on amounts surpassing this amount. If this is ‘2’, then the agent will always hold onto at least ‘2’ units of ‘wants’ before they consider selling. 

vision – this determines the size of the radius in patches that the agent is able to examine before deciding what direction to go. If it goes below 1 (the red line on the graph) the agent is effectively blind to its environment and the direction of its movement will become completely random.

wander – this determines an agents propensity to jump several patches in order to go exploring. 

Programme Graphing and Monitors

Standard Graphs and Monitors 

The first window is the simulation itself. A green environment indicates the renewables are at their most accessible. The agents start with different colours that their children inherit. If the environment starts to turn grey (or orange) that means the harvesting of non-renewables is impacting the accessibility of renewables. The darker the grey or orange, the larger the impact. It is possible for the agents to harvest non-renewables with little impact on the environment and this sometimes happens once the technological society has declined.

Average Age – ages of the agents has been summed and divided by the population. 

Birth and Death – this graphs the total births in that iteration divided by the population and the negative of total deaths divided by the population. The middle orange line is 0 and there are two red reference lines currently set at .9 for birth and -.9 for death.

Farmers – this measures the population of agents who choose harvesting ‘needs’ over ‘wants’. As industrialists, these agents will only harvest ‘wants’ if they can’t harvest any ‘needs’ during a cycle (tick). See also order-of-trade under Evolutionary Variables. 

Fixed-Resources – sums the patches’ ‘wants’. It is useful to see whether the agents are still harvesting ‘wants’ from the patches when the amounts are too small to register on the graph.

Hunter/Gatherer, Traders, Innovators and Industrialists – this measures the population in each category. 

Non-Renewable Resources – graphs the sum of the patch’s ‘wants’.

Peak – reports the highest population recorded since the Reset Peak Population button was pressed. 

Per Capita Productivity – graphs the amount of harvesting of ‘needs’ and ‘wants’ in that cycle (tick) divided by the population. ‘Wants’ has been graphically repressed in order to prevent the graph from shooting meaninglessly straight up every time the industrialists take over. This graph is softened by summing the reported number for the last 6 ticks and dividing them by 6.

Per Capita Surplus Labour – this graphs the sum of all the agent’s ‘needs’ in excess of the sum of all the ‘needs’ required to fulfil the ‘living-cost’. If ‘living-cost’ is 3 and there are ten agents with 24 ‘needs’, the total living cost is 10 x 3 (living-cost x agents) = 30 so this graph would report back 24 – 30 / population = -.6. This means most agents would be living hand to mouth meaning if they didn’t harvest ‘needs’ in that iteration some would certainly die of starvation. This would mean creating or harvesting non-renewables would be a low priority for most agents. Reporting back ‘0’ would mean the society as a whole had 1 iteration of excess food. 

Population (graph) – graphs the simulation’s population over time.

Population (monitor) – reports the number of agents in the simulation at any one time. 

Prices – adds up all the prices in that cycle and divides them by the number of transactions. The purple line is the price for ‘wants’ and the green line is the price for ‘needs’.

Societal Assets – the sum of all non-renewable assets held by agents. 

Total Wealth Per Capita – sums the agents’ ‘needs’ and ‘wants’ and divides by population. As in the per Capita Productivity graph, the ‘wants’ has been suppressed.

Wealth Top, Wealth Mid and Wealth Low – this is the population in each wealth band. 

X (Lower Wealth) = Higher Wealth – using the simulation’s definition of wealth, X is the variable graphed, low wealth is the sum of wealth of the lowest wealth class divided by the population of that class and high wealth is the sum of wealth of the highest wealth class divided by the population of that class.

# of Trades / Capita – sums the trades and divides by the population. The purple line is the per capita trading of ‘wants’ and the green line is the per capita trading of ‘needs’.

The Evolutionary Graphs 

For almost all these graphs, the black line is the system average (the sum of the agent’s variable divided by the agent population) while the red line reports the highest value of this variable held by any agent in the system while the green line reports the lowest value.

Further definitions of the variables used can be found under ‘Evolutionary Variables’. 

Charity – reports the variable ‘charity’.

Children by Wealth – reports two variables: the sum of children had by people from the lower class divided by the sum of people in child bearing age in the class and the sum of children had by middle and upper classes divided by the sum of people of child bearing age of these two classes. The red line is the lower class while the turquoise line is the upper and middle classes. 

Fecundity – reports sum of ‘children’ had by the living agents divided by the sum of all agents above the age of 11 (the age they need to be before they can have children).

Fertility – reports the average only of the average ‘fertility’ of the lowest class by wealth and the combined average of the middle and upper class by wealth. The turquoise line is the upper and middle classes while the red line is the lower class. 

Health-Choice – reports the variable ‘propensity-health’

Industrial-Usage – reports the variable ‘industrial-efficiency’ 

Labour-Choice – reports the variable ‘propensity-labour’

Leisure-Choice – reports the variable ‘propensity- leisure’ 

Modern Travel – reports the variable ‘migration’. This is how intensively agents use technology to help them move.

Movement Desire – reports the variable ‘movement’. 

Mutation – reports the variable ‘mutation’.

Nature Vision -- reports the variable ‘vision’. 

Nurturing-money – reports the variable ‘nurturing-money’

Nurturing-Needs – reports the variable ‘nurturing-needs’

Nurturing-Tech – reports the variable ‘nurturing-wants’

Savings Rates – reports two variables: the sum of savings rates for ‘needs’ divided by the population and the sum of savings rates for ‘wants’ divided by the population. The purple line is ‘wants’ and the green line is ‘needs’.

Some Details on How the Simulation Works

How Trading Works

If an agent is at least a ‘trader’ and has ‘needs’ or ‘wants’ above their savings rate, or they have any money, the agent will attempt to initiate trades with any agents that are within a small range of patches around them (begins at 1). If selling, it will only consider buyers whose offer price is equal to or above the asking price. If that is the case, the agent will seek the lowest price (regardless of quantity) and then initiate an exchange. The agents trade as much as they can in any transaction but always go for the best price. An agent will choose a small trade with the best price over a larger trade with a slightly worse price.

If an agent still has money or goods left to trade after the first trade, it will attempt to trade again. If the agent fails to trade on the first attempt, it shifts its price to enhance the chances of achieving a trade and tries again. For example, if selling goods, it will lower its asking price and try once more. This is how prices move.

All agents start the simulation with bid and offer prices of ‘50’.

The agents’ key trading variables are:

buy-price-needs – At each iteration, an agent will compare its price to that of the price calculated[35] by the patch on which it currently resides. If the agent’s price is higher than the patch, it will lower its price by the amount ‘price-down’ (see price-down and price-up under the Evolutionary Variables section). It the price is too low, it will raise it by the amount ‘price-up’. This is an agent’s price offered to purchase ‘needs’.

buy-price-wants – This is an agent’s price offered to purchase ‘wants’.

sell-price-needs – This is the price asked for ‘needs’ sold.

sell-price-wants – This is the price asked for ‘wants’ sold.

How Reproduction Works

If an agent is within breeding age (currently arbitrarily set at 12-40 ticks or cycles old) a random number (random 1.5 x age) is compared to its evolved fertility (an agent variable called ‘fertile’) and if ‘fertile’ is the higher number then a child will be born. An agent can have a maximum of one child per year of fertility. The user can implement population control policies as outlined in the User Population Control section.

How the Agents Move

Primarily, agents will move .9 (in order to prevent military-like lines forming) of a patch length if they cannot harvest what they seek or their harvest of ‘needs’ is below ‘living-cost’. In seeking either ‘needs’ or ‘wants’, they can access a ‘vision’ code that gives them an idea[36] of the location of a likely fertile patch.

If they desire to do so, they can also do a blind jump of several patches and then explore the new area using the method above.

At present, there is no code to help them locate places where trading conditions are better.

Some Last Notes

For most environment configurations, the resources in many of the patches will be inaccessible by single agents at the beginning of the simulation. While there must be ‘seed’ patches where agents can harvest even if they are by themselves, in order for society to grow they need to combine their efforts to access the full richness of their environment. The user can watch this happen especially if the state of the simulation remains at the Hunter/Gatherer or Trader level. The agents themselves have no code to help them cooperate. Cooperation can only occur through a self-organisational process at the societal level.

The industrial phase lowers the need for direct cooperation to access resources as technology makes the agents much more individually powerful in this respect. The most powerful agents can stride their world harvesting anything they want at will. However, if the agents ‘choose’ to disregard the environment it can have rather severe consequences once the non-renewables hit a critical level of scarcity. It is rare that the agents remove all the ‘wants’ from the patches before the population collapse occurs. This can result in interesting situations in the ‘post apocalyptic’ era when a society sometimes rises again with habits and processes much more conservative in its use of non-renewables.

This is only a brief overview. One could quite literally write a book simply from the results of this simulation.

Part III

Conclusion

Repeated success in building artificial evolutionary systems means scientists and engineers have a good practical understanding of designing such a system to solve specific problems. Human society contains the necessary systemic prerequisites in order to be an artificial problem-solving evolutionary system. In the absence of convincing evidence that any system with evolutionary pre-requisites can avoid entering an evolutionary cycle, we are left with a strong possibility that human society has been acting as an artificial problem-solving evolutionary system for some time now. The key forces in modern society aim the societal evolutionary forces to solve for wealth creation.

Our understanding of evolutionary systems tells us this would lead to a system optimising for wealth creation over all other things. This explains our current situation in the world and much of what goes on in society today. However, our practical knowledge of artificial evolutionary systems also tells us how to alter society’s evolutionary imperative in order to make a real impact on some of the most serious problems facing human society today and to create a society much more sympathetic to a broader set of human needs, desires, and dreams.

The solutions avoid many of the current pitfalls of international negotiation and policy creation, in particular, the need to agree on detail. Computer simulation can give us a high degree of certainty that our assumptions and interventions will accomplish what we wish.

Problems would include resistance from people who feel that human intent must be the most potent force on earth (or the universe), the effect on human awareness that the current system exerts on humanity, and the fact that humanity does not have a great deal of time before change is thrust upon it (for example, the ocean is due to run out of fish in 2050[37]).

In order to visualise the possibilities, consider how often you, your family, or your community (or any organisation) felt forced to do something you did not quite want to do because of circumstances that are, directly or indirectly, tied to money. When this happens, do you tend to accept circumstances as a fact of life or did you find yourself wishing for different circumstances? Our understanding of evolutionary systems informs us the former is not true and exactly how to achieve the latter.

Now imagine a world where circumstance empowers behaviour that supports you, your family and all life on earth while disempowering destructive behaviour[38].

Further Research: What’s Next?

Free Will Within an Evolutionary System

One of the most intriguing areas that have come up in this research is how an evolutionary system might optimise any free-willed agents it finds within its environment. Possible answers to this may lie in the techniques used by a group of magicians called mentalists.

Good mentalists create a show around their ability to get people, of their own free will, to choose what the mentalists want them to choose. Once this skill is honed, it is a simple matter to appear both psychic and prescient by setting up elaborate and unalterable indications predicting what people are going to choose in the show.

For example, one mentalists hired a promotional company to come up with an advertising campaign for an imaginary company. He handed them an envelope which he asked them not to open and then left. On returning, he asked them to pitch their campaigns (there were two). Then had them open the envelope. Inside were all the elements (the logo including a stylised drawing, the catch phrase, the actual pitch) of the two advertising campaigns. Then the mentalist showed us exactly how he did it.

What he did was pick up the promoters and then controlled their environment. For example, the stylised drawing was on the shirts of a teacher and school children crossing the road in front of their car. Other elements were placed in shop windows, etc. To sum up this ‘trick’, the mentalists said he was pleased that the marketers themselves were vulnerable to the methods they routinely apply to consumers.

The point is our ‘free will’ is tremendously influenced by what we see in the environment and how often we see it. Even repeating a simple logo creates a sense of familiar comfort that is later triggered when the product is seen in a shop. Again, no thought is required. That is why companies pay millions to have their logos displayed at large events and hire companies to count how often their logos appear so they can compare actual buying patterns to the actual number of times their logos appear. While any results from this research is an industrial secret, they continue to pay millions to have their logos displayed.

If the environment is empowered when it creates wealth and disempowered where it does not, then it becomes a simple matter to create a self-reinforcing cycle that quickly turns society into an environment generating ‘free will’ behaviour that serves the evolutionary imperative. This reaches into all areas of our life. Consider, do people ever have to buy something new simply to attend a social event? Do we need ever more advanced electronic devices in order to communicate with people? Do we ever buy anything in order to enter courtship activities? Christmas?

It might be that stigmergy is far more relevant to societies built by human beings than it ever was to termites.

When programming an evolutionary system, there is no functional difference between programming ‘free-willed’ action and random action.

Preliminary Work in Deriving Evolutionary Imperatives From First Principles

For the purposes of this paper, evolutionary imperative is being defined as the problem being solved by the overall evolutionary system. In the case of simple artificial systems the answer is straight forward: the evolutionary imperative arises directly from the fitness test. In natural evolutionary systems, the test must be automatic and must arise from the conditions and processes in the system.

There is a chart at the end of this section that directly compares the two different kinds of evolutionary systems.

In simple artificial systems the fitness test dominates the process.  However, in natural evolutionary systems, the process creates the circumstances that determine the fitness. At the root of the process will be a key factor determining the problem that the evolutionary system is trying to solve.

For natural evolution, it is replication. As Hod Lipson demonstrated in his TED lecture[39], an evolutionary system with no goal defaults to self-replication. This makes sense because if something with a limited life span does not replicate, it will not exist in the future while anything that does, will. Therefore, anything that accidentally self-replicates will suddenly appear in the future. Only if the self-replication stops will it disappear in the system.

The moment something begins to self-replicate, it will start influencing its environment because it will take up resources that would otherwise end up in non self-replicating ‘things’. As the self-replication process accelerates, the new life will consume the resources necessary for life the expel that which is not. Over time, this will self-organise the environment such that all the items necessary for self-replication will be contained in self-replicators and anything not will become more concentrated in their environment.

This will inevitably ‘pollute’ the environment. As far as we scientifically know, in earth’s history, the first pollutant was oxygen and it started an extinction event. However, this is where the magic of evolutionary solutions made a master stroke. In order to save the self-replicators that relied on sunlight and expelled oxygen, a new kind of self-replicator appeared: one that preyed upon the original self-replicators. At first, this does not seem like a solution at all until one considers that the new kind of self-replicators absorbed oxygen and expelled the declining life-giving element of carbon dioxide. This change in circumstances vastly improved the conditions in which all self-replicators could grow and led to something known as the Cambrian Explosion of life.

So, how did this incredible level of cooperation emerge at all? The answer is simple: the goal is to self-replicate, not to be the ‘fittest’. If one species can help another self-replicate, then its presence is empowered. Think of the coral reefs.

A goal of self-replication does not usually create an environment of urgency in life. Anyone who actually watches wildlife over a long period of time will notice that there are great lengths of time when wild animals don’t really do anything urgently. Most live very relaxed lives. In economic terms, animals have considerable surplus labour that they use mostly for relaxing and, if they are that way inclined, socialising. They are not driven to be the best. In fact, it is possible that characterising nature as ‘survival of the fittest’ is not only incomplete, it is incorrect.

If you define self-replication as life, then you have an evolutionary system with an overall evolutionary imperative of life over all other things. If this is true, the result would be evolutionary solutions to all environments on earth. Species would be deeply entwined with one another, each supporting the other either directly or indirectly. The health of the natural environment as a whole would be positively influenced by all species. The waste of one species would become food for another in a perfect balance over time. Only resources that are be used renewably are useful in such a system because otherwise self-replicating ends the moment the non-renewable resource ends. Thus the system is completely ‘green’.

Turning to human society, we see something different. An agent is an economic unit that possesses wealth[40]. In order to remain present over time, instead of self-replicating, any economic unit must maintain a flow of incoming wealth that at least balances out the outgoing wealth. However, ‘return on capital’ ensures that in order to simply remain in the same position, incoming wealth must always exceed outgoing wealth. In other words, everyone always must have more coming in then going out just to maintain their position.

This creates an inherently parasitic system, because no system can sustain all its agents receiving more than they give unless the system constantly takes from somewhere else.

Better practice is determined by how much wealth results. Better practice propagates throughout the system by agent adoption rather than agent replication. Bad practices are either dropped by the agent or forced to disappear through bankruptcy (or some other form of financial death) or disempowerment.

Areas of society better organised to create wealth will then receive more wealth. This, in turn, empowers the wealth creating organisation. The recent economic experience of Zimbabwe demonstrates what happens if moves are made that ignore the wealth creation imperative.

When an evolutionary system constantly and consistently promotes agents based on a single factor, then that factor can be equated to the fitness test in artificial solutions. Since the fitness test directs the evolutionary system we can then know the evolutionary imperative.

For human society, that single factor is wealth. As such, we can conclude that wealth is the evolutionary imperative of human society.

Changing the Evolutionary Imperative

The most exciting aspect of this work is developing ‘realistic’ societal evolutionary goals and the functional changes necessary in society to start the new evolutionary goal in action. While consciously changing an evolutionary imperative for human society might sound like a fantasy to some, there is precedence.

Adam Smith addressed our most recent successful attempt in ‘The Wealth of Nations’ when he described an ‘invisible hand’ guiding peoples’ selfish desire to better all in society (in material terms). This is exactly how an evolutionary system works. The moves that England made in switching from a feudal system to a capitalistic system was a process of successfully institutionalising the wealth creation imperative.

The key is the fitness test. In simple artificial evolutionary systems, determining the evolutionary imperative is easy: the fitness test is applied equally over all agents and never changes. As such, the long term goal is linearly defined by that fitness test.

However, this is not true when the fitness test emerges in a complex system. This kind of fitness test unusually involves survival and does not directly answer the question of where the system is headed.  Survival strategies change with circumstance and in dynamic systems the agents themselves can alter the circumstances in the environment. Many prematurely see the fitness test in nature as ‘survival of the fittest’ which is driving the system to evolve the ‘best’ animal. However, since circumstances keep changing, the criteria for the ‘best’ animal keeps changing. In such an environment, there is no ‘best’ animal over the long-term as advantages come and go as circumstances change. Therefore, evolving the ‘best’ animal’ cannot be the evolutionary imperative.

The key to understanding the evolutionary imperative in such a system is to look for the core of what the agents needs to do in order to succeed. In the natural evolutionary system of earth, that which most efficiently utilises the available resources in order to express itself (and therefore expressing life) will be the most empowered. In other words, that which most powerfully organises resources for the purposes of life best succeeds. Because different circumstance lead to different optimisations, the evolutionary system also organises all its resources to ensure life can best express itself in many circumstances. That is why there are so many species, many expressing cooperation even in between species, and why life appears in every place we have so far looked on earth. Summing this up, the evolutionary imperative of nature is the creation of life.

The question becomes what imperative might be best for human society and what technical steps do we need to accomplish in order to activate it? Can we do this without disruption to society? Can we do this before society as it is today hits a critical point and begins its likely collapse?

Tim Gooding

tim@humansos.org

Evolutionary System Comparative Chart 

Chart comparing artificial evolutionary systems designed by Danny Hillis[41] and Hod Lipson[42] with the natural systems arising in nature and human society.