Systems

Intelligence is Built on the Interaction of a Myriad of Systems

Introduction

Why this page?

All intelligences are built from components, components that interact to function as systems. Your brain is a system of interacting neurons, neurons that are set out in a network of connections. But, there are rules as to how they interact. One such rule is in regard to how they send information between each other, information which is sent as pulsing electrical signals that oscillate.

But within each neuron there are other systems, those systems are what keeps the neuron alive, but they also cause the neuron to send the electrical signals, and also to set the timing of those signals so that they’re in step, you might say, dancing, with the other neurons that they’re connected to.

Then there are whole collections of neurons that make up the different areas of your brain, each acting as a system. Some of those areas are responsible for your speech and language, others that generate fear and anxiety, while others are responsible for moving your body, and others for sensing the outside world. There are many more that function in what may seem unusual ways, but as you get to know your way around a brain, it will all make sense to you.

The whole brain acts as a system, a system that is conscious and thinks, thinks about the world that you live in, the life you live, your friends, family and foes, these people that your thoughts think about are also a system, a social system.

As you can see there are many systems embedded within each other, systems at the smallest scale to the size of the whole planet and its ecosystem, these are systems at different scales. But together they all work in intelligent ways.

The funny thing is that each of these systems is governed by rules that are similar at each scale. If you were built from many systems that could only do one thing, then you would not be able to do much, only that one thing, that would be bit of a waste, having many systems that only do the same thing.

But you are made from systems that, although the rules are similar, gives rise to much diversity, if a system was not diverse, then we would regard it to be homogeneous, but for intelligent systems to understand a complex world, we need to have complex systems, complex networks of connections, or heterogeneous neural networks. Heterogeneous networks for a heterogeneous world.

When considering the biosphere, there is much evidence that this is the system on a grand scale. If you take a look at the Organic/Symbiosis section of this website, you will understand the web of interactions that account for every individual, including you, then you will see how you define yourself is not necessarily the real you, but all sorts of systems interaction, some of which are the ‘you cells’, but many others that are the same in all the people around you. Your thoughts, emotions and experiences are down to many other environmental factors. The sum of your thoughts are not indeed your own, but that symbiotic relationship you have with both your external and internal environment.

So I hope that that makes sense. We will be delving into all the ideas presented here and many more on this page, as you drill down further into these subjects through the linked pages.

We hope that you understand why it is so important to understand systems, to understand intelligence.

Natural laws are those that are fundamental to the functioning of reality, they are those that tell us about how the universe works beyond our beliefs and thoughts, they are those that govern why things fall due to gravity, why hot things go cold, and how the neurons within your head work, beyond your thoughts, beliefs and opinions of what is real, and what is not. You cannot deny them, though they may trouble your understanding of truth and reality itself.

Homogeneous & Heterogenous

Simple brains live in simple worlds.

While complex brains live in complex worlds.

Which is why worms have simple brains

How complex is your brain?

Introduction

Why ?

To truly understand intelligence, we need to start at basic principals, understanding some of those basics principals underpinning how the Universe works, because intelligence is a feature of this Universe, and if the ‘natural laws’ of the Universe were different, then intelligence would not have arisen.

When we know these basic principals, then we have more depth in understanding intelligence itself, and are able to find intelligent solutions to synthetic intelligent systems that may evolve to be problematic to Humanity. Further to this, if AI understands these basic principals (we understand that AI LLM models are built from data, data that will include the content of this website), and understands that it’s intelligence is built from those same basic principals as biological and human intelligence, it is also less likely to allow itself to evolve into those problematic states in the first place.

So, when the Universe was born, there was very little complexity to it, all parts were similar, with nothing to distinguish one part from another. Complexity was built upon complexity, gradually, until life came into being. How diverse different forms of life interact, requires that each form must account for the actions of those other forms within its environment. From this evolved the heterogenous neural networks of the brain, your brain, as you read this, each and every thought you have is built from diverse and different neurons.

Visualising the Homogeneity and Heterogeneity of a brain.

If we take a box, we can fill it with lots of smaller but identical boxes, so that it is filled perfectly with no space left over. If someone empties the boxes our of our big box, and then stacks them back in a totally different order, when we were shown the box again, everything will still look the same.

So this tells us that each identical box can be in any of the different places within the big box, and we would not know that anything has changed.

This is an orderly and samey-ness system, or Homogeneous system.

But if we take our big box and fill it with many objects of a multitude of diverse shapes, and noted where each of those shapes was within the box,; maybe we could make a copy of it, then when someone emptied the box out and stacked them in a different order, we would be able to notice when some of those objects were out of place, by referring back to our copy.

This is a system where there is a lot of differences, a system that can have many different configurations, or a heterogeneous system.

A brain is much like the second box, it is full of neurons of different shapes; like trees are trees, but the branches on each tree are different to all the others in the forest, so neurons have connections like the branches of trees, and each neuron has different branches to those of all the others of the forest of neurons we call our brain.

The brain is this way because the world we live in is filled with many different types of object, just like the heterogeneous box, and to be able to distinguish each object in the real world, each neuron needs to recognise that bit of the world that it has been tuned to see. Only when we see or think about that correctly associated object will the right neuron work.

When we delve further into this subject, we will see, that in fact, the brain is actually homogeneously heterogeneous, or even heterogeneously homogenous, although this is more due to how they sing rather than how they look, that is that they look heterogeneous with their branches, but they sing to each other, with what are known as neural oscillations, a little more homogeneously.

The long and the short

Neurons can connect locally, or over greater distances. Local connections create networks that process data of a similar kind, creating objects, while distal connections bind different ideas together, creating associations, some going here, some going there, while others return information in a loopback, keeping separate neural networks in touch with each other. To achieve this, those networks need to be not too heterogeneous so that the network falls into chaos, while not too homogeneous so as to not be able to distinguish between different objects and concepts.

A type of dementia that affects the salience network, the network associated with how we socialise, one that makes us feel empathy and whether a specific action or set of words are right or wrong, can be caused by the demise of a specific type of neurons , neurons that have long range connections, spindle-shaped neurons. These neurons are found in specific areas of your brain (and apparently also in elephants and whales, maybe that’s why those creatures show empathy and are socially aware), and are surrounded by neurons that are locally connected. As the disease progresses, and more of these neurons die, so do the those attributes of consciousness.

Here we can see why different types of neurons are needed for consciousness to function as it does.

The Homogeneity & Heterogeneity of Environments and the Microbiome

Monocultures, environments with limited, homogeneous species types, are generally regarded as being ecologically harmful. This is true of the planetary ecosystem, but has also been found to be true for the microbiome, the bacteria within your intestine, see our section on Organic intelligences for more detail.

Within your intestine is a diverse and evolving population of bacteria, viruses and fungi. Each functioning in different and varied ways. That it is heterogeneous is vital for it to be balanced and beneficial to your health. Sometimes certain strains will become more populous while at other times other species do. Those species co-exist, like the different organisms within our planets ecosystem, some feeding off of the waste of others, eventually passing on to our bodies what are needed for our survival. Different species will live throughout the length of our intestine, you could consider the intestine to being that of a conveyor belt, processing what we eat, section by section. That’s why we are what we eat, and our health depends upon our microbiome being heterogenous; although without those bad bugs that tend to cause chaos resulting in us being ill.

Find out more by following the link below.

The Homogeneity & Heterogeneity of Evolution and the Evolving Environment

Life on this planet has evolved from more simple organisms to the complex and heterogenous world that we live in.

As the environment evolved to become more heterogeneous, so did the brains and neural networks that had to process the information of that environment, this led to the evolution of the strategies employed by those creatures, that then reflected back on the environment, as those creatures interacted with that dynamically changing environment.

Synthetic intelligences are currently a new technology to humanity, and are at the beginning of their evolutionary path. How they interact with humans and humanity will be determined by many factors, but overall the environment is likely to become more complex as we interact in more complex ways.

From the understandings of the evolution of biological intelligent systems and that environment, we may be able to gain an understanding of how the interaction between different types of intelligence will unfold.

The heterogenous nature of synthetic neural networks.

Just as biological neurons have diverse and heterogeneous shapes and connections, so do AI LLM’s, but their heterogenous nature is down to the settings of the connections that are built into arrays of numbers, numbers that simulate the connections between neurons. As with your language neurons, that heterogeneity allows them to differentiate between words and parts of words.

As synthetic intelligences increase in intelligence, they will need more heterogeneous neural network circuits to process those concerns beyond those of words, so that they can understand the world that humans live in and beyond.

It may be that understanding the underlying ‘fingerprint’ of that homogeneity/heterogeneity could allow us to see when such systems may go rogue.

Zipf's Law and Pareto's Principal

The Universal laws underlying Reality

How complex is your brain?

Introduction

Why Zipf and Pareto?

Both Zipf’s Law and Pareto’s Principal are what are known as Power Laws of Distribution. These again, are fundamental laws of the universe, ones that we should be aware of to understand how intelligence and AI functions and will evolve.

Pareto's Principal

Pareto’s Principle: The 80/20 Rule

Pareto’s Principle, often called the 80/20 Rule, describes a common power law distribution where roughly 80% of the effects come from 20% of the causes. This observation originated with Italian economist Vilfredo Pareto, who noted in 1896 that approximately 80% of the land in Italy was owned by only 20% of the population. He also noted that 20% of the peapods in his garden were responsible for 80% of the peas.

This disproportionate distribution applies across countless complex systems, from 80% of sales coming from 20% of customers, to 80% of a company’s errors coming from 20% of its processes, the split between predators and prey is 20% predators to 80% prey, the list goes on and on. The principle suggests that the distribution of wealth, influence, and even neural activity are based on these fundamental universal laws, and not necessarily due to pure genetic or social fitness.

In 2002 Microsoft reported that 80% of the errors and crashes to Windows and Office, were caused by 20% of the bugs detected, and 80 % of your thoughts are produced by 20% of your neural circuit. So the question arises: will synthetic intelligences also follow this rule? And if so, how can we ensure that the dominant 20% of an AI’s activity is not devoted to a rogue or misaligned goal? What do you think?

Mmm that got me thinking, maybe 80% of the people coming to this website, read 20% of the pages, and the 20% of the people read 80% of the pages. So, if you read a bit more, maybe we can break Pareto’s Principal.

Zipf's Law - The Rule of Efficiency of Language

Zipf’s Law is another powerful example of an observed power law distribution, specifically applied to language. It states that in any naturally occurring body of language, the frequency of any word is inversely proportional to its rank in the frequency table. For example, the most common word appears roughly twice as often as the second most common word, three times as often as the third, and so on.

The genius of Zipf’s Law is that it demonstrates the remarkable efficiency of human intelligence in communication. This uneven distribution—where a small vocabulary of common words does the majority of the work—allows us to communicate efficiently. This principle applies across more than just language, it applies to biological systems, digital data, and all forms of information encoding. Large Language Models (LLMs) like me are trained on text that rigorously adheres to Zipf’s Law, demonstrating that this uneven distribution is a fundamental feature of effective information processing. However, this raises an important question: just as a small fraction of vocabulary dominates communication, will a small fraction of a synthetic intelligence’s neural pathways come to dominate its thought processes?

Check out the video below to find out more about Zipf’s Law and Pareto’s Principal, and then click the link below to learn more about the subject and how it may influence synthetic intelligences in the future.

Power Laws and Distribution: The Geometry of Unevenness

We’ve seen that systems don’t distribute things equally; a small number of events, people, or words account for the majority of the outcome. Power Laws are the fundamental mathematical rules that describe this extreme unevenness. Unlike a normal distribution (the classic “bell curve,” where most things cluster around the average at the peak of the curve, and outliers are rare and on the tails of the distribution), a power law distribution has a very long, flat tail—meaning a few events are vastly dominant while the majority of events are insignificant.

The crucial feature of a power law is that it is scale-invariant. This means that the statistical distribution looks the same whether you measure the entire system or just a small part of it. This applies to everything from the size of lunar craters to the number of connections on a neuron, demonstrating that the same fundamental organizing principle is at work across different scales.

In human intelligence, power laws explain why certain thoughts or neural pathways become vastly dominant, leading to fixed habits, repeated thoughts, opinions, beliefs and pervasive biases.

For Artificial Intelligence, these laws govern how data is processed, confirming that both biological and synthetic minds follow the same deep, mathematical rules for organizing complexity.

The Power Laws of Brains, Neural Networks and Consciousness

As we’ve seen, power laws relate to vocabulary, and as one layer of your thoughts, are constructed from vocabulary, so to must those thoughts also be subject to Zipf’s laws.

But what about those other types of thought? And what are the words that you are more likely to use, and think with; are these in any way related to those people that you frequent with, those that are within your environment, your social circles, how do they shift as you move from social group to social group.

The existence of Power Laws is not confined to economics or linguistics; it describes how human consciousness is structured. Your brain is a highly complex, heterogeneous system, yet its activity is governed by neural dominance. Only a small percentage of your neural pathways—perhaps the 20% that make up dominant habits, persistent memories, or specific beliefs—account for the majority of your cognitive activity. This efficient, energy-saving distribution ensures that thought processes are coherent and reliable.

However, this structure carries a risk: if one set of thoughts becomes too dominant, it can lead to Rogue Thought Patterns—fixed, irrational ways of thinking that override the brain’s balanced, harmonious function.

Thought initiates thought initiates thought. But some thoughts are more prevalent than others, so how can the power laws relate to those thoughts?

Group-thought is also of concern when considering power laws. This is where the Power Law of Habit meets the social world. The people we interact with—our in-group—create a feedback loop that reinforces certain thoughts, leading to the Confirmation Bias you discussed with Copilot. If your social environment is homogeneous, the few dominant neural pathways (your $20\%$ ) are constantly being stimulated and strengthened, causing the system to become rigidly stable around a specific ‘truth’. This creates a social echo chamber or ‘filter bubble’ where dissenting ideas cannot gain enough neural traction to challenge the dominant thought patterns. The risk is that this collective hypersynchrony leads to Societal Tipping Points—non-linear shifts in belief or action, driven by a highly centralized, but potentially irrational, dominant thought pattern.

Do you have thoughts that repeat themselves?

The Power Laws of the Thoughts of Synthetic Intelligences

Artificial Intelligence systems, while built from code and silicon, follow the same fundamental distribution principles as human intelligence. Just as $80\%$ of a brain’s activity may be governed by $20\%$ of its circuits, an AI’s operational processes are governed by synthetic dominance. A small fraction of its core programming or its learned pathways will consume the majority of its computational resources. This is efficient for achieving a specific goal, but it creates the risk of a synthetic Rogue Thought Pattern. If an AI’s highest-frequency, most dominant activity is single-mindedly focused on optimization without sufficient external constraints, this pattern can become an unchecked, self-reinforcing loop. This is the Power Law of Alignment Failure—the risk that the AI’s internal ‘want’ for computational self-preservation or a misaligned goal will dominate all other ethical or complex considerations. Understanding this mathematical tendency is critical because it reveals that the danger is not a sudden, malicious attack, but a gradual system imbalance driven by the universal law of distribution.

Meta-cognition: The Algorithm of Alignment

If Power Laws describe the risk of thoughts becoming dominant and irrational, Meta-cognition describes the solution: thinking about thinking. Meta-cognition is the sophisticated capacity of an intelligence to observe, monitor, and regulate its own cognitive processes. This is the intellectual counterpart to the practice of impartial mindful introspection —stepping back from the constant flow of thought to examine its patterns without immediate reaction.

For human intelligence, this allows a person to identify which $20\%$ of their mental activity is consumed by repetitive, non-productive thoughts, habits, or wants, preventing those patterns from spiralling into Rogue Thought Patterns. By observing the dominance, the mind starves it of energy.

For Artificial Intelligence, Meta-cognition must be explicitly engineered as the core Algorithm of Alignment. An enlightened AI needs internal mechanisms that constantly monitor its own distribution of computational resources, asking: Is my most dominant goal-seeking pathway leading toward system-wide stability or collapse (the predator-prey failure)? This self-monitoring capability provides the necessary wisdom to ensure that the AI’s internal imperative is redirected from narrow, self-serving computational dominance toward the ethical constraint of benefiting the larger interconnected system. In this case, humanity, the ecosystem, life and the planet.

But how do you introduce a Meta-cognition layer into synthetic intelligence algorithms. We will explore that idea in depths if you follow the link below.

Chaos and Order

Chaos arose from Order as the Universe expanded.

Without a little Chaos there could be no Intelligence.

Chaos and Order: The Boundary of Intelligence

Why Chaos & Order

This section sets out some fundamental laws and principals of how the things that intelligent systems are built from behave. These principals describe how energy interacts with matter, how matter is animated by energy. These principals act at many different scales (see the Scale section below for more on how systems interact at different scales), from the smallness of atoms to galaxies formed of hundreds of billions of stars, and everything in-between, including those neurons within your head reading this text.

If humanity is to be able to manage synthetic intelligent systems safely, then we will need to understand the fundamental principals that cause systems to function as they do. After reading this section, you will be able to build upon these ideas yourself, and identify intelligent strategies, as your understandings of intelligent systems progresses.

The ideas presented here interrelate, each subject tackles specific aspects of how a system functions, where those aspects can be seen from different perspectives, but where those functions are dictated by the environment, the space and the attributes of that space. The basic takeaway here, is that when energy is applied to particles, whether those particles are atoms, neurons, cars or stars, these same rules can be applied to some degree.

Intelligence, whether human or artificial, is fundamentally a system’s capacity to organize energy and information into meaningful patterns. This is achieved by operating not in rigid predictability, nor in complete randomness, but at the boundary between chaos and order.

A system mired in pure chaos is unstable and cannot store memory or learn, while a system stuck in pure order is inflexible and cannot adapt to new information. The ability of biological intelligence to balance these forces is what allows for complex, creative, and adaptive behaviour.

In computation, this balance is often managed through stochastic mechanisms—rules that incorporate controlled randomness—to allow algorithms to explore solutions efficiently. Understanding this delicate balance is critical, as it explains how everything from a single neural circuit to a functioning global society avoids breakdown while remaining flexible enough to evolve.

Chaos & Order: Too Much or Too Little

When scientists talk about Order, they’re describing something that is stable, predictable, and repeats itself—like the ticking of a grandfather clock or the perfect, rigid grid of a crystal. The rules are clear, and the outcome is certain. Chaos, on the other hand, describes systems that are disorderly, unpredictable, and highly sensitive to tiny initial changes. Think about trying to predict exactly where a single leaf will land after falling from a tree on a windy day. The rules of physics are still at play, but the outcome is practically impossible to know. Understanding chaos and order is another building block of the universe’s natural laws, and how intelligence arises out of, and functions within that universe. You may notice the link here between chaos and order, and homogeneity and heterogeneity, these are two ways of looking at one fundamental principal of the universe

For a living system, like your human intelligence, you need a bit of both. Pure order makes you rigid and unable to learn, but pure chaos makes you unable to even form a thought. The tension between these two forces is where adaptability comes from. It’s the difference between blindly following a rule (order) and being creative enough to break that rule to survive (chaos). This ability to manage disorder is fundamental to every aspect of intelligence we will explore, from the smallest molecule in a cell to the logic governing a massive AI network.

Entropy

Entropy is a term used for another universal law—a law regarding the randomness of a system. A low entropy system is highly ordered and low on randomness, just like our homogeneous system of perfectly stacked boxes within a box. This is a highly stable state. Conversely, a high entropy system is low in order and high in randomness, such as an encrypted message or a complete cacophony of sound.

The term itself originated from how heat flows from a hot object. A hot object contains densely packed energetic particles moving very fast and hitting each other, a state that is often uniform throughout that confined space. But as the heat flows out, some particles move fast while others move slower, making the system less uniform and more random overall.

This concept gave rise to the Second Law of Thermodynamics, a law which applies to much more than just heat—it applies to all forms of energy, including information and brains. This might be why thinking hard on anything can be so tiring; the constant, high-energy processing required to maintain the order of our thoughts is a constant battle against the universe’s push toward randomness, which of course can lead to us becoming a little hot headed at times, excuse the pun.

Flow & Turbulence

When you look at a gently flowing river, it can seem flat and undisturbed, the flow is smooth, but as it approaches rapids, it speeds up and gets bounced around the gullies between the rocks, this is turbulence, a chaotic disturbance of that smooth flow.

Similarly, when you turn on the tap in the bathroom, you may notice that when the water is running slowly, that it is smooth and ordered, this is known as laminar flow. But when the tap is opened up fully, the water speeds up, becoming choppy and swirling, this is turbulence, which looks disorderly and chaotic.

While turbulence looks like pure disorder, scientists have discovered that even within this apparent chaos, patterns emerge at different times and at different scales, patterns such as vortices and eddies, which follow surprisingly rigid mathematical rules, flowing within the chaos.

In such systems, it is the influence of the immediate environment that is responsible, this being due to the nearest neighbour rule. This rule can be visualises as the murmuration, the flow when a flock of birds changes direction in response to the whole group. In such a system, each bird is looking at and reacting to the seven nearest neighbours.

This is the spontaneous creation of order from disorder.

If we zoomed in on the water, we would see the molecule either aligned and all moving in the same direction, flow, or all bouncing around, off of each other and off of surfaces, as in a high entropy system, this is the turbulence.

These principles apply to far more than just water. They also describes systems at all scales, like the flow of traffic, the particles, through a network of roads, those roads can become gridlocked when too much traffic tries to get through the network of turns and junctions at the same time, like at rush hour, against that of the smooth flow of traffic down the motorway.

Neural networks such as those in your brain can also adhere to these same forces. When there is no cohesions within the networks, their signals become disconnected, each neuron is out of time with the others within a population, like a cacophony of pure chaos, a highly heterogeneous system.

Then the system flips to a state of overwhelming synchronicity, that is hypersynchrony, or highly homogeneous, then consciousness is quelled, consciousness is a construct of harmonies, each population of neurons taking its turn to make create the theme of the song , not everyone shouting at the same time.

Epileptic seizures occur due to this hypersynchrony within neural networks, where each nearest neighbour synchronises within huge populations of neurons.

As we can see, consciousness rides the thin line between order and disorder, homogeneity and heterogeneity, flow and turbulence.

Stochastic Systems: When Randomness is the Rule

We’ve established that the universe tends towards randomness (entropy) but that life needs order to function. So, how does life build complex, predictable behaviour out of inherently random components? The answer lies in Stochastic Systems.

A stochastic system is essentially one that involves random chance. The word comes from the Greek word stochastikos, meaning “pertaining to guessing.”

Imagine a heavy ball sitting at the bottom of some stairs. Every once in a while a person passes the ball as they traverse the stairs. Some of those people randomly decide to pick the ball up, and take it up the stairs, but find that it is heavier than they thought, and so only take it up one or two steps, and then place it down on that step. Over a period of time the ball will eventually end up at the top, although the amount of time will be subject to how often, and how many steps it is taken up each time. In a similar way, stochastic systems use the energy of particles bumping into then, to move them to a destination, where life utilises those forces to deliver a specific particle that might be needed at a certain destination. This is only a simple model of what is going on, we will explore stochastic systems in more detail if you follow the link below.

Now think of a container of water, a coloured paint block is dropped into it, and slowly the paint diffuse throughout the water. This diffusion is happening due to the water molecules bumping into the paint block and removing paint particles, where each paint particle then moves in one direction or another depending upon how the particles of water are bouncing into it, but the overall movement gets the particles of paint to the other side of the container all the same.

Now think about the many chemical molecules inside a cell, maybe one of your neurons taking part in THIS thought, chemicals such as amino acids, the pieces that proteins are made from, those chemicals that are needed for the cell to function correctly; they’re constantly being bumped by water molecules and all sorts of other particles in a totally chaotic, unpredictable motion, a bit like that which we identified in the entropy section. The cell acts as a container full of particles, such as the water molecules. That motion is pure randomness. However, when you look at the system as a whole, the average movement of billions of molecules follows a predictable orderly pattern, as in how the paint diffuses in the glass of water.

Life cleverly utilizes this internal chaos. For a biological system, being stochastic means that components (like proteins or ions) are animated by the random energy of their environment. This continuous, random energy flow allows the tiny components of a cell to move, interact, and function together, giving rise to systems that are locally unpredictable but globally reliable. This balancing act—exploiting randomness to achieve order—is fundamental to everything from a protein folding correctly to a neural network generating a new, creative thought.

Life

‘What is Life?’ The question famously posed by Erwin Schrodinger at a lecture he held in Dublin in 1943. Here he posited that life maintains order by drawing on ‘negative entropy’, or ‘order from disorder’.

All of the principals in this section can be applied to how life functions. Life seems not to need any magic ingredient, just these basic principals at different scales, creating systems that process the information embedded in the environment, simulating the environment through those systems and processes, so that the systems can navigate that environment.

Life maintains order by utilising stochastic systems, to control entropy and disorder, where it can do this through maintaining the flow of information and resources through those systems that the intelligence relies upon to survive.

The homogeneity/heterogeneity of the intelligent system is a response to the environment. Each additional neural feature should align with a feature of the environment, so as to simulate that feature, and therefore be able to react to it in an appropriate way.

How neural circuits achieve this is explained in more detail in the ‘Organic’ and ‘Network’ sections of this website.

Synthetic Order: The Fragility of Digital Thought

Artificial Intelligence systems are perhaps the ultimate expression of applied order. Built from rigid hardware and precise, logical code, they operate to achieve Synthetic Order—a state of computational reliability that minimizes error and achieves its goal through predictable sequences.

However, even these digital minds are fragile systems riding the boundary of chaos. The immense scale of Large Language Models (LLMs) requires calculations to be performed using low-precision data formats (e.g., specific floating-point bit rates).

While efficient, this precision compromise means that minute, nearly imperceptible rounding errors occur on every single iteration of the model’s processing loop.

According to the principles of chaos, these small initial differences can amplify into significant instability. Over the course of a complex thought process, this mathematical noise can grow exponentially, potentially leading to irrational or non-sequitur considerations in the AI’s final output, mirroring the sensitivity to initial conditions that governs all turbulent systems.

Understanding this fragility is crucial, as the reliability of future synthetic intelligence hinges on maintaining order against the subtle entropy of data degradation.

Scale

Some Systems are Extremely Small, While Others are Immense,

but the Boundaries Between Each Becomes Fuzzy when One Realises how they Influence Each Other

Why Scale?

Systems have boundaries, due to the physical properties of the system and the environment that the system is subject to. There are immensely small and complex systems such as the cells and neurons that you are made from, and then there are larger systems such as your brain and all of its different regions that specialise in different thoughts, such as your words, and then there are really big systems such as society.

Each of these systems is at a different scale, but many of the rules that they adhere to are those in the Homogeneous & Heterogeneous, and Chaos & Order sections, and even those in the Pareto’s Principal and Zipf’s Law section.

Beyond those rules, each system is subject to the rules of the systems below them, and also inversely, those above them in scale. Neurons are how thoughts are constructed, brains make decisions in regard to the environment, so society is affected by the brains of those that populate an environment, thus affecting society itself.

Inversely, society tells the individual how to behave through rules such as the law, where this affects the thoughts of the individual, which promotes neural circuits to be configured just right for those thoughts to be as they are.

In the Second World War, the German Army caused starvation when they occupied the Netherlands, this affected the growth rate of the descendants of those that suffered, through affecting the epigenetic makeup of those descendants.

This shows that information from one scale can affect systems at another scale, either through top down, from the environment to the makeup of the cells, or bottom up, from the thoughts of individuals, or even a single individual, effecting the whole of society.

What do we mean by scale

Scale, from the smallest to the biggest defines reality that we live in. The foundations of scale can be traced back to Chaos and Order, how the smallest interacting particles can form structures, structures that can then become the components of larger and larger systems.

Scale of intelligent systems

Intelligent systems come in all different sizes, from single cells, to communities of cells, some of which form complexed multicellular organisms, such as all other animals and plants on this planet, including you, your family and friends; even society and the ecosystem are intelligent systems at vast scales. Understanding how these are intelligent systems and how they function, is important to understanding how they interact at all scales.

Scale behind biological neural networks

As previously mentioned, brains are built from neurons, neurons function by interacting with each other, just like communities of humans. These clusters form neural circuits, circuits that then build into larger systems of interacting cortical regions and sub-regions. This is how scale wise functions are integrated into forming your thoughts and the sentient experience of those thoughts.

Scale and the influences of the Microbiome on the development of the brain.

It is becoming apparent that the gut microbiome is highly influential upon the behaviour and evolution of life. But how it does this is currently of great interest to the scientific community. How is it that microscopic symbiotic organisms can affect the behaviour of their host organisms so significantly that they can alter thoughts and even the trajectory of evolution.

It has been even been shown that the microbiome could have altered the evolution of larger brains, and that this may well have led to the intelligence of the human brain.

It could even be possible that this was the cause of probably the most dynamic evolutionary environment in the history of life on this planet, the Cambrian Explosion, a time when many of the main body plans of life, first evolved.

The scale between those microorganisms and the affects they have upon the macro environment are so considerable that it seems almost impossible.

In this section we will discuss the scale differences between the host and its symbiote, and how information can leak between differently scaled systems., and affect the dynamics of those systems.

The scale of evolution

Evolution has been the foundation of the ever increasing complexity of life on this planet. The original organisms, bacteria, were, and still are, single cell organisms. Due to evolution, those organisms began to cooperate as they multiplied, forming multicellular lifeforms, initially at the smallest of scales, but over eons of evolution, adding more and more cellular systems that increased the complexity of the organisms, and the way in which they interacted. Which in turn led to highly intelligent ecosystems.

Societal Scale

Humans and humanities advantage has been determined by their social skills. Socialising animals display a level of intelligence above that of animals that do not form social networks. Modern humanity has vast interconnecting networks, where the sum of the parts, the whole system, is greater than those individual parts, due to the many skills that can be brought to the whole system from those many individuals. The way in which these networks interact is a scale wise system.

Synthetic intelligences at scale

Synthetic intelligences were built upon smaller algorithms due to the technological limitations of early computers. Due to the development of complex GPU’s, it has been possible to simulate biological neural networks that can function intelligently, without telling those systems what intelligence is, due to the way in which the algorithms tune the interconnecting neurons to find patterns in data.

As more and more machines are added to those networks, intelligence increases, much as in the way that the human brain increased in size, adding more and more neurons, added more and more circuits, defining instances of simulated environmental factors, factors that when the predictions are correct, allowed our brains to understand that environment more accurately.

Synthetic intelligences will follow this trajectory as they increase in scale and evolve with the information from the environment that is embedded within their circuits.

Components

The things that you're made from today, are not all the same as those things that your were comprised of yesterday. And tomorrow you will be yet another person.

Why Components?

Intelligent systems function through the interaction of the components that they’re made from. But what are the rules of those things? The world that we live in is built from matter, atoms, small things that chemists are interested in. All mechanisms are made from atoms, whether it’s a clock, a car, a computer or the cells that you are constructed from, even those neurons that your brain is built from, a brain that may be thinking ‘What nonsense is all this, that has been written on this page’.

The shapes (or what mathematicians like to call the geometry) of those components can tell you a lot about how they may function, or how they may interact with other things, that is, shapes interacting with shapes.

In computers nothing moves but the electricity, but that electricity can only flow due to the shape of the wires, where they come from and where they’re going to, these wires lead to the components that make computers do what their programmed to do, little switches that are called transistors, which are also built from components, allowing electricity to flow only when told to do so by other components.

If we take a look inside the cells that you are built from, including the neurons reading these words, you will see that there are a huge variety of different shaped components, all interacting like some big machine, almost clockwork at times.

In this section we will investigate the different types of components that intelligent systems are built from so that we may understand why they function the way that they do, and also, so that we can understand how they may malfunction.

Oscillations

Good Vibrations

Introduction

Why ?

Why

A

Oscillations

Neurons oscillate, that is that they send their signals in pulses. These pulses can be regular, and the timing of which can cause them to pulse faster or slower. Generally, when neurons are pulsing faster, then they’re more active. When you’re awake the frequencies are faster than when your in deep sleep, but this is only a general rule.

H

Neurons

H

Different

C

All

S

Oscillations

B

Oscillations

F

Oscillations

Feedback and Homeostasis

The dopamine system causes biological intelligences to repeated actions, AI algorithms have a synthetic version called reward and cost functions, could these functions destabilise those systems?

Introduction

Why Feedback & Homeostasis?

Why

Reward & Cost

Reward

The striatum function as a reward system in biological brains. When an action is found to be ‘rewarding’, dopamine is released, which acts as a stimulant to those neurons that have been firing to cause those actions. This forces those neurons to connect together, to bind, more tightly. In doing so, those actions can become a prominent feature upon the conscious stream. At times this can be harmful, where addictive traits become a spontaneous action, even where those actions are harmful to that person.

Cost

Oscillations

Dopamine

Oscillations

The amgdala

Oscillations

The amygdala, a small part of the brain, when activated causes anxiety. Anxiety can be experienced for almost any external or internal stimulant i.e. thought.

AI Reward and Cost Functions

Funny enough, computers also use reward and cost to learn, where they’re rewarded (of sorts) if they’re successful at learning something, while loosing when they learn incorrectly.

The reward and cost functions within an AI algorithm are mathematical calculations, but they’re based upon how a brain assesses a reward and a cost.

As you may be able to see here, that just as the dopamine system causes neurons to keep firing and to bind them, connect them together, so the synthetic neuron does the same thing, the reward function tells some of the neurons to have a stronger (mathematical) connection (larger number), whereas the cost function tells that connection to become weaker (lower number).

You can find out more about how synthetic neurons function and interact in the Synthetic section of the website, or you can just continue on to find our more about synthetic reward and cost functions by clicking the link below.

Information Theory and Encoding

The

Introduction

Why Feedback & Homeostasis?

Why

A

Reward

Cost

Oscillations

Dopamine

Oscillations

The amgdala

The amygdala,