Key themes required for understanding

It now seems beyond any shadow of reasonable doubt that all of our understandings and experiences of reality are some sort of model, and therefore necessarily some level of simplification of reality itself.

That has both power and risk.

The power is that our simple models work in the situations that they do.

The biggest risk in such a system is that the context may have changed such that our simple models no longer work as well as they once did, but we are blind to that because our experience is of our models, not of reality directly (but not many are yet aware of that fact).

We need to be conscious of both the strength and the risk in this aspect of what we are!

So what sort of ideas are important to getting some sort of a “broad brush stroke” picture of what we are, what we might be capable of, what sort of dangers and opportunities might be present, and how we might explore the opportunities while mitigating the risks?

This is intended to be a very brief introduction, a brief sketch of a couple of dozen or so ideas that are fundamental to beginning to build a modern understanding of what we are, and what sort of systems we need in place to maintain the sort of social and ecological conditions necessary for individual life and individual liberty, at the same time as we develop effective strategies to mitigate the many different sorts of risk that we already know about, at the same time as we search for other risks beyond the boundaries of our current understandings.

The list of ideas:

what does it mean to know (epistemology and ontology);

complexity and infinities;


nature of systems and boundaries;


hierarchies of competence;

strategic environments (the need to ritualise combat);

sources of uncertainty (model, QM, error, …);

uncertainty and long tail distributions;

nature of freedom;

cultural evolution and personal development;


markets and capitalism;

the boundaries between order and chaos;

the necessity for both liberals and conservatives;

AI and knowledge;

decisions under stress (are you friend of foe, baby or snake);

appreciating the depths of wisdom embodied in our biology and culture;

the distinction between competence and power, particularly in respect of hierarchies;

some problems with political notions of our past; and

what sort of futures might be stable enough to last a long time?

In a very real sense, answering these questions will be a book, and I always intended it to be a book, and let’s call this an outline for the book.

What does it mean to know (epistemology and ontology)

Ontology is about what is there, what is “being”, what is “real”, the nature of existence.

Epistemology is about knowledge, about how we know, about what it might mean to say we know something.

It does not seem possible to entirely separate the two domains, as to know something implies something real that is doing the knowing (whatever “real” is).

And the history of these two domains of enquiry and understanding is interesting, as they have slowly, over thousands of years, built from very simple to very complex webs of understanding.

Quantum mechanics seems to now be telling us that this reality in which we find ourselves is a finely balanced mix between the lawful and the random, that in some contexts can very closely approximate hard causality, and in other contexts deliver very high degrees of freedom.   And that claim will receive much more attention to backing arguments in the book, and for now I will leave it at that.

Evolution does now give us a mechanism as to how simple systems capable of copying themselves with some level of fidelity can, over vast time, with repeated recursion, and repeated emergence of new levels of cooperation, turn into extremely complex multi-leveled systems capable of making models of themselves within their models of reality, and reading and writing sets of encoded symbols like this.

So these two domains of enquiry, that once seemed so separate, have, with evolutionary epistemology, come back together as a coherent whole.

Complexity and infinities

Complexity is a very deep topic.

One of the best simple approaches to it is David Snowden’s Cynefin Framework – which divides infinite gradations into 4 general classes, simple, complicated, complex and chaotic, and gives appropriate management responses to each different class.   David has a great video explaining it.   Some classes of things cannot be predicted.   Some classes of things are uncertain, and cannot be predicted ahead of time if they are solvable or not.  It gets really complex.   The only thing we can know with confidence is that absolute knowledge is not an option, ever.

Infinity is a really difficult idea to deal with.   The idea that there may be no end to something, that however far you have gone in your explorations, what remains unexplored makes what you have explored a close approximation to nothing.  The old Zen Buddhist notion “that for the master, on a path worth taking, for every step on the path, the path grows two step longer”; does capture something about the nature of any infinity.

One of the simplest infinities is the infinity of whole numbers (integers – like 1, 2, 3, 4, 5, …..).  Between each integer lies an infinity of all possible fraction – all the bits that can be represented by the relationship of two integers – like 1/2, 1/3, 2/3, 1/4, 3/4, 1/5, 2/5, 3/5, ….  Then there is the infinity of all possible fractions, between all possible integers.

There is a mathematical proof that the infinity of fractions that cannot be represented by the relationship of two integers is greater that the infinity of those that can.  One of these “irrational” numbers that most of us have heard of is Pi (the relationship of the diameter of a circle to its circumference in flat {Euclidean} space) .  We normally write it as 3 and 1/7, but that is just a useful approximation.   The real number has no finite representation.  One may continue computing Pi for the rest of eternity and never reach the end.   That is a very odd idea.   The idea that a number so fundamental to our understanding of reality can never be known with certainty, only ever approximated to some level that is close enough to be useful in whatever context we need it.   If that one idea alone does not deal a death blow to the idea of certain knowledge, then you need to think about it a bit more.


Recursion is a really neat idea, the idea that something can fold back on itself, potentially forever.  In some classes of systems this can lead to exponentially expanding complexity.

A very simple example is the child’s story my father taught me over 50 years ago:
“It was a dark and stormy night, and the little ship was tossed around by the waves, and the skipper said to the mate, ‘Tell us a story mate’. and this is the story he told.
“It was a dark and stormy night, and the little ship was tossed around by the waves, and the skipper said to the mate, ‘Tell us a story mate’. and this is the story he told.
“It was a dark and stormy night, and the little ship was tossed around by the waves, and the skipper said to the mate, ‘Tell us a story mate’. and this is the story he told. …..”””

In complex systems, things like recursion, with slight variations added, can lead to infinite complexity, infinite diversity, and to really complex systems – like us.

Nature of systems and boundaries

Systems have been a passion of mine for over 50 years.

Any system requires boundaries to give it form.

Without boundaries, everything mixes up to a uniform sort of randomness.

It is boundaries that allow complexity to exist.

A minimum set of boundaries are required for any complex system to exist.

The really difficult question when dealing with really complex systems is determining what that minimum set of boundaries is in any particular context.   In many real cases that can be a problem of sufficient complexity that it can have no clear and definitive answer.


Evolution is a  really simple idea in a sense, yet it rapidly gets very complex.

All it takes is three things:
Something that can replicate;
Sufficient errors in the replicating process (or from the other perspective, sufficient fidelity in the replicating process); and
Differential survival between the variants in different contexts.

And it very rapidly gets very complex from there.

If the context is such that most threats to individuals comes from other individuals from the breeding population they belong to, then competitive strategies tend to emerge and dominate and the system is driven to some “local minima” – some simple state from which change is difficult (hence we see examples in the fossil record that seem to be stable throughout time).

If the threat to individuals comes largely from factors outside the population, then cooperative strategies can emerge, and eventually stabilise with sufficient attendant strategies to detect and remove strategies that cheat on the cooperative.   This can lead to something of a strategic evolutionary arms race.   We humans seem to embody about 20 levels of complex, cooperative systems.   Our complexity is fundamentally grounded in this cooperation, and we can each compete, and do so hard, if the context demands it.   And in being the most complex entities we know of, we are also, almost by definition in this evolutionary sense, the most cooperative entities we know of.

When dealing with evolution one needs to give up “either or”s and accept “both and”s.

It isn’t “nature or nurture”, it is nature and nurture, to differing degrees in differing contexts.   We are not competitive or cooperative, we are competitive and cooperative and how much we each are of which depends very much on the context.  And as a general rule of thumb, our existence is far more predicated on our abilities to cooperate than it is our abilities to compete.   Yes we all have our competitive side, and in the big picture, it is a good first order approximation to say that all of our culture and technology, and our survival as a species, is predicated upon levels of cooperation.

Hierarchies of competence

Hierarchies of competence are universal in the animal kingdom, they are not simply “cultural constructs”.

It makes perfect sense in an evolutionary context for hierarchies to develop – they are efficient ways of using resources and reducing conflict and threat in many contexts – but not necessarily all contexts.

Hierarchies do not need to be about power or force, and power can be one of the many different factors embodied in any particular hierarchy.

Reality often demands competence across multiple domains simultaneously.

Strategic environments (the need to ritualise combat)

In complex animals like ourselves, where raising children takes decades of work by both parents, conflicts that cause serious injury have a very high cost (in terms of not leaving children in the next generation).   So there is a lot of pressure to solve conflict without resorting to all out physical violence – a lot of “ritualised combat”, at many different levels.

Every level of cooperation is vulnerable to exploitation by cheating strategies, and so must be accompanied by sets of strategies that can effectively detect and remove cheating strategies.   Arguably that is what most of our emotions, and much of our morality, and supposedly our legal systems, are about.   Unfortunately, we seem to be in one of those periods where “cheating strategies” have gained a dominant role, and we need to redress that, and bring those who have been using such “cheating strategies” back into the cooperative of humanity.

Sources of uncertainty (model, QM, error, …)

Uncertainty is everywhere.

Any non-trivial situation contains many levels of uncertainty.

I have had many discussions on the nature of uncertainty and this is one of the better ones.

We have many classes of uncertainty.

Measurement error, Hiesenberg uncertainty, Goedel incompleteness, infinities, irrational numbers, etc.

We must adopt a certain humility, even as we strive to find the most effective approximations possible for the contexts of our present and future.

And one does not have to go far into this complexity to realise that all knowledge, all understanding, is based in uncertain heuristics (best guesses at some level).

Reality doesn’t seem to give us any other option, so we may as well accept it and get on with doing the best we can with whatever limited approximations we have.

Uncertainty and long tail distributions

Most things have random aspects that give them probability distributions.  Sometimes those distributions have “long tails”.

A “long tail” on a probability distribution means that the further out you get, the lower the probability becomes, and in many cases it doesn’t go to zero.

If we are serious about long term survival, we need to look very closely at some of those “long tail” distributions.   Some things that occur very infrequently can be very important.

The Nature of freedom

As mentioned above, all systems require boundaries.

Complex systems often require complex boundaries.   Cell walls might seem simple to some, but they are very complex, being selectively permeable to some things (letting some things in easily, and others out easily) while being much more resistant to the movement of other things.   Some things are actively transported in, others actively ejected.   Those things can change with context (internal or external).

We are very complex systems.   We live in complex biological and social contexts.

Freedom cannot be an absence of boundaries (that is suicide).

Freedom, if it has any real meaning, if it is to survive, must involve acknowledging and respecting all the necessary levels of boundaries required to give the forms of complexity present (including ourselves).

Given the profound levels of complexity and uncertainty present, that seems to require of us sets of conversations to establish what sorts of boundaries are reasonable in what sorts of contexts.   That is not simple.   It is profoundly complex.

Legal and ethical systems can get very complex indeed, and one needs to be eternally alert for cheating systems within that complexity, even as one acknowledges the need for such complexity.   The current economic system may have reached (or be rapidly approaching) a stage in our development where its dangers far exceed its benefits.

Cultural evolution and personal development

Evolution seems to apply to any system that is capable of replication, and to every level where boundaries give some sort of form to some level of entity.

Thus we are used to thinking about the evolution of genes encoded in nucleic acids (RNA and DNA) leading to bodies, populations and ecosystems, but our brains and our technology (writing, pictures, stories, digital technology) now allow information to be encoded and copied symbolically in sounds, picture, actions, and a variety of digital electronic and magnetic and optical media.

Evolution works on any system capable of replication, and the speed of evolution is very sensitive to the error rate (degree of creativity from a different perspective).   Too few errors (too little creativity) and not much happens.   Too many errors (too much creativity) and the system doesn’t have enough boundary to sustain the orders of complexity present, and it breaks down to something simpler (a big issue for cultures of our past – and a danger for us now).

As human beings we seem to have between 15 and twenty levels of evolved systems embodied within us, and only one of them seems to be actually conscious, and all of them are active, doing “their thing” continuously.   Every system influencing every other system at each level.    It is far too complex to get anything other than a broad brush stroke sort of a picture.    Even using advanced computer simulations, the degree of context sensitivity to subtle changes in context make any sort of absolute prediction impossible, and one can get a reasonable idea of the sorts of things that are probable in most contexts.

Cultural evolution is closely tied to technology, and technology is on a double exponential curve of productivity.   20 years ago cell phones were “yuppy phones”, now most people have them, and they are much more powerful and much cheaper and more reliable than those early models.    Changes like that will come more quickly, with deeper impact, as rates of change continue to increase.

Stable cooperative cultures used to be limited to about 200 people (Dunbar’s number), which seems to be largely a function of our ability to reliably recognise individuals and recall our interactions with them over time and context.   Modern digital technology allows us to enhance our memories for network interaction and extend that number to any population our sun is capable of supporting.   That will mean deep change at many levels of culture, as any form of cheating on the cooperative becomes easy to identify, and track back for decades if need be.   Yet the long term benefits available, even to those whose current assets have come from cheating, far exceed anything possible by extending our current protections for high level cheating.

At the personal level, individuals will need to evolve to a level that they can accept the diversity that must result from real freedom.

We must all acknowledge the power of whatever culture gave us birth, and the necessity for such things, and we are entering a post cultural age, where all individuals’ ability to choose their own developmental path will make communication about non-trivial issues very challenging, as sometimes there are no sufficiently useful analogies in shared experience.    In such cases, stability demands an agreed set of basic values and a degree of trust.   That will be a difficult transition for many.


Ecosystems can be thought of as a level of entity, if the boundary conditions are sufficiently defined.    Those entities can be cooperative if the risk from agencies external to the population is greater than risks from within the population.

Whether a system goes towards competition or cooperation is very largely determined by where the dominant locus of risk is located.    That seems to work at every level of structure or organisation, including the ecosystem level.

Ecosystems can develop very complex relationships between different organisms and the physical environment.

Markets and capitalism

As mentioned above, markets are at base a mechanism for measuring a particular type of value – value in exchange.   When most things were in fact scarce, then value in exchange was a good proxy for value more generally, and exceptions like the air we breath (which has no market value), could be ignored.

Markets are not simply mechanisms of exchange.

Markets also perform many other complex and valuable functions including distributed information processing, distributed governance, distributed trust networks, distributed risk assessment and mitigation, several levels of resilience, and a variety of other complex network effects which are very valuable for cooperative social organisms like ourselves.

The fundamental flaw with markets, is that the value they measure is based in scarcity, and anything universally abundant has zero value in a market by definition.

In an age of exponentially expanding computational ability, with the class of goods and services that can be fully automated expanding exponentially, then the native incentive of markets fails to deliver the abundance generated universally.   That is a huge problem for fairness and justice (both highly evolved and necessary functions in cooperative systems).

Simply going to a different distribution mechanism, without first creating replacement distributed systems to perform all the many complex functions markets currently perform would not be a smart move.   And we have plenty of alternatives possible, there is just no native economic incentive in the existing system to develop and implement them.

See my page on money for a deeper look at the systemic issues with our current economic system.

The boundaries between order and chaos

The idea of a boundary between order and chaos is as old as mythology itself.   It is embodied in many spiritual traditions, perhaps most commonly known in the Yin-Yang symbol.

In evolution, the “error rate” in copying is essential to the process.   Too few errors (too little chaos – ie too much order) and there is little or no change.   Too many errors (too much chaos or too little order) and the systems cannot maintain form.   The errors imposed by sterilizing radiation destroy all living order for example, and while snap freezing may preserve order, it prevents function.

This same balance occurs at every level of existence, from the quantum all the way up.   The balance is very context sensitive.

At the highest level we know, this seems to be human existence, finding that boundary between order and chaos – a very personal thing.

Too much order, we have fixed pattern and eternal boredom.

Too much chaos, we have anxiety and terror and destruction.

We need that balance, now more than ever before, because the extremes of imbalance available to us pose existential level risk to most life on this planet.

If our models of who and what we are become too simple – we are in deep trouble.

We are complex.   We must own that complexity, and respect the diversity that comes from it.  Jordan Peterson does a superb job of pointing to some of the deep lessons of history encoded in our cultures and mythology in this context.

The necessity for both liberals and conservatives

This is just one simple example in a very real sense, of the need for balance between order and chaos.

We need both in our lives – with conservatives generally being on the ordered side of the spectrum, and liberals on the chaotic side.

And we are all liberal on some issues, and conservative on others – the need for this balance is as real within us as it is between us in the wider political situation.

Setting them up as alternatives to each other tends to create an oscillator that can tear systems apart.   That is not a stable approach.   The two need to be integrated as essential parts of the system, and must be seen as such by all involved.   The idea that one can or should exist without the other must be seen for the insanity that it is.

AI and knowledge

Artificial intelligence will get much smarter than humans in areas it has experience, and in areas that are computationally simple.

Yet there are aspects of being human that have been encoded into our genetics and our cultures over deep time that are not easily available to AI until it has been around for quite a few thousand years and run into some of those issues for itself.

There are also classes of problems that do not scale simply with computational ability.

While AI will be more powerful than us in many areas, there are likely to remain areas where conversations between humans and AIs (and the entire spectrum of entities in between) will be valuable to us all for a very long time to come.

Decisions under stress (are you friend of foe, baby or snake)

We humans have evolved in a set of contexts where stress causes our brains to simplify the situation for rapid decision making.

Faced with a charging big cat, or a coiled snake about to strike, such responses were very useful in our past.

Faced with the threat of global nuclear annihilation, or global financial meltdown, or complex cultural conflicts, that tendency to simplify things under stress can over-simplify what really are very complex situations, leading to far from optimal outcomes.

The key to making effective decisions in high stakes situations is remaining calm, and using every tool-set available in the time available to evaluate what is the most appropriate response.   That can involve understanding many very complex sets of systems and their inter-relationship.   And it seems clear beyond any shadow of reasonable doubt that the only hope any of us have for a long term sustainable outcome is if that outcome delivers justice to every individual at every level.   And justice in this sense must involve and acknowledgement of all the necessary sets of boundaries required to maintain the complexity present (and becoming), at the same time as it maximises the freedom available to every individual (and freedom in this sense must involve a reasonable set of tools and materials to exercise that freedom in a responsible fashion).   And that is a problem of sufficient complexity that it is likely to be an ever evolving thing.

Appreciating the depths of wisdom embodied in our biology and culture

Biology, particularly evolutionary biology, has had to find workable solutions to some very complex problems and some very uncommon yet important situations – the fact of our existence is strong evidence for that.

There are deep lessons to be learned from biology, lessons about trust relationships, communication networks, distributed redundancy, and many others.

Perhaps the deepest lesson is that complexity requires cooperation, and cooperation requires security within.   Those lessons need to be internalised at every level and applied in practice within our societies – universally.

The distinction between competence and power, particularly in respect of hierarchies

One of the great errors of modern times is to confuse hierarchies of competence with hierarchies of power (an over simplification that has sent post-modernism down a very destructive path).

It makes sense to have the most competent individual in lead positions.   That is of benefit to all.

Certainly hierarchies of power can and do exist, and most power hierarchies embody some aspect of competence, and in many hierarchies it is vastly more about competence than it is about power.

When danger is high, and time is short, there is real need for hierarchies to function, for the survival of the group.   When there is the luxury of time, a lull between the storms, then individuals can experiment with new and potentially better ways of doing things.   But in the middle of a storm is not the best time for such experiments.  (And of course one must be alert for artificially created storms that are present simply to sustain some power relationship – such things can happen, and that does not mean that all storms are such creations.)

Some problems with political notions of our past

There are many notions from our past that have worked well enough to survive in the contexts of their time.   None of that necessarily means that they are suitable for a future involving fully automated systems, and levels of abundance that even the most wealthy of a hundred years ago could not aspire to.

Many of our political systems seem to be based in framing order and chaos as opponents in a battle, rather than as necessary parts of the cooperative system that must be in a balance with the specifics of any particular situation.   That framing as opposition sets up a tendency to oscillate that creates dangers that are now causing existential level risk in and of themselves.

It now seems beyond any shadow of reasonable doubt that we must accept that at every level there must exist some sort of balance between order and chaos, and that such a balance must be very sensitive to the specific contexts of the situation.   That demands of each of us all sorts of qualities that go under headings like wisdom, judgement, tolerance, acceptance, strength, forgiveness, love, etc.

Such a reality must contain uncertainties, that is a given.

What now seems beyond any shadow of reasonable doubt, is that the only alternative to the acceptance of such uncertainty is the certainty of our destruction.   If that is not clear, then all that means is that the models being used are not sufficiently complex to give a reasonable approximation of the reality that does in fact seem (beyond any shadow of reasonable doubt) to be present.

The old politics of a conflict between opposites must be replaced by an acceptance of the necessity of diversity and the fundamental necessity of both order and chaos at every level of structure.

There is no shadow of reasonable doubt that human beings are the most cooperative species yet to emerge on this planet, even if most are not yet consciously aware of that fact.   The complexity that sustains us is predicated on cooperation, not on competition.

Competition can be real, and it is not the source of either creativity or complexity, its systemic direction is towards simplicity.

What sort of futures might be stable enough to last a long time

Futures that empower every individual, with all the tools and resources to do whatever they responsibly choose, seem to be the most likely sorts of future to have a reasonable probability of lasting a very long time.

And to choose in this sense an entity must be of sufficient complexity to create models of reality that are sufficiently complex that they contain a model of itself as an actor in reality, and are sufficiently flexible that they can imagine multiple possible alternatives, and choice is some function that allows actions to be taken in reality that favour one of those options over others.   And such a choice function must contain some balance between order and chaos.   All human beings do in fact seem to be such complex entities.  Some AIs seem to be rapidly approaching such capabilities.

And what is reasonable will likely always be context sensitive in some very real senses, at the same time as it demands social and ecological responsibility from every one of us – a general duty of care to maintain those boundaries necessary for all complex life (human and non-human, biological and non-biological), and a simultaneous need to test them from time to time – just to check they haven’t changed significantly.