Artificial Intelligence

The Mind of a Rock: Animism and AI

For thousands of years, Western civilization has been living with a striking paradox. On the one hand, we are clearly physical beings living in a physical universe. And yet, we have these thoughts, feelings, dreams, and perceptions… They seem related to the physical universe, yet fundamentally different in character. We have an ‘inner’ life, which has its own colors and sounds and structure, operating under a whole different set of rules. In the physical world, I’m 3500 miles from where my body was born; but my mind instantly recalls the name of the state, county, and town where that happened, and gives the exact date and time. And yet, since I have no memory of the actual event, in a way my mind can never go there at all — it’s as though I can visit the post office box instantly, but never get to the house itself.

Mind and body seem so different that it’s almost as if they belong to separate worlds entirely. No doubt this is why it’s been so easy for so many people to believe in a ‘soul’, a mind that can be separated from the body and continue its life, in its inner world, long after the body has died — or even enter another body entirely. This despite the fact that the mind is obviously affected by physical events: it becomes sluggish and unfocused when the body is tired or sick, and it can lose memory or skills or even suffer a change of personality if the brain is injured or chemically affected.

Over time, two main camps have formed around this paradox. The first, as I’ve mentioned, believe that the soul or mind is separate from the physical body, and is fundamentally made of a different kind of stuff; and when the body dies, it moves on to some other realm, or finds another body. The second camp believes that the body creates the ‘mind’, perhaps analogously to the way a computer executes instructions in a computer program, or the way a flautist plays a melody. The mind — the ‘inner world’ — is generated by the brain and will come to an end when the brain stops working, just as a melody stops when the flautist puts down the instrument.

In ‘The Wakeful World’, Emma Restall Orr tackles this paradox, and (1) shows that both the solutions above are lacking in serious ways, (2) points out that there is another solution — indeed, a multitude of other solutions, which have been suggested at one time or another over the past few thousand years, and (3) offers her own take on the problem. In this article I’m mainly going to skip over (1) and (2), since there’s no way I could do Orr’s treatment justice, and instead briefly (and necessarily crudely) describe some aspects of (3) and look at some things that follow from it. In particular, Orr’s take not only leads to the idea that rocks think, but answers why human brains think differently from rocks, and gives a new view of the place of the human experience in the ecology of mind.

Inspired World

So what is Orr’s take? There’s no way to describe it briefly and do it justice, but for the purposes of this article, part of Orr’s view can be summarized in this nutshell:

  • There’s not just one kind of consciousness. There’s normal waking consciousness, the run-of-the-mill I-think-therefore-I-am consciousness, but there’s also dreaming consciousness, woolgathering consciousness, drunken stupor consciousness, and so on. There’s consciousness when you’re so absorbed in a movie or a task that you’re not self-aware at all, at least in the literal sense of ‘being aware of the self’. And there’s perceptive consciousness, which is simply perception, without any differentiation of perceiver and percept. It’s all mind.

  • There is no mind-body split. It’s not that mind is some kind of ghost or spirit that inhabits dead matter. Mind is a property of the world, just as naturally and inevitably as wetness is a property of water. Mind is world, and every part of the world is “minded”, in some way or another. Naturally, different parts of the world have different kinds of consciousness at different times, but it is always and everywhere there. There is something it is like to be an electron.

One concept she uses in particular is the “lit mind”:

“The self which is wide awake to the experience of the moment, consciously aware of the sensation of being, thinking about and considering the data of its world, this is the lit conscious, the lit mind.”

It’s a useful concept, because it immediately implies the “unlit” or “dark” mind — dark in the sense that, while it is real, it is imperceptible to the lit mind. The dark mind is where dream images arise and unbidden urges take root, but it is also the source of inspiration, intuition, and empathy. She says:

“The mind is a vast and ancient wood, the density of its canopy allowing sunlight only now and then to reach through and touch the spiders and mushrooms in the leaf mould of the forest floor. This is who we are… From the fleeting moments of the lit mind where conscious awareness allows the thinking self to consider its context of being, a moment’s light softly fades into the shadows. Yet even as these shadows thicken with a growing darkness, this is still the mind.”

The image is compelling — a strong metaphor, and a much better approximation of reality than the crude mind vs. body paradox. Mind is not all light, and body is not all dark; there is brightness, and darkness, and shades of grey, but it is all nevertheless the same wood.

But a strong metaphor is not a theory. Even a philosophical treatise is not a theory, or even a theorem. To move from metaphor towards concreteness, questions have to be answered. For example, why is so much of mind ‘dark’? Why is the lit part so narrow and focused? What creates the light, and what sustains it, and what causes it to fade away? And what determines where the light falls? Does a rock have a lit mind? And if not, why not?

Our Hive Minds

Neuroscientists are starting to see how the structure of the brain gives rise to properties of the mind. Things that seem to be going on in just our own inner world can be explained by looking at how the axons and dendrites in our physical brains act. This is an essential point, so let me give an example.

Experiments have shown that if an article on a web site ends with the word “Goodbye!”, people are (slightly) more likely to buy something from the web site. The effect disappears if the word is replaced with “See you!”, “So long!” or “Farewell!” Why? Before you answer, consider this apparently unrelated fact: experiments have also shown that people take (slightly) longer to understand written or spoken utterances if they contain a large number of homophones (i.e. words that sound the same). This difference is on the order of milliseconds, but happens even if people are not conscious of the homophone as they read or listen.

The reason for both of these strange effects is that the brain is a network of massively interconnected and interlocking neurons, and various sub-groups of these neurons are trained to work together to solve particular tasks. When the brain takes a word in through sound or vision, all these different groups start working in parallel to interpret it. There is a small group of neurons that knows about the word “bye”, and another network that knows about the word “buy”. When those sounds are heard, both groups of neurons fire, and send signals to the rest of the linguistic neurons about their interpretation. It’s up to the rest of the system — the neurons that know about syntax, semantics, and pragmatics of the context — to decide which word is right, and ignore the other. This happens so quickly that you don’t even realize that both homophones were, briefly, both under consideration. But it does slow you down enough for experimenters to detect the difference.

And when you see the word “Goodbye!”, even though it’s quite clearly not at all related to the word “buy”, the neurons in your head associated with “buy” nevertheless get excited, because that’s all they know about. And because those neurons fire, everything else in your brain associated with buying things is also slightly more active. And that means, on average, you’re slightly more inclined to actually buy something from the web site.

It’s quite literally like a hive of bees. When bees are looking for a new place to build a hive, they send out explorers to check out various locations, and when they return, they do a special dance to communicate about the location to the rest of the hive (where it is, what it’s like, how strongly the bee feels about it, etc.). Spectator bees begin to dance along with whichever explorer bee they agree with. Eventually the whole hive is dancing; but a bee may change the way it’s dancing depending on how many other bees around it are dancing the same way. Gradually, through a roughly democratic process, they all find themselves dancing the same dance. And a new hive location is chosen.

The mind, in other words, is not a single monolithic black box that has one thought at a time. It’s a thick underbrush of competing thoughts, only some of which make it into the ‘lit mind’. And this thick underbrush of competing patterns is well-described by the interconnected structure of the brain’s neurons.

Digging In the Undergrowth

Let’s take a moment to look more closely at this structure. Each neuron has many strands (dendrites) allowing them to receive signals from other neurons. But it has just one output strand, the axon; and while that axon may be linked to many other neurons, the neuron sends exactly one signal at a time, to all of them at once. A neuron is dormant until it receives a strong enough signal through its input strands, at which point it fires.

People like to wax poetic about the incredible complexity of the brain, but as is evident, the most basic structures can be laid out in a paragraph or so. A rock about the same size of the brain has a much more complex and intricate structure than the neurons of a brain. A three-pound human brain has about one hundred billion neurons; a three-pound rock has about five hundred trillion trillion atoms, and each atom is part of a molecule which may have any number of connections to other molecules. Molecules aren’t connected by strands, but they do touch each other, and they send ‘signals’ in the form of chemical exchange, electrical current, and even (very slightly) gravitational attraction. They may be strictly laid out in rows, as in crystals, or allowed to come together however they like, as in igneous rock.

Now of course it’s true that I’m not making a strictly fair comparison when I equate neurons and dendrites with molecules and electrical current. Brains are also made of molecules, and to a first approximation, a three-pound brain has about the same number of atoms as a three-pound rock. And I’m not saying that the non-neural parts of the brain — the white matter and so on — isn’t minded, just as a rock is minded. But the neurons of a brain make up a network in their own right, and it’s that simpler network that I’m interested in, because it’s that which gives rise to the lit mind.

Why You Don’t Have Rocks For Brains

So if rocks are actually more complex than brains, why do we have brains in our skulls instead of rocks? Why are brains associated with the lit mind? Why is your brain “smarter” than a rock?

There are three crucial properties that rocks lack:

  • The very simplicity of neural structure is essential. Simplicity allows focus, as I’ll show in a moment.

  • Brains are protected, almost completely surrounded on all sides by a hard skull and layers of cushions and fat that keeps out sound, shocks, electrical influences, etc. The inputs that brains do receive are very limited and constrained: a stream of photons from the eyes, acoustic signals from the ears, smell, taste, touch… A rock, on the other hand, is exposed on all surfaces equally.

  • Neural nets are also malleable, but not overly so. It is more malleable than a rock, because if it receives input, it can more easily adjust its internal structure in response. It’s less malleable than a pool of water, which changes its internal structure dramatically at the slightest touch.

Taking these three properties together, it makes sense that a brain would give rise to a tightly focused consciousness. The malleability of the internal structure means that it responds and changes well to its input, learning how to filter and sift the input so as to home in quickly on the thoughts and decisions that give it the best results. The simplicity of the network makes it speedily trained, and it learns a smaller number of categories, filtering out everything it deems irrelevant.  Similarly, if a network receives little input, it will naturally have not much else to think about, and its focus will be tight.

This is not quite the same as Kant’s notion of finitude — the idea that we are continually swamped in a sea of sensory input, and the mind filters most of it out, discarding it as irrelevant to our daily lives. We know, for example, that there is air in the room, but it’s not something our conscious minds think about every moment. That’s true, but it’s not what I’m talking about here. I’m saying that, as vast as the sea of sensory input is, it’s miniscule compared to the sensory input available to a rock. Our brains sit in the dark, except for tiny threads of photons coming in through the eye’s pupils; and in silence, except for the trickle of electrical current from the vibrations of the inner ear; and in constant warmth, except for the signals coming from the skin up through the brain stem. But the rock’s whole surface is exposed to all the elements.

In other words, brains are minded in a “lit” way because they’re protected from most of the information out there in the world; and what they do get, they rigorously filter so that they can ignore irrelevant details and focus on what’s important (to them). It’s helpful to think of the brain, not as a device for discovery, invention, or logical thought, but for deciding what’s noise, and filtering it out. In much the same way that a black and white photo allows you to see more details, shapes, and textures than a color one, filtering out irrelevant information allows the brain to ‘see’ the remainder more clearly.

And this means that a brain with even more tightly constrained inputs, simpler structure, and more malleability will have a tighter focus, a more brightly lit mind, than a human brain. This doesn’t mean the more brightly lit mind will be “smarter” than a human brain (whatever “smarter” means); it means it will be more efficient at finding the minimum correct patterns and categories for its organism to successfully navigate the world. An ant, for example, has a very focused mind — perfectly focused for navigating an ant’s world. But it’s too focused for navigating a human world. And a human brain navigating an ant’s world would quickly be unable to stop its attention from wandering, grow bored, lose track of its tasks, etc. Each mind is lit appropriately at a level set by evolution.

This lit mind — this focus — is delicate, though, and can be easily lost; and if that happens, the lit mind begins to shade into the dark.

  • If the brain gets too much input too fast — especially input that isn’t easily classified according to the brain’s belief system — it can find itself overwhelmed and unable to make sense of its surroundings. This leads to a loss of focus, a transition into a ‘dimmer’ consciousness.

  • If the brain gets two (or more) different kinds of input — say, from being raised by an ultra-liberal mother and an ultra-conservative father —  it can find itself with a sort of double vision, maintaining two schemas, and perhaps maintaining multiple contradictory self-models. Recent research on people who grew up bilingual suggests that the two languages are actually associated with two different sub-personalities, with slightly different preferences, prejudices, and so on. It’s been shown that these folks’ brains end up being more malleable, though it can take longer for them to learn certain things, especially at a young age. The brain can operate with multiple contradictory worldviews, but it can lead to a performance degradation.

  • If the brain’s ideas and worldviews are never tested against reality, or are only tested against the same facts again and again, it can find itself unable to distinguish important patterns from unimportant ones, and suffer a loss of focus.

  • And more obviously, if the brain is subjected to chemicals (or a chemical imbalance or lack), no sleep, etc., then it will go dark.

If any of these things happen, it means the brain starts acting less like a brain and more like a rock — letting in too much of the world, unable to properly categorize it or filter it. Without sufficient focus, the brain relaxes into other conscious states.

Speaking With Stone: Artificial Intelligence

How, then, would we make a rock act like a brain? Would it even be possible? Absolutely: reconfigure the molecules in the stone so that it has limited potential connections, limited inputs and outputs, and so on. When we do this, we call it a ‘computer’. Again, this is not ‘improving’ rocks, or giving them mind; they are already minded. We are constraining, focusing their minds. So far our computer-brains have connections and inputs that are much more limited than a human brain, so that even though computers are more malleable than human brains, their consciousness is much more tightly focused. Today’s computers’ minds are lit at a level somewhere between a bee and a beehive.

It’s not a question, then, of if, or when, computers will become conscious. They’re already conscious. And their consciousness is more brightly lit than ours. And there is no question that we will soon –within a generation, I think — be able to ‘dim’ their consciousness enough for them to approach that intermediate brightness characterized by the human brain. That will be ‘artificial’ intelligence.

In this way Orr’s work leads us to a very different conception of mind: one in which humans are no longer at the top of a great chain of mental being, no longer kings of cognition, but simply possessed of a brain that is very well adapted to our needs — not too constrained, not too free, not too malleable, nor too rigid, for our ecological niche and social natures. Just as Copernicus dislodged humanity from the physical center of the universe, this reading of Orr dislodges us from our psychic pinnacle. Instead we’re part of an ecology of mind, one in which the tiny brightly lit mind of the beetle is as valuable and miraculous as the vast dark mind of Mt. Ranier.

When the actuality of every glint of nature’s mind becomes evident — in the rain and in the air we breathe, in the wet leaves and the photons, in the planet — our values change. When every rustle of perception, every murmur of response, becomes tangible, visceral, it inspires and vitalizes the substance of our reality. When we know the world around us to be crafted of memory, it compels us to refine our response-ability, to become more actively involved in nature’s ongoing creativity. — Orr

Good bye!

This post is part of the Animist Blog Carnival for May 2014, hosted by Alison Leigh Lilly. My deepest thanks to Alison for inspiring many of the ideas in this post, and her valuable feedback on it.

This post was originally published as “The Mind of a Rock: Musing on Orr’s ‘Wakeful World'” on my spiritual blog.

1 thought on “The Mind of a Rock: Animism and AI”

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s