Unpredictable Patterns #47: Oracles, rationality and the need for noise

About our shallow understanding of rationality, the power of evolution, mental models as proteins and Orgel's many lessons, the I Ching and the digitization of the future state spaces of our universe

Dear reader,

This week is the week it is formally ok to start playing Frank Sinatra Christmas tunes. I am just saying, to give you the go-ahead if you feel you need it. We are closing in on December, one of my favorite months, and the Swedish Glögg is flowing freely in the Berild Lundblad household (well, you can only have a small glass, and then the sweetness is overwhelming, but still) — and as we start to near the end of the year, we are very naturally turning our attention to what is next - the year to come. But how can we think about that when our own lives and our work increasingly seems stuck in ill-structured messes? The answer is not so much blowing in the wind as it is found in the oracles of old…

Ill-structured problems, messes and complexity

You are stuck. You want new perspectives, but you have thought this thing through a million times - and you don’t know how to deal with it. The problem you are facing is a wicked problem, a mess - something that does not lend itself to structured problem-solving in any conventional way. What do you do?

The answer to this question is surprisingly important, and that is because most of the problems that we face are truly ill-structured, wicked or messy - they do not admit of simple solutions and easy answers. Each answer seems to lead to a new dilemma and second order effects multiply exponentially, until you almost feel like giving up even before you have tried any solution at all.

Take a second and diagnose your own problems - of the problems that you care about : how many are well-structured and well-formed problems where there are existing methods that will allow you to solve them? And how many are ill-structured and messy problems? I would guess that if you can say that 1 out of 10 problems is well-structured you will belong to the lucky ones!

And the same holds, of course, for our personal problems as for our professional problems.

So, what, then is the rational approach to problems like this?

Noise has structure - all things do - and so can help interact with other things that have structure…

One possible, and perhaps surprising, answer is that you may want to inject a little bit of noise or randomness into the process in order to stabilize it. By approaching an ill-structured problem with a complex noisy method or process - by injecting noise - you may be able to re-balance the problem.

The benefits of noise

Noise is useful in many different ways. In machine learning you can combat overfitting with noise, and create a more robust training process. In cryptography noise can be added to create completely secure communications. Statistical noise can protect your privacy.

The usefulness of noise points to something profound - randomness can be used to manage complexity.

This is not just a theoretical insight, but also a strategic insight. Thomas Schelling suggested that projecting a little bit of irrationality, injecting a component of random noise into your opponents model of what you are likely to do, is a great way of gaining a strategic advantage: if they know that you will decide some things as if they were random - then the ability to calculate your next action is severely reduced.

But how can you inject noise into a problems solving process? What methods of randomness can we benefit from and how can we benefit from them?

Our desire for rationality

Before we dig into this we should discuss rationality. There is a growing movement in tiger world today that is centered around the ambition to become more rational. A movement that argues that we need to ensure that our own behavior can be defended from a rational point of view.

This movement is not new, but rather builds on the ideals and ideas embedded in the enlightenment, but it has sharpened its tools and arguments and increasingly is turning an individual’s level of rationality into an ethical yardstick. To be irrational is to be increasingly irresponsible.

A brilliant book - but does it dismiss the value of biases too fast?

Helped by books like Kahneman’s books on noise and decision making, or the LessWrong community, this movement is arguing that rational argument and action now his within our grasp, and we would be delinquent - morally and intellectually - if we do not take the opportunity to become more rational.

There is great value in that thought, and I believe that the enlightenment and the new rationalist movement has enormous value, but I also believe that we tend to underestimate the random, noisy and irrational.

The human mind’s capacity for rational thought is, indeed, based on a certain level of irrationality - on, for example, satsificing, as Herbert Simon pointed out. Our mortality forces a certain irrationality into the question of how long we should ponder a problem or how we choose to live our lives. There is no right way to spend our seconds and minutes on this earth.

Death is not rational, yet remains the source of all rationality.

Randomness, noise and irrationality are all anchored in death, and so as inevitable as death really is (the height of the rationalism that is emerging today is the wish to conquer even death, thus irredeemably turning life irrational. Eternal life has no anchor points for rationality, no center of gravity.

But that is a controversial view - the idea that our mortality is a prerequisite for our rationality is not widely embraced and certainly seems somewhat depressing if you only look at in passing.

The rationality movement is geared towards eradicating superstition and biases, and seeks a fairer humanity through deliberation - goals that seem impossible to disagree with, but we do not need to disagree with them to suggest that they may be limited to the well structured problems we encounter, and leave us somewhat fledgling when we look at the more complex, ill-structured messes that we more often live in.

So how do we deal with those? How do we inject noise into our thought processes?

Bits and chunks

The biologist Leslie Orgel is known for formulating a few concise and interesting rules around evolution. The most well-known law is the second one, that states - acidly - that ”Evolution is cleverer than you are”.

Leslie Orgel - reminding us that evolution are cleverer than we are.

This is a stark reminder to anyone that seems to think that the human mind can be restructured according to the principles of abstract rationality, and that evolution failed to do so. If we believe that the biases and short cuts the human brain avails itself of, are mistakes, we are very likely to have simply forgotten Orgels second law.

Biases are artifacts of cognitive economy and a much more realistic chunking of the world than the mathematical and binary chunking we find in decision theory and mathematical econometrics.

Information comes in at least two forms - the exact bit and the much more nebulous chunk (another concept from Simon). The chunk is how we group information when we remember it, integrate into our own thoughts and the problem with studies of rationality is that they are often based on a theory of information as bits, but evolution has developed our brains into tools for chunks.

Chunks are groupings of bits based on history, path dependencies and individual idiosyncrasies - and the best way to navigate these are most likely the ways that we have evolved, what is often condemned as cognitive biases.

This does not mean that they are right - from an is no ought follows - but it does mean they merit close attention.

When we say that we understand the world not as propositions, but as narratives we are making a point about chunks and the linking of chunks in time into narratives, and if we accept that narratives are more foundational to human thinking than predicate logic, we also accept that any analysis of human decision making based on the latter is going to be weaker than an analysis based on the first.

Orgel, however, also had a first rule, and that is almost never discussed - yet it is perhaps even more interesting for our discussion.

Orgel’s first rule is: ” Whenever a spontaneous process is too slow or too inefficient a protein will evolve to speed it up or make it more efficient."

So what mental constructions akin to proteins have developed to manage the weaknesses in rational thought in dealing with ill-structured messes?

Mental models are proteins

Anyone sufficiently mired in the idea that our mental capabilities are structured around mental models or tools will soon find that it is intriguing to think about this mental models and memes as analogous to proteins in biological evolution.

Proteins catalyse, replicate, respond and provide structure to cells and organism, just as mental models replicate as memes, respond to adaptive problem solving and provide structure to our cognitive work.

By playing with the idea that mental models and proteins are - at some level - analogous, we can also suggest that our mental models are not just arrived at willy-nilly but that they have evolved to respond to a slowness or inefficiency in our own thinking!

So what mental models have evolved to respond to our slowness and inefficiency in dealing with ill-structured problems?

Injecting noise with mental proteins

One - incredibly large - answer is art. Art is a way for us to gain new perspectives and develop entirely new cognitive structures. The study and understanding of art as expression without utility is ultimately misguided, but neither is art easily reduced to a simple tool. It is, however, probably understandable as a very, very complex tool for injecting noise into our own thinking. Music, literature, images — they all provide noisy inputs that re-calibrate us in ways that help us out from the iron-prison of rationality.

There is usefulness in art, and that is an oft-ignored demarcation criterion!

But delving deeper into the role of art in thinking would require more space than we have here - but it may be worthwhile coming back to.

So, what we can do here is to look at a much simpler set of tools, the tools that have been used for ages by the ancients: different tools of prophecy.

Mankind has tried to understand the way the future will unfold as long as we have a had a concept about the future overall. Different cultures, different languages, different traditions - but everyone of them has sought to divine the fate and future of our ambitions.

One, large strand of the history of magic - and magic is nothing else than a mental model where causality is layered in complex ways - is the history of fortune telling and divination.

Delphi - a mental model in stone

Many different methods have been used for divinations, but they all have a few elements in common:

• First, divination is a dialogue with nature. If it is through the laying of the tarot, or the throw of the dice, if it is through the throw of coins or the visits with the oracle in Delphi, mankind has sought the future in questions and queries made to nature.

• Second, the answers that fortune telling delivers have always been extraordinarily open and generative. Few fortune tellers answer the question we ask - most give us an answer that requires an enormous interpretative effort.

• Third, the art of divination is associated with answers we do not want. The key to a good fortune is that it is hard for us to integrate in to our own beliefs about the world. All the stories we have about fortunes and divination are tales about how we deny, suppress or ignore the true import of the message that is right there, in the fortune.

Together these three qualities suggest that fortune telling is a mental model that has developed to force us to adopt interpretations on ill-structured problems that are counter to what we want.

Divination is perhaps foremost a mental model developed to deal with the inefficiency of rationality and the slowness of the analytical process. It short cuts the process of deliberative thinking and requires us to restructure our approach to the problems we are solving.

Asking the I Ching about tech lash

Let’s look at an example, or a case study, from one often oldest known divination methods - the I Ching, or Book of Changes.

The I Ching is a remarkable book in many regards. It is, for starters, a digitization of the future. The I Ching represents the future in 64 hexagrams with whole or broken lines - a binary system that represents all possible futures.

This is your future - digitized, coarse-grained and frame-breaking

That the I Ching is a digitization, coded in binary code, was obvious early on to thinkers like Leibniz. When we look at it today we can see that it has important lessons to teach us about digitization. Not least that digitization is not, necessarily, a more exact or precise representation of the world. The I Ching is noisy, and exceedingly so - it represents all future possible state spaces through 64 hexagrams - even if that simplicity is only illusory.

We should step back, and study a (much simplified) use case.

Let’s say that we want to understand how to deal with a current political problem - like the tech lash.

The tech lash is a huge problem for anyone working with the technology industry, but it is an ill-structured problem: all attempts at simplifying and structuring this problem are in vain. The problem remains a mess even after we do polling, focus grouping, sentiment analysis, conceptual mapping…it is simply a mess. So how should approach the problem? The suggestion that we have here is that we should inject noise, and use the memetically evolved tools we have to do so.

We should, then, ask the I Ching what we should do to solve this problem.

So we need to frame a question, and we should spend some time doing so - really agreeing what the question is that we want to ask the Oracle. In this case, for practice and in some humility, we will use the question ”How should we deal with the techlash?”.

This focus question is our starting point, and as we have jotted it down we proceed to use the random process used to consult the I Ching.

Here we should note that the idea that randomness is used to combat complexity is incredibly modern. The I Ching here represents a very modern view of complex problem solving, and the use of a random method to cast the divination is key.

Now, the I Ching offers two traditional methods to consult the oracle. One is through the use of coins, where we cast three coins with clear heads and tails.

When the cast the coins and if we get three tails we get changing Yin - a broken changing line (one of the forces in Yin and Yang), if we get 2 tails and one head we get a whole line, Yang, if we get two heads and one tail we get a broken line, Yin. If we get three heads we get changing yang, a changing whole line.

We will come back to the changing, whole and broken lines - but for now let’s just accept that this is how we build a hexagram, a set of six lines.

Once we have thrown the coins six times they will have allowed us to answer the question we have asked with one of the 64 hexagrams.

For our example, I cast the coins and the answer I got was this:

Hexagram 52 is called The Mountain, and the enigmatic lines that it offers suggests that:

”Above this Mountain’s summit another more majestic rises:The Superior Person is mindful to keep his thoughts in the here and now. Stilling the sensations of the Ego, he roams his courtyard without moving a muscle, unencumbered by the fears and desires of his fellows. This is no mistake.”

This is classical divination - enigmatic, compressed, difficult to read - and this is exactly the value of fortune telling. It forces us to unpack what we find and then re-structured our ill-structured problem with the help of the noise in the process.

Here, if we wanted to just do a quick reading, we could suggest that tech companies are spending far too much time on the tech lash and that this is detrimental to them. That it is better to focus on new products, and not the ”fears and desires of his fellows”.

The hexagram then contains two chaining lines and when you shift these to their opposite they reveal another hexagram, the way our issue will evolve, the next state it can enter:

This one is representative of where the situation is heading and what you need to do to get there -”responsive devotion, receptive influence” — and then the interpretation:

”This is a time for dealing with reality as it is, not as you would have it be.If you realize that in this situation you are the receptor, not the transmitter of the stimulus, you will find yourself reaching goals that seemed unattainable under your own steam. If you persist in futile efforts to be the Shaper rather than the Shaped, you will completely miss this unique opportunity.”

Now, this is a real example. I just consulted the Oracle, and it suggests that — what?

This is the real question: what is happening here is that we are forced to re-frame our problem after the random, noise-like - but structured and historically evolved - injection of the Oracle. And suddenly we can see that there is a world in which we do not care about the tech lash, and that success may be found in realizing that we are the Shaped, not the Shaper!

This brutally simple example shows how divination can be used to re-frame a real, complex problem, and the way it does that is through the injection of random noise into our own thought process.

Divination is a mental model - a protein - evolved to address the inefficiency of rational thought.

So what?

What is the real take-away from all of this, then? It is obviously not that we all should rely on divination techniques, oracles and fortune telling for problem solving. But it is the very real question of how you inject noise into your own cognitive patterns. If it is through art, conversations with others who are cognitively diverse from you or ancient oracles does not matter that much, but you need to cultivate tools for shaking yourself out of your own frames.

These tools, universally, have in common that they impose a certain interpretative cost on you: they force you to spend time integrating a randomly or pseudo-randomly selected pattern into your own.

And this is where the genius of long-established forms of fortune telling like the I Ching, the Tarot or even astrology has a quiet strength: they are patterns adopted and evolved over time to sum up future possible spectrums and events. The I Ching is not a random set of 64 hexagrams, it consists of a carefully curated model of possible future branches of events assembled over millennia. It is a coarse-grained but exhaustive digitization of possible futures.

We should not abandon our rationality; but we should not enshrine and adore it either. We should weave it into that complex weave of humanity that is open to the structured randomness of fate and prophecy, emotion and irrationality - because that is what the world looks like.

This is who we are.

Thanks, as always, for reading.

Nicklas

---