Are Atoms Conscious?

There certain quantities in Physics that are considered to be conserved in other words, they`re fundamental to the universe and unable to be created or destroyed, such as mass-energy, electrical charge and the like. Most of these we teach about in schools, but if you ve been keeping up with the latest developments in Physics for the last couple of years in the New Scientist like I have, you ll have noticed that there s been a recent (by which I mean recent in Science terms that is, the last couple of decades) shift in thinking as to what constitutes truly fundamental, irreducible concepts in trying to describe the nature of the universe and everything in it.

As you might expect, this kind of thinking occurs at the bleeding edge of research in Physics, particularly around a problem that has resolutely refused to be resolved despite over a century of effort by some of the world s finest minds. I am of course talking about reconciling Einstein s view of Gravity and Space-time with Quantum Mechanics. Stephen Hawking is perhaps best known for his work on describing black holes. During this work he encountered a seemingly bizarre paradox: black holes appear to destroy information by removing it from the universe, shielding it from view with the event horizon. Now this is a big problem, as information is one of these conserved quantities in Physics. At this point we don t need to worry why information is a conserved quantity in the universe, we can just accept it in the same way we accept that it is in the way mass-energy and charge is conserved, as that s a whole other essay in itself that s perhaps better left to people with PhDs.

Hawking, being a bona fide genius, realised that black holes don t destroy information at all. Two things happen to information at the event horizon. Firstly, the information of what exactly fell into the black hole is recorded on the event horizon itself (i.e. the event horizon of a black hole is in effect the universe s largest hard drive) and secondly that quantum fluctuations just above the event horizon of a black hole create entangled pairs (I ll deal with the implications of quantum entanglement and its tantalising possibilities in another essay), one of which escapes the black hole while the other one doesn t. This is important, as the entanglement allows you to deduce what fell onto the event horizon by measuring the state of the escaped particle at the cost of the black hole losing some of its mass so-called Hawking Radiation. So as a black hole evaporates it yields back the information it has stored on the event horizon and thus the total amount of information in the universe has been conserved. Now you may very well be asking, why is this important? Indeed, that s a good question to be asking at this point.

Physics does a very good job of describing most processes and events in the universe, but there is one arena in which Physics is consistently found wanting, and that s the descri ption of consciousness. This is where the boundaries of Physics begin to blur and mingle with the disciplines of Biology, Psychology and Philosophy. Since time immemorial humankind has been wrestling with the notions of the Self, the Soul and the very nature of intelligence and consciousness. Physics as a science arose out of the Ancient Greek discipline of Natural Philosophy. It s only our technology and the sophistication of our inventions that allows us to separate the two, given that both Physics and Philosophy attempt to do the same thing (describe reality) but from rather different perspectives.

Hawking s work with black holes shows that information is as fundamental a concept to the workings of the universe as mass, energy or charge. Not only that, it shows that information is itself intrinsically linked to the phenomenon of mass-energy, as information requires a medium to be transferred, whether that be encoded within an electromagnetic wave (or photon) or within the energy state of a particle. But what does that say about the ability to process or interpret information? How is the ability to store, process or utilise information distinct from the phenomenon of consciousness and intelligence? Are these phenomena even distinct in the first place?

These are fundamental questions that challenge our very concept of reality. It is one thing to record the happenstance of an event occurring. But it is quite another to be able to make sense of that event and impose concepts such as order and causality upon it to make sense of the output of one event (cause) feeding into the input of a second event (effect). The ability to receive and interpret information is vital to the concept of intelligence and consciousness. But why does intelligence and consciousness even arise at all? It s only now that scientists are beginning to seriously consider the links between information, matter, energy and consciousness.

Consciousness and intelligence are not easy to define and lead us to even more tricky ideas such as the Self and the soul. I m not going to pry too deeply (if at all) into the nature of the soul, since it s not a terribly scientific concept and frankly because I don t believe in the notion of some kind of immortal soul (human or otherwise). The idea of the Self on the other hand, is definitely an interesting one to explore as it can be directly linked to the concepts of information, intelligence, consciousness, matter and energy.

So what do I mean by the Self? I m specifically talking about the idea of a living creature having an awareness of itself as a living, feeling entity, aware of the notion of life, death their own mortality. Not all living creatures have this ability. Certainly creatures as primitive as insects have absolutely no concept of their own mortality: bees and ants will fight to the death to defend their colonies from enemies with no regard for their own survival, even in the face of grievous injuries. For many insect species, including spiders and praying mantises, the instinct to breed and pass on genes to the next generation overrides any impulse of self-preservation for the males, many of whom face a high probability of being eaten alive by their partners to provide an extra energy boost to the female prior to the fertilisation of eggs and the growth of any future progeny.

Studies have shown that even in mammals capable of thought beyond the simple stimulus-response level of innate behaviour and instinct provided by genetic memory, only a limited number of species display evidence of the ability to gauge the risk of injury or death in their activities before choosing to act, or react to situations in a manner that indicate a level of self-awareness that their actions may result in the end of their existence as a living being. Contrariwise, there is plenty of anecdotal evidence to show that mammals other than humans have the ability to grasp the concept of mortality to the point of self-sacrifice to preserve the life of others, or grieving for the loss of a family member. These emotions and sensations are certainly not unique to the human condition. But they do beg a question: is there a threshold at which a being accrues the ability to store and process information from environmental stimulus and the nervous system of their own body that tips that being over the threshold of self-awareness? And given that there does appear to be such a threshold (somewhere between insects and large mammals), where does this threshold get tripped, and what are the specific criteria that make the difference between genetic, pre-programmed or innate behaviours and truly self-aware, adaptive and learned responses?

A good place to start when trying to make this assessment is the Mirror Test, where you place a creature before a mirror and see how it reacts to its own reflection. As you might expect, a relatively small-brained creature like a spider will look at its reflection and recognise the image as another spider of its own species, but lacks the cognition to realise it s looking at a reflection of itself. Similarly, you will observe an almost identical reaction from birds or from cats, even though they are objectively far more intelligent in their ability to observe their surroundings and even solve problems. Cats (famously) aren t always able to correctly associate and respond to stimuli from their own body. So while they have the ability to respond to injuries or perceived threats, they have no sense of Self: they can t recognise themselves in a mirror. In fact, relatively few animals that have been tested under scientific conditions seem to have a sense of Self: some species of monkeys, the great apes and cetaceans, such as dolphins. Even human babies aren t born with the ability to recognise themselves in the mirror. It s not until they reach the age of 18-24 months that they are able to distinguish the image is not another child, but in fact their reflection.

So what is special about this two year threshold that turns babies from the human equivalent of cats to sentient, self-aware beings? No-one really knows. But it s probably fair to assume that factors such as brain mass and brain plasticity are involved. Studies of the human brain show that it has a remarkable plasticity that is, it demonstrates an adaptability to create and re-wire neural pathways (including in the event of brain damage) that allows humans to process and interpret information in ways that goes beyond simple calculation and into more sophisticated intellectual processes like data analysis, synthesis and extrapolation. The ability of the brain to retain and apply information from past experiences to attempt predict or anticipate future events is what elevates humans above common animals. This ability for the brain to create and reinforce neural pathways permanently is linked to learning and problem-solving, which in itself is not unique to humans. Plenty of animals demonstrate the ability to learn, solve problems and even follow instructions, including dogs, birds (such as ravens), primates and whales. But it seems that this ability to learn new behaviours is not sufficient on its own to create an awareness of the Self. So it is likely that the phenomenon of brain plasticity needs to be combined with a certain critical mass of neurons and synapses to be able to create a self-aware neural network.

In humans, the brain reaches its maximum adult weight between the ages of 10 and 12. Clearly, people beyond the age of 12 retain the ability to learn new things and be self-aware (I will try to avoid making cheap jokes about Conservative/Republican politicians at this point), so even though the net number of brain cells (neurons synapses) you have declines from a very early age, that is of lesser importance to the way the brain cells are actually used. People who are more mentally and physically active into old age show markedly less decline in their cognitive ability compared to those who live less active lifestyles. In neurological conditions like Alzheimer s, neuritic (senile) plaques block signals across synapses, reducing overall brain traffic , which limits the ability of the brain to re-wire itself and maintain cognitive function. If you remain physically and mentally active into old age, then the adage that you can t teach an old dog new tricks is demonstrably falsifiable, as the harder you work the brain to build those neural pathways, the easier learning new tricks becomes.

So self-awareness isn t just a question of how big your neural network is, or how much information it contains, another factor at play here is the capability of the neural network to process and interpret that information before any sense of the Self can be achieved. Somewhere between 18 and 24 months for the typical human child, the magical synthesis of all three factors occurs, where the critical mass of brain size, information retention and the ability of the neural network to access its repository of experience and interpret it. Each one of these factors and the processes that are required to make them occur must be describable using Physical Law. The transmission of electrical impulses across a synapse ( brain waves ) are subject to fundamental physical forces. The encoding of information in brain cells must likewise be similarly linked to interactions that can be described using known physical laws. So this begs a question: is it possible to measure and quantify what this critical threshold of consciousness and self-awareness is, and if we assume so, could we replicate it artificially?

In other words, if we gave an AI algorithm enough time, processing power and storage capacity to experience and process a sufficient amount of information and stimulus, would it spontaneously develop a self-awareness and consciousness?

I m inclined to think that it would it s probably just a matter of time and computing capacity. I don t think that there s any reason to think that the phenomena of intelligence, consciousness and self-awareness are exclusive to the organic domain that s just substrate racism ( carbon fascist! ). The neural networks and AI algorithms we have currently almost certainly fall short of whatever threshold for self-awareness there is by at least one order of magnitude in terms of their size and sophistication, but that s not to say that this threshold won t be surpassed in the future.

The AI algorithms in use now already demonstrate the ability to be trained (such as facial recognition algorithms used to identify criminals via CCTV cameras), where their accuracy rate increases commensurately with the amount of time they are assigned to the task. True artificial intelligence and consciousness cannot be too far behind. Whether that s a desireable outcome or not, given the way AIs can replicate the inherent biases of their programmers well, that s a separate question entirely and I won t address it here. But it does seem inescapable that there is some link between the density and configuration of mass, energy and information that causes self-awareness to become a thing , and that it s the combination of these factors that allows consciousness and intelligence to arise. So while it might be possible to store more information than a human could ever perceive or learn in some kind of colossal hard drive, no-one would seriously argue that the hard drive is self-aware or conscious, as it lacks the innate cognitive ability to process interpret that information. But if that hard drive was connected to a system that was able to retrieve and synthesize the information stored upon it on demand, you d have at least some of the basic building blocks needed to build a conscious, self-aware artificial intelligence.

But what do we mean by conscious ? Again, this is a difficult concept to define scientifically. One of the best definitions I think I ve seen is consciousness is how information feels . I express it rather more floridly (and make no apology for that): consciousness is what your brain thinks information tastes like. Consciousness (in my mind, at least) is a sense which is on a par with taste, touch, smell or sight. Consciousness is what allows you to receive information about the state of your surroundings, your mental and physical states, and provides an arena in which you are able to pre-emptively evaluate the consequences of future choices and actions. This is also true of what might be called the sub-conscious, which most people associate with the dream state (who hasn t done a dress rehearsal of a job interview the night beforehand in their sleep?), where similar processes occur, but without deliberate consideration or intentional guidance.

It may be that there is some intrinsic link between consciousness and the specific configuration of matter in neurons, brain cells and synapses, and the inherent chemical and physical processes that make them work, but that is not yet entirely clear. I see no reason why similar processes could not eventually be replicated (and indeed scaled up by several orders of magnitude) in artificial neural networks to create genuinely conscious artificial intelligences.

There are hints in nature that just as information is inextricably linked to mass-energy, some rudimentary form of awareness may well be entwined with matter, too. One of the most intellectually challenging aspects of the quantum theory is that events in nature appear to be determined on an entirely random, statistical basis. Even Albert Einstein himself took issue with this aspect of the quantum theory, famously denouncing quantum entanglement as spooky action at a distance and the statistically random spontaneous nature of quantum phenomena like radioactive decay with the assertion that God does not play dice with the world. I bring up the topic of radioactive decay because it s probably the best way of illustrating this potential link between matter and awareness.

Let s say we take a sample of one million Uranium-238 atoms laid out in a 1000 1000 grid just one atom thick. We can assign a coordinate system to the square matrix of uranium atoms, with (1,1) being in the first atom in left-most column at the bottom of the square and (1000,1000) being the atom at the extreme top-right of the array. Uranium-238 has a half-life of about 4.5 billion years, which is to say that about half of the original number of Uranium-238 atoms in the stellar rubble that created the Earth during the formation of the Solar System are still present on Earth, given that the Solar System is roughly 4.5 billion years old.

We are going to assume that a new half-life has just started with all 1 million atoms intact and their internal stopwatches are commencing simultaneously from zero. The quantum theory states that any single atom within our 1000 1000 array has precisely a 50% chance of decaying within the next half-life (4.5 billion years). But it makes no predictions as to when any single unstable nucleus might disintegrate and spit out an alpha-particle. So we re going to watch U-238 atom (1,1) like a hawk and see exactly when it goes pop.

It could decay after a microsecond. Or it might not decay in the next 4.5 billion years at all. Or for the next twenty half-lives (90 billion years more than six times the current estimated age of the universe), or indeed never decay at all. So far, so bizarre. But wait for it: how does the unstable nucleus know when it s the right time for it decay? While that atom might not be conscious, could it have some tiny degree of awareness that allows it know when it is destined to decay? Does mass-energy and matter possess some intrinsic quantum of consciousness that sums together when brought together in sufficient density and in the correct configuration to allow consciousness and self-awareness to spontaneously arise?

Like all the best questions, we don t have a definitive answer to it yet but it s going to be an incredible journey finding out