Homo Deus

by Yuval Noah Harari

  What exactly are the conscious experiences that constitute the flow of the mind? Every subjective experience has two fundamental characteristics: sensation and desire. Robots and computers have no consciousness because despite their myriad abilities they feel nothing and crave nothing. A robot may have an energy sensor that signals to its central processing unit when the battery is about to run out. The robot may then move towards an electrical socket, plug itself in and recharge its battery. However, throughout this process the robot doesn’t experience anything. In contrast, a human being depleted of energy feels hunger and craves to stop this unpleasant sensation. That’s why we say that humans are conscious beings and robots aren’t, and why it is a crime to make people work until they collapse from hunger and exhaustion, whereas making robots work until their batteries run out carries no moral opprobrium.

  And what about animals? Are they conscious? Do they have subjective experiences? Is it okay to force a horse to work until he collapses from exhaustion? As noted earlier, the life sciences currently argue that all mammals and birds, and at least some reptiles and fish, have sensations and emotions. However, the most up-to-date theories also maintain that sensations and emotions are biochemical data-processing algorithms. Since we know that robots and computers process data without having any subjective experiences, maybe it works the same with animals? Indeed, we know that even in humans many sensory and emotional brain circuits can process data and initiate actions completely unconsciously. So perhaps behind all the sensations and emotions we ascribe to animals – hunger, fear, love and loyalty – lurk only unconscious algorithms rather than subjective experiences?2

  This theory was upheld by the father of modern philosophy, René Descartes. In the seventeenth century Descartes maintained that only humans feel and crave, whereas all other animals are mindless automata, akin to a robot or a vending machine. When a man kicks a dog, the dog experiences nothing. The dog flinches and howls automatically, just like a humming vending machine that makes a cup of coffee without feeling or wanting anything.

  This theory was widely accepted in Descartes’ day. Seventeenth-century doctors and scholars dissected live dogs and observed the working of their internal organs, without either anaesthetics or scruples. They didn’t see anything wrong with that, just as we don’t see anything wrong in opening the lid of a vending machine and observing its gears and conveyors. In the early twenty-first century there are still plenty of people who argue that animals have no consciousness, or at most, that they have a very different and inferior type of consciousness.

  In order to decide whether animals have conscious minds similar to our own, we must first get a better understanding of how minds function, and what role they play. These are extremely difficult questions, but it is worthwhile to devote some time to them, because the mind will be the hero of several subsequent chapters. We won’t be able to grasp the full implications of novel technologies such as artificial intelligence if we don’t know what minds are. Hence let’s leave aside for a moment the particular question of animal minds, and examine what science knows about minds and consciousness in general. We will focus on examples taken from the study of human consciousness – which is more accessible to us – and later on return to animals and ask whether what’s true of humans is also true of our furry and feathery cousins.

  To be frank, science knows surprisingly little about mind and consciousness. Current orthodoxy holds that consciousness is created by electrochemical reactions in the brain, and that mental experiences fulfil some essential data-processing function.3 However, nobody has any idea how a congeries of biochemical reactions and electrical currents in the brain creates the subjective experience of pain, anger or love. Perhaps we will have a solid explanation in ten or fifty years. But as of 2016, we have no such explanation, and we had better be clear about that.

  Using fMRI scans, implanted electrodes and other sophisticated gadgets, scientists have certainly identified correlations and even causal links between electrical currents in the brain and various subjective experiences. Just by looking at brain activity, scientists can know whether you are awake, dreaming or in deep sleep. They can briefly flash an image in front of your eyes, just at the threshold of conscious perception, and determine (without asking you) whether you have become aware of the image or not. They have even managed to link individual brain neurons with specific mental content, discovering for example a ‘Bill Clinton’ neuron and a ‘Homer Simpson’ neuron. When the ‘Bill Clinton’ neuron is on, the person is thinking of the forty-second president of the USA; show the person an image of Homer Simpson, and the eponymous neuron is bound to ignite.

  More broadly, scientists know that if an electric storm arises in a given brain area, you probably feel angry. If this storm subsides and a different area lights up – you are experiencing love. Indeed, scientists can even induce feelings of anger or love by electrically stimulating the right neurons. But how on earth does the movement of electrons from one place to the other translate into a subjective image of Bill Clinton, or a subjective feeling of anger or love?

  The most common explanation points out that the brain is a highly complex system, with more than 80 billion neurons connected into numerous intricate webs. When billions of neurons send billions of electric signals back and forth, subjective experiences emerge. Even though the sending and receiving of each electric signal is a simple biochemical phenomenon, the interaction among all these signals creates something far more complex – the stream of consciousness. We observe the same dynamic in many other fields. The movement of a single car is a simple action, but when millions of cars move and interact simultaneously, traffic jams emerge. The buying and selling of a single share is simple enough, but when millions of traders buy and sell millions of shares it can lead to economic crises that dumbfound even the experts.

  Yet this explanation explains nothing. It merely affirms that the problem is very complicated. It does not offer any insight into how one kind of phenomenon (billions of electric signals moving from here to there) creates a very different kind of phenomenon (subjective experiences of anger or love). The analogy to other complex processes such as traffic jams and economic crises is flawed. What creates a traffic jam? If you follow a single car, you will never understand it. The jam results from the interactions among many cars. Car A influences the movement of car B, which blocks the path of car C, and so on. Yet if you map the movements of all the relevant cars, and how each impacts the other, you will get a complete account of the traffic jam. It would be pointless to ask, ‘But how do all these movements create the traffic jam?’ For ‘traffic jam’ is simply the abstract term we humans decided to use for this particular collection of events.

  In contrast, ‘anger’ isn’t an abstract term we have decided to use as a shorthand for billions of electric brain signals. Anger is an extremely concrete experience which people were familiar with long before they knew anything about electricity. When I say, ‘I am angry!’ I am pointing to a very tangible feeling. If you describe how a chemical reaction in a neuron results in an electric signal, and how billions of similar reactions result in billions of additional signals, it is still worthwhile to ask, ‘But how do these billions of events come together to create my concrete feeling of anger?’

  When thousands of cars slowly edge their way through London, we call that a traffic jam, but it doesn’t create some great Londonian consciousness that hovers high above Piccadilly and says to itself, ‘Blimey, I feel jammed!’ When millions of people sell billions of shares, we call that an economic crisis, but no great Wall Street spirit grumbles, ‘Shit, I feel I am in crisis.’ When trillions of water molecules coalesce in the sky we call that a cloud, but no cloud consciousness emerges to announce, ‘I feel rainy.’ How is it, then, that when billions of electric signals move around in my brain, a mind emerges that feels ‘I am furious!’? As of 2016, we have absolutely no idea.

  Hence if this discussion has left you confused and perplexed, you are in
very good company. The best scientists too are a long way from deciphering the enigma of mind and consciousness. One of the wonderful things about science is that when scientists don’t know something, they can try out all kinds of theories and conjunctures, but in the end they can just admit their ignorance.

  The Equation of Life

  Scientists don’t know how a collection of electric brain signals creates subjective experiences. Even more crucially, they don’t know what could be the evolutionary benefit of such a phenomenon. It is the greatest lacuna in our understanding of life. Humans have feet, because for millions of generations feet enabled our ancestors to chase rabbits and escape lions. Humans have eyes, because for countless millennia eyes enabled our forebears to see whither the rabbit was heading and whence the lion was coming. But why do humans have subjective experiences of hunger and fear?

  Not long ago, biologists gave a very simple answer. Subjective experiences are essential for our survival, because if we didn’t feel hunger or fear we would not have bothered to chase rabbits and flee lions. Upon seeing a lion, why did a man flee? Well, he was frightened, so he ran away. Subjective experiences explained human actions. Yet today scientists provide a much more detailed explanation. When a man sees a lion, electric signals move from the eye to the brain. The incoming signals stimulate certain neurons, which react by firing off more signals. These stimulate other neurons down the line, which fire in their turn. If enough of the right neurons fire at a sufficiently rapid rate, commands are sent to the adrenal glands to flood the body with adrenaline, the heart is instructed to beat faster, while neurons in the motor centre send signals down to the leg muscles, which begin to stretch and contract, and the man runs away from the lion.

  Ironically, the better we map this process, the harder it becomes to explain conscious feelings. The better we understand the brain, the more redundant the mind seems. If the entire system works by electric signals passing from here to there, why the hell do we also need to feel fear? If a chain of electrochemical reactions leads all the way from the nerve cells in the eye to the movements of leg muscles, why add subjective experiences to this chain? What do they do? Countless domino pieces can fall one after the other without any need of subjective experiences. Why do neurons need feelings in order to stimulate one another, or in order to tell the adrenal gland to start pumping? Indeed, 99 per cent of bodily activities, including muscle movement and hormonal secretions, take place without any need of conscious feelings. So why do the neurons, muscles and glands need such feelings in the remaining 1 per cent of cases?

  You might argue that we need a mind because the mind stores memories, makes plans and autonomously sparks completely new images and ideas. It doesn’t just respond to outside stimuli. For example, when a man sees a lion, he doesn’t react automatically to the sight of the predator. He remembers that a year ago a lion ate his aunt. He imagines how he would feel if a lion tore him to pieces. He contemplates the fate of his orphaned children. That’s why he flees. Indeed, many chain reactions begin with the mind’s own initiative rather than with any immediate external stimulus. Thus a memory of some prior lion attack might spontaneously pop up in a man’s mind, setting him thinking about the danger posed by lions. He then gets all the tribespeople together and they brainstorm novel methods for scaring lions away.

  But wait a moment. What are all these memories, imaginations and thoughts? Where do they exist? According to current biological theories, our memories, imaginations and thoughts don’t exist in some higher immaterial field. Rather, they too are avalanches of electric signals fired by billions of neurons. Hence even when we figure in memories, imaginations and thoughts, we are still left with a series of electrochemical reactions that pass through billions of neurons, ending with the activity of adrenal glands and leg muscles.

  Is there even a single step on this long and twisting journey where, between the action of one neuron and the reaction of the next, the mind intervenes and decides whether the second neuron should fire or not? Is there any material movement, of even a single electron, that is caused by the subjective experience of fear rather than by the prior movement of some other particle? If there is no such movement – and if every electron moves because another electron moved earlier – why do we need to experience fear? We have no clue.

  Philosophers have encapsulated this riddle in a trick question: what happens in the mind that doesn’t happen in the brain? If nothing happens in the mind except what happens in our massive network of neurons – then why do we need the mind? If something does indeed happen in the mind over and above what happens in the neural network – where the hell does it happen? Suppose I ask you what Homer Simpson thought about Bill Clinton and the Monica Lewinsky scandal. You have probably never thought about this before, so your mind now needs to fuse two previously unrelated memories, perhaps conjuring up an image of Homer drinking beer while watching the president give his ‘I did not have sexual relations with that woman’ speech. Where does this fusion take place?

  Some brain scientists argue that it happens in the ‘global workspace’ created by the interaction of many neurons.4 Yet the word ‘workspace’ is just a metaphor. What is the reality behind the metaphor? Where do the different pieces of information actually meet and fuse? According to current theories, it certainly doesn’t take place in some Platonic fifth dimension. Rather, it takes place, say, where two previously unconnected neurons suddenly start firing signals to one another. A new synapse is formed between the Bill Clinton neuron and the Homer Simpson neuron. But if so, why do we need the conscious experience of memory over and above the physical event of the two neurons connecting?

  We can pose the same riddle in mathematical terms. Present-day dogma holds that organisms are algorithms, and that algorithms can be represented in mathematical formulas. You can use numbers and mathematical symbols to write the series of steps a vending machine takes to prepare a cup of tea, and the series of steps a brain takes when it is alarmed by the approach of a lion. If so, and if conscious experiences fulfil some important function, they must have a mathematical representation. For they are an essential part of the algorithm. When we write the fear algorithm, and break ‘fear’ down into a series of precise calculations, we should be able to point out: ‘Here, step number ninety-three in the calculation process – this is the subjective experience of fear!’ But is there any algorithm in the huge realm of mathematics that contains a subjective experience? So far, we don’t know of any such algorithm. Despite the vast knowledge we have gained in the fields of mathematics and computer science, none of the data-processing systems we have created needs subjective experiences in order to function, and none feels pain, pleasure, anger or love.5

  Maybe we need subjective experiences in order to think about ourselves? An animal wandering the savannah and calculating its chances of survival and reproduction must represent its own actions and decisions to itself, and sometimes communicate them to other animals as well. As the brain tries to create a model of its own decisions, it gets trapped in an infinite digression, and abracadabra! Out of this loop, consciousness pops out.

  Fifty years ago this might have sounded plausible, but not in 2016. Several corporations, such as Google and Tesla, are engineering autonomous cars that already cruise our roads. The algorithms controlling the autonomous car make millions of calculations each second concerning other cars, pedestrians, traffic lights and potholes. The autonomous car successfully stops at red lights, bypasses obstacles and keeps a safe distance from other vehicles – without feeling any fear. The car also needs to take itself into account and to communicate its plans and desires to the surrounding vehicles, because if it decides to swerve to the right, doing so will impact on their behaviour. The car does all that without any problem – but without any consciousness either. The autonomous car isn’t special. Many other computer programs make allowances for their own actions, yet none of them has developed consciousness, and none feels or desires anything.6

  —<
br />
  If we cannot explain the mind, and if we don’t know what function it fulfils, why not just discard it? The history of science is replete with abandoned concepts and theories. For instance, early modern scientists who tried to account for the movement of light postulated the existence of a substance called ether, which supposedly fills the entire universe. Light was thought to be waves of ether. However, scientists failed to find any empirical evidence for the existence of ether, whereas they did come up with alternative and better theories of light. Consequently, they threw ether into the dustbin of science.

  Credit 1.15

  15. The Google autonomous car on the road.

  Similarly, for thousands of years humans used God to explain numerous natural phenomena. What causes lightning to strike? God. What makes the rain fall? God. How did life on earth begin? God did it. Over the last few centuries scientists have not discovered any empirical evidence for God’s existence, while they did find much more detailed explanations for lightning strikes, rain and the origins of life. Consequently, with the exception of a few subfields of philosophy, no article in any peer-review scientific journal takes God’s existence seriously. Historians don’t argue that the Allies won the Second World War because God was on their side; economists don’t blame God for the 1929 economic crisis; and geologists don’t invoke His will to explain tectonic plate movements.

  The same fate has befallen the soul. For thousands of years people believed that all our actions and decisions emanate from our souls. Yet in the absence of any supporting evidence, and given the existence of much more detailed alternative theories, the life sciences have ditched the soul. As private individuals, many biologists and doctors may go on believing in souls. Yet they never write about them in serious scientific journals.