The Hard Problem of Consciousness

There are five levels — and we are at Level 4

The Problem

There is the brain, with its trillions of interconnecting neurons, with neural impulses travelling back and forth, sometimes reinforcing a pathway, and sometimes inhibiting. Coupled with this is a chemical system, interlinked with these pathways at the level of the individual synapses, but also affecting the brain as a whole, with the ratio of one chemical (or neurotransmitter) to another potentially affecting our moods.

Then there is our experience — often referred to as qualia — of various sensations, ranging from the physical (the pain of a hot stove) to the almost metaphysical (the awe when you watch the sun rising behind a mountain).

And on top of it all, there is the feeling of a separate self, an “I” behind the eyes — a person that seems separate and distinct from the body, looking out and sensing the world through the imperfect intermediaries that are the sensory organs. This “I” goes by various names: Self-awareness, sentience, consciousness, and of course, soul.

How does the physical structure and electrical/chemical activity of the brain create the experiences, the qualia? Why is our brain structured to create the illusion of a separate self, a separate I? What evolutionary advantages did it confer, and how does the brain create this illusion?

The first question has been called the hard problem of consciousness, but I suggest that the key to a true understanding of consciousness requires that we also focus on the second and third questions.

A Working Theory

Now my day job is that of a construction project director, but I do write science fiction on the side, and my latest project, my magnum opus, as it were, requires that these questions be answered, because the whole basis of the story — which starts off with an AI becoming sentient at the end of the first book, and goes forward from there in the second and third books, requires a working theory that sounds plausible, at least from a SF writer’s perspective.

So here is my working theory: There are actually five types of consciousnesses, arranged hierarchically in terms of increasing complexity, and we are at Level 4.

Level 1

Let’s start by imagining a bacterium floating around in a soupy substrate. The interactions of this bacterium with its surroundings are basic and simple. Find food, escape from potential dangers, and procreate like crazy. Though the bacterium is a living organism, it does not strain credibility to imagine that all these functions are catered for by simple feedback mechanisms.

In that sense, a bacterium is no different from a thermostat, which is designed to stay within certain set parameters, except that the bacterium is alive and capable of self-replication, so it has the advantage of evolution on its side. It is possible, though there is room for debate, that most (if not all) plants also fall into the same category.

Level 2

This simple thermostat approach would not work with multi-cellular and complex animals, because the feedback loops are now much more complicated.

Think for a moment of a Komodo dragon feasting on some animal that it has killed. Unfortunately, the carcass is lying in the sun, and the dragon has no means of internally controlling its body temperature, so it periodically needs to go back into a shaded area, and then come back to eat once its inner temperature has gone down to a non-lethal level.

This balancing act now requires a more elaborate feedback mechanism, with a brain, no matter how rudimentary, which creates both a sense (i.e. qualia) of discomfort (sun, too hot, painful, seek out the shade) as well as comfort (food, good to eat, pleasant, go for it). Of course, we do not know the exact mechanisms by which a hot sun creates the sensation perceived as pain, or how eating generates a sense of pleasure in the reptile’s brain, but at least we only have to explain how these physical sensations arise.

Level 3

However, in addition to these purely physical sensations, birds and mammals have a brain structure (the limbic system) that permits additional, more subtle sensations; there is now an additional type of qualia, an emotion, that is generated in response to the outside world, and which is qualitatively different from a purely physical sensation (such as pain).

Let me give you an example. We have a pet African Grey parrot in our home, and during the daytime, she generally has free run of the house, only returning to her cage when she is really hungry. For whatever reason, she is very attached to me, and as my work takes me away from home for days and sometimes weeks, her reaction on seeing me after a few days is obviously joyful.

If I want to leave the house in the first few hours after she’s seen me, I have to do it quietly, because otherwise she will come and sit on my head, as if trying to stop me from leaving. No one can tell me that she is not feeling emotions — joy, anxiety, worry, happiness — that go beyond the physical.

Anyone who has kept both a pet iguana and a dog will immediately be able to tell the difference between an animal that feels on the purely physical plane, and an animal that additionally also feels emotions. This is why a pet lioness you raised from birth will never see you as food, but a pet crocodile will; the latter feels no emotions.

Any species that developed an ability to feel emotions acquired a huge evolutionary dividend, because it allowed members of the same species to coordinate and live flexibly in groups, unlike (say) bees, where the group instinct had to be baked into the DNA. With emotions, you now have a glue that bonds the individual members of the group, by creating positive emotions for each other, and also (unfortunately) by creating negative emotions for those who are non-members.

Emotions also enhanced the care that the young of the species received, further increasing the evolutionary advantage.

The First Three Levels

So, we can now theorize that (so far) there are three kinds of consciousnesses: Automatons which have practically zero consciousness, a second level where the organism feels physical stimulation only, and a third level of consciousness, where emotions are introduced.

The introduction of emotions complicates the hard problem of consciousness even further. It is already difficult enough to explain how touching a hot stove generates the qualia of pain in my hand; now, on top of this, we have to explain how seeing my wife generates a feeling that I call love.

Maybe one day we will understand how the specific combination of hormones (such as oxytocin) and neural impulses, when these interact with my memories, generate this feeling; but even then, it will not tell us how we get the jump from those specific brain states to the experience and feeling of love.

The Problem of Self

Let’s now move into more speculative territory. As human beings, we have an additional quality — the feeling of a separate self, which I will refer to as self-awareness rather than the soul, in order to avoid making value judgments. One can argue as to whether other species have this sense of self or not; but by its very nature, this feeling must be binary — either you have it, or you don’t, for it is difficult to imagine having a partial sense of self.

So, assuming for a moment that this is restricted to humans only, when did it appear in our evolutionary line? If we could discover the physical structures within the brain that create this self-awareness, identify the specific genes responsible for it, and then (somehow) trace back this evolutionary history using our current genome as a starting point, we may be able to answer this question one day.

For now, all we can be reasonably confident of is that Neanderthals — from the fact that they buried their dead with flowers — had a similar sense of self, for reasons I will explain further below. As the branching between the Neanderthals and modern humans occurred around 800,000 years ago, and the Denisovans branched out from the Neanderthals after this, we can surmise that Denisovans were also sentient beings, and that self-awareness is at least a million years old.

Level 4 — Past and Future

I would like to make two points here: This sense of self makes our consciousness qualitatively different from that of other animals, to the point where it can be called Level 4; and that it is actually a by-product of something which is more fundamental — the ability to live, inside our brains, in the past as well as the future.

According to this viewpoint, other animals only live in that fleeting moment that we call the present — like a frame in a movie, where the frame rate is the inverse (perhaps) of the Planck time (the shortest possible period of time).

Sure, my parrot feels love, but only because her specific internal brain chemistry and neural pathways interact with her memories to generate that feeling; in this scenario, if I were to ill-treat her, the memory interaction when I meet her the next time would now generate negative feelings/emotions, but she would still be living in the present only.

To understand better what living in the present means, find a video of a lioness stalking and then hunting down its prey, let’s say a herd of wildebeest, on the plains of Serengeti. Once the lioness starts its attack, the herd scatters in a high state of panic, fear etched into the body movements of the individual wildebeest. But as soon as one member of the herd has been brought down by the lioness, and the attack is over, the other members go back to grazing on the grass, as if nothing has happened.

Can you imagine a group of human beings being so nonchalant after one of them has been brought down by a lioness? The past would still be resonating in their brains, as if it was still happening, again and again. But wildebeests live only in the present, so once the present threat is gone, the past disappears from their attention; only the vestiges of the past remain, as retained in their memory, which in turn will affect their future behavior when their memories come back into sync with the outside world, i.e. when the lioness attacks again.

Imagining the Future

So, what is unique about this ability to live in the past and the future? And how is it generated (qualia again, but at a much deeper level)? Let’s try to answer the first question by rephrasing it in a different way: What evolutionary advantage did this confer on our ancestors?

A credible answer, if you think about it, is that it allowed them to learn not only from experience, whether their own or by learning from others, it allowed them to create hypothetical scenarios in their minds, and then plan for it. In other words, it allowed our ancestors to simulate the future.

For example, before leaving his cave to kill a deer, a hunter could relive past experiences in his mind, and based on that knowledge, then plan out a whole future scenario in his mind; for example, approach the deer from this angle to get the best shot, but stay close to the forest line, in case a lioness is nearby, so that he had an escape route.

While not perfect, it was far better than learning only from experience as retained in memories, because experience is costly in a way that mental simulations are not — you can get killed many times by a lioness in a mental simulation, but only once in real life.

Now one could argue that other animals also exhibit this kind of planning behavior, but my argument is that this planning behavior is based on memories of past experiences modulating and affecting present actions, thereby creating the illusion of planning for the future. The animal is still living only in the present.

By being able to simulate the future in the brain, we also acquired the ability to imagine a future that is independent of past experience — we could not just imagine a lioness in the future, we could also imagine a centaur.

And though it took hundreds of thousands of years before the full force of this ability bore fruit, it has allowed us as a species to create what is not there, simply by imagining a future, and the possibilities that lie out there. In fact, we can even imagine multiple alternative pasts that never occurred, and the futures that these pasts could have generated, all existing only within our brains.

Emergence of Self

In any event, just think about what happened once our ancestors developed this simulating ability; using the example above, the imagined future does not just have a lioness and a deer, it also has the hunter, a future self, walking close to the forest line.

The hunter had to now visualize himself in that future, as a third actor on the stage that he was creating in his brain. This could only work if he also developed a sense of self, an “I” behind his eyes, someone who is him and yet not him, because it is a future him.

Emergence of Death

It then follows, quite naturally if you think about it, that a future in which the hunter could visualize killing the deer also had an alternative possible ending, where the lioness killed him; so, the ability to live simultaneously in the past as well as the future meant that our ancestors also had to face the possibility that their existence would end in some future moment of time.

Other animals lived in the present, so they had no fear of death; but now the poor hunter had to contend with the possibility that one day, he would cease to exist.

Any evolutionary advantage conferred by living in the past and future would therefore have been negated by the fact that the hunter would now be really, really scared to leave his cave, and therefore starve to death, just as a wildebeest herd that developed the ability to live in the past and the future would constantly be worrying about lionesses, rather than eating grass and procreating.

For our ancestors, the workaround for this problem was simple: Come to the firmly-believed conclusion, rightly or wrongly, that their separate self, their soul, would live on after death, so that death was merely a transition point to another phase of life. By this reasoning, therefore, a species that carefully buries its dead can be safely considered to be sentient.

Connection to the Quantum

How did this ability to live in the past and present develop? How does the brain go about doing this? To borrow an idea from Roger Penrose, let’s assume that there are specific structures within the human brain which manage to stay in quantum superposition for extended periods of time, and that this connection to the quantum realm (for lack of a better phrase) is the basis of consciousness.

Let me now extend this assumption in a manner that Roger Penrose never intended (I think).

In this extension, we posit that the brain structures that allow this connection also permit us, as a by-product, to escape the small sliver of time that is the present. Now you may ask why would that be the case, i.e. what is so special about this connectivity? How exactly does it allow our brains to escape the present moment?

I have so far gone out on a (really extended) limb; please bear with me, as now I have to take you into woo-woo territory.

Consciousness

Imagine that there is a consciousness field that permeates the whole universe, in the same way that there is an electron field. And just as two electrons exist as two separate waves in the one electron field, you and I are separate waves in that consciousness field, with these special brain structures (whatever they are) being responsible for “generating” the wave.

Imagine further that this field exists in dimensions beyond our three spatial and one temporal — say in at least six dimensions. In six dimensions, all past and future possibilities (i.e. quantum states which have not collapsed) exist, so this connection to the consciousness field allows our brain, no matter how imperfectly, to “live” in the past, present, and future, all at the same time.

This connection also gives us the sense of self, because we are borrowing, in a manner of speaking, from the consciousness inherent to that field, just as the properties of the electron wave are defined by its field.

The Ocean of Consciousness

This is not an original idea. Ancient Vedic philosophers, in treatises such as the collection of books known as the Upanishads, posited the existence of a universal consciousness, which was likened to a sea. Individuals appear as separate waves in that sea, feeling different from each other simply because each person sees his wave as all there is, when in fact the waves are all part of one ocean.

As very preliminary evidence for this, consider that people taking various hallucinogens (think mushrooms) report that when they are deep within the psychedelic experience, the separation between the self and the universe disappears, with the self realizing that it is part of a much bigger whole.

You can dismiss this as a straightforward hallucination, if you like, but then consider that people who have been meditating deeply for years also approach the same end point. If correct, then this is potentially a fifth level of consciousness, where self-awareness becomes universal, and connected to everything else.

Can an AI Become Conscious?

If the above analysis is correct, then any AI which is software-based can never become sentient or self-aware, because the interaction with the consciousness field can only be mediated through physical structures; so, if we wanted to deliberately create an AI which can become sentient, we would have to construct it with such structures built into its hardware.

If we truly set out to do this, then we will have to make sure that we first instill emotions (i.e. Level 3 consciousness) within this AI, because otherwise all we would be doing is creating a super-intelligent sentient being which has no emotions, which is unlikely to end well for us.