Tag Archives: Artificial Intelligence

Why this will be the Century of Complexity…


ALTHOUGH science may have spent the last 400 years honing its understanding of The Linear Dynamics of Cause and Effect”, the reality of everyday life in the 21st century is that the really interesting stuff is increasingly the result of “The Nonlinear Evolutionary Dynamics of Adaptive Integration and Emergent Complexity”…

Linear and Nonlinear Dynamics

“Physics” is the ultimate science of cause and effect.  Physicists like to describe their science as the hardest of “hard science” because physics can claim to be governed by hard and fast scientific “Laws”.  This of course would seem to imply that many of the so-called “soft sciences” are in some way not quite as elevated, not quite as good.

In truth however we could say that physics is an “easy science”, and the soft sciences are “difficult” because the “laws” of physics only really work in the absence of “noise”, and yet the everyday world of the soft sciences is full of noise because most everything is continually battered and buffeted by “constantly changing feedback” which can generate wild “nonlinear dynamics”.

In reality all dynamics have feedback (and resultant nonlinearity), it is just that some dynamics have much less feedback than others.  Physics is, in a sense, the science of “dynamics with negligible feedback”, the science of “linear dynamics” — or in other words it is the science of the nonlinear stuff that can be safely “compressed” into neat linear differential equations which express neat linear “cause and effect”.

New Paradigm

In the simplest possible terms, linear dynamics are dynamics where the effect is proportional to the cause, and nonlinear dynamics are where the effect can be disproportional to the cause.

Physics, it would be fair to say, has throughout its 400 year history, actively tried to steer clear of messy nonlinear dynamics, and in so doing has actively established a paradigm of linear dynamics; a linear paradigm of cause and effect.

But then, out of the blue, in the latter part of the 20th century, along came both “Chaos Theory” and “Complexity Theory” which between them seemed to hint strongly at a completely different paradigm; a nonlinear paradigm of “Integration and Emergence”…

Chaos Theory and Complexity Theory

Unfortunately however nobody seems to have been paying the proper attention, and so chaos theory and complexity theory, as they stand today, are still a bit of a mishmash of concepts and don’t really have agreed upon definitions.

Chaos Theory, for example, is generally associated with the relatively vague notion of the so-called “Butterfly Effect” (or as the academic community like to say “sensitivity to initial conditions”), but this association has, in my opinion, done more harm than good — for it is misguided, and its misdirection has merely served to mask the true nature of chaos.

Complexity Theory is similarly afflicted, but rather than analyse all the pros and cons of all the various definitions of both Chaos and Complexity, I will instead simply offer my own definitions…

Defining Chaos and Complexity

In my opinion chaos is not primarily characterized by sensitivity to initial conditions; but by emergence of decisions points and the resultant sensitivity to choice.  Chaos is simply “adaptive instability”; it is “unresolved internal adaptation to feedback, surfacing as turbulent diversity on the system level”.  So, in the simplest possible terms, we could say that

“Natural Chaos is Incompressible Adaptive Diversity”…

Complexity is simply the resolution of adaptive instability.  Complexity is the result of the “adaptive  integration of co-emergent diversity”, which ultimately results in “emergent complex systems” that have effectively “organised themselves into existence”.  So, in the simplest possible terms, we could say that

 “Natural Complexity is Self-Integrated Diversity”

Matrix of Universal Dynamics - Copyright - Kieran D. Kelly

Integration for Free

So is any of this important?  Absolutely it is!   We live in a world of nonlinear dynamics, some of it compressible, most of it not.  Physics and Chemistry generally deals with the compressible stuff, but Natural Evolution and Emergent Complexity deals with the rest…

To understand Natural Evolution and Emergent Complexity is to understand how in any system of adaptive entities many diverse things can randomly occur and be reinforced; but furthermore, and much more importantly, it also tells us that

With the co-emergence of diversity,

Complex-Integration comes for Free!…

This “Integration for Free” is evolution’s secret sauce.  Evolution drives the emergence of great diversity, which leads to the natural integration of co-emergent diversity, which drives the next level of emergence.  This constant interplay of integration and emergence means that evolution naturally ratchets-up complexity over time, and consequently “the complex whole is forever becoming greater than the sum of its less complex parts”

Natural Creativity

Some years after its publication, the English philosopher Herbert Spencer summarized Darwin’s theory of evolution as being the “Survival of the Fittest”, but unfortunately this description, although popular, is somewhat misleading.

Evolution is not about the “Survival of the Fittest”; evolution is about the “Integration of the Optimally Adapted” (or more precisely, the optimal integration of optimally adapted diversity).

There is a subtle difference between these two descriptions; the former implies anti-synergistic competition, while the latter implies synergistic collaboration.  When we look at the natural world it is obvious that Nature favors integrated diversity over uniformity.  Mother Nature does not employ an asymmetric “winner takes all” strategy, but prefers a more chaotic, but ultimately more creative, strategy of “mutual reinforcement”…

Accelerated Evolution

More and more in the early part of this 21st century we are being made to realize the creative power of complexity dynamics and its potential for system self-integration and emergence.  In some arenas such as a multicultural society, the economy, technology, the arts, and even our daily lives, self-integration and emergence is a source of great diversity and creativity; but in other areas such as financial markets, terror networks, and the global climate it can be a source of great instability and destruction.

Over the last 400 years cause and effect has told us a lot about the dynamics of simple systems void of feedback, but the dynamics of complex systems alive with feedback is a subject that is becoming increasingly relevant and important to understand.

Essentially natural complexity is, in fact, “self-integrated information”.  All natural complex systems are characterized by the fact that they have “low thermodynamic entropy, but high information entropy”; consequently these systems can be considered to be highly organized but unpredictable nonetheless.

So while most people might think about evolution in terms of “an ecosystem of plants and animals”, the reality is that plants and animal (and the ecosystem itself) are really just “information structures”, and evolution simply keep creating evermore complex information structures (structures that are ever more difficult to mathematically compress).

Evolution is fundamentally a “universal process of change”.  Evolution is not just about biology, but about all “information creation”.  Evolution created us, and we in turn create complex information structures.  In fact it could be reasonably argued that in our ever-more rapidly interconnecting world, we are likely fast approaching a phase transition in human development, a transition to a whole new age ; “An Age of Accelerated Evolution and Information Creation”.  And this coming age will be dominated not by the old linear paradigm of predictable cause and effect, but by a whole new Nonlinear Paradigm of unpredictable “Integration and Emergence”…

Algorithmic Search

The new physics of the 21st century and beyond, will be the physics of self-integrating systems and accelerated evolution.  Understanding this “new physics of evolution” will be essential if we want some control over our ever-increasing inter-connected, co-dependent world.  Artificial Intelligence (AI) is seen by many as a means of dealing with complexity and already AI is being used to get computers to learn, but ultimately this will turn out to be really rather small potatoes; the really big pay-off will come from getting computers to explore.

After studying chaos and complexity for so many years, it strikes me that the universe is not fundamentally (as it so often suggested) purely mathematical, but is more generally “algorithmic”.  The process of evolution is, as Darwin himself more or less suggested, a continual process of the emergence of ever greater complexity; a process which would seem to suggests that Mother Nature is, in fact, ceaselessly executing a form of “Algorithmic Search”, constantly seeking out the most successful combinations of integrated diversity.

If this is indeed the case then it begs the question, “Is what Nature finds simply random, or are some things more likely to be found than others?”  Well, as it turns out, chaos suggests the latter…

The discovery of chaos has alerted us to an algorithmic universe that was previously hidden from our awareness; a nonlinear universe of “complex strange attractors”.  The existence of such algorithmic attractors begs yet another question, “are there some, or even many, hidden gems (or dangers for that matter) in this nonlinear universe that we are as yet unaware of?”

In these early days of the 21st century, we are only just beginning to reach the computational power needed to address this question; and although computational exploration of complex system behavior is still an activity very much in its infancy, it is destined to grow to great importance because, for good or bad, it is safe to say that

The 21st century will be a century that embraces the Creative Power of Natural Evolution and Complexity Dynamics…

Matrix of Emergent Dynamics - Copyright - Kieran D. Kelly