The Nature of Consciousness: How the Internet Could Learn to Feel

"Romantic reductionist" neuroscientist Christof Koch discusses the scientific side of consciousness, including the notion that all matter is, to varying degrees, sentient.



If you had to list the hardest problems in science -- the questions even some scientists say are insoluble -- you would probably end up with two:

  • Where do the laws of physics come from?  
  • How does the physical stuff in our brains produce conscious experience? 

Even though philosophers have obsessed over the "mind-body problem" for centuries, the mystery of consciousness wasn't considered a proper scientific question until two or three decades ago. Then, a couple of things happened. Brain-imaging technologies finally gave neuroscience some high-powered tools to peer inside our brains while we think. And a few renowned scientists -- most famously, Francis Crick -- claimed that neuroscientists had to tackle consciousness if they were ever going to understand the brain.

By the 1980s, Crick had jumped from molecular biology to neuroscience and moved from England to California. There he found a brilliant young collaborator, Christof Koch, the son of German diplomats who'd recently landed a job as an assistant professor of biology and engineering at the California Institute of Technology. For the next 16 years -- until Crick's death in 1994 -- they worked together, searching for the neural correlates of consciousness.

Koch remains on the front lines of neurobiology. In fact, he will soon leave Caltech to work full-time as Chief Scientific Officer of the Allen Institute for Brain Science in Seattle. After years of publishing scientific papers, he has now written a trade book, Consciousness: Confessions of a Romantic Reductionist. Somewhere between memoir and popular science, the book offers a highly personal glimpse into the mind of an unconventional scientist: a lapsed Catholic who teamed up with the staunch atheist Crick, and the eminent neuroscientist who speculates about the consciousness of bees and squid and even bacteria. In the first of a two-part interview, we talked about the wiring of our brains and the possibility that the Internet itself may become conscious.


Why have you devoted so much of your life searching for the neural roots of consciousness?

Koch: Consciousness is the central factor of our lives. The only way I know I exist is because I'm conscious. I might be mistaken about who exactly I am -- for example, how attractive I am to the opposite sex -- but there's no doubt I have feelings of pain, pleasure, anger, of being a man, of waking up. Until recently, science has neglected to incorporate the fact of consciousness into its theories. If science wants a complete understanding of everything in the universe, it has to include consciousness.

What makes consciousness such a difficult problem for scientists to explain?

Koch: Well, unlike black holes or the Higgs boson or molecules, consciousness has both an external perspective and an intrinsic perspective. In other words, you can weigh the Higgs boson and the molecule. You can poke them. You can measure them. Scientists and engineers are very good at doing that, but we don't think a black hole feels like anything. We don't believe the Higgs boson or a single nerve cell feels like anything. But a healthy human brain feels something if it's awake.

You actually see a world. How does this picture get into your head? That's the mystery. And because it has both an exterior, third-person perspective as well as an interior, first-person perspective, it's unique among all the phenomena in the universe. This means it's a little bit more difficult to attack using a scientific point of view. It doesn't mean it's impossible.

Or maybe a lot more difficult if we consider the complexity of the brain. How many neurons and synapses are in the human brain?

Koch: The average human brain has a hundred billion neurons and synapses on the order of a hundred trillion or so. But it's not just sheer numbers. It's the incredibly complex and specific ways in which these things are wired up. That's what makes it different from a gigantic sand dune, which might have a billion particles of sand, or from a galaxy. Our Milky Way, for example, contains a hundred billion suns, but the way these suns interact is very simple compared to the way neurons interact with each other.

So it doesn't matter so much what the neurons are made of. It's how they're organized and wired together.

Koch: Correct. Unless you believe in some magic substance attached to our brain that exudes consciousness, which certainly no scientist believes, then what matters is not the stuff the brain is made of, but the relationship of that stuff to each other. It's the fact that you have these neurons and they interact in very complicated ways. In principle, if you could replicate that interaction, let's say in silicon on a computer, you would get the same phenomena, including consciousness.

Are you saying the Internet could become conscious, or maybe already is conscious?

Koch: That's possible. It's a working hypothesis that comes out of artificial intelligence. It doesn't matter so much that you're made out of neurons and bones and muscles. Obviously, if we lose neurons in a stroke or in a degenerative disease like Alzheimer's, we lose consciousness. But in principle, what matters for consciousness is the fact that you have these incredibly complicated little machines, these little switching devices called nerve cells and synapses, and they're wired together in amazingly complicated ways. The Internet now already has a couple of billion nodes. Each node is a computer. Each one of these computers contains a couple of billion transistors, so it is in principle possible that the complexity of the Internet is such that it feels like something to be conscious. I mean, that's what it would be if the Internet as a whole has consciousness. Depending on the exact state of the transistors in the Internet, it might feel sad one day and happy another day, or whatever the equivalent is in Internet space.

You're serious about using these words? The Internet could feel sad or happy?

Koch: What I'm serious about is that the Internet, in principle, could have conscious states. Now, do these conscious states express happiness? Do they express pain? Pleasure? Anger? Red? Blue? That really depends on the exact kind of relationship between the transistors, the nodes, the computers. It's more difficult to ascertain what exactly it feels. But there's no question that in principle it could feel something.

Would humans recognize that certain parts of the Internet are conscious? Or is that beyond our understanding?

Koch: That's an excellent question. If we had a theory of consciousness, we could analyze it and say yes, this entity, this simulacrum, is conscious. Or because it displays independent behavior. At some point, suddenly it develops some autonomous behavior that nobody programmed into it, right? Then, people would go, "Whoa! What just happened here?" It just sort of self-organized in some really weird way. It wasn't a bug. It wasn't a virus. It wasn't a botnet that was paid for by some nefarious organization. It did it by itself. If this autonomous behavior happens on a regular basis, then I think many people would say, yeah, I guess it's alive in some sense, and it may have conscious sensation.

I think we need to back up for a moment. How do you define consciousness?

Koch: Typically, it means having subjective states. You see something. You hear something. You're aware of yourself. You're angry. You're sad. Those are all different conscious states. Now, that's not a very precise definition. But if you think historically, almost every scientific field has a working definition and the definitions are subject to change. For example, my Caltech colleague Michael Brown has redefined planets. So Pluto is not a planet anymore, right? Because astronomers got together and decided that. And what's a gene? A gene is very tricky to define. Over the last 50 years, people have had all sorts of changing definitions. Consciousness is not easy to define, but don't worry too much about the definition. Otherwise, you get trapped in endless discussions about what exactly you mean. It's much more important to have a working definition, run with it, do experiments, and then modify it as necessary.

We assume humans are conscious, and most of us also think dogs and elephants and mice have some degree of consciousness, but what about other animals? Do lizards have consciousness? Are ants conscious? What about bacteria? Are these useful questions?

Koch: Not right now. In the fullness of time, they have to be answered. But right now, let's stick with cases that are undoubtedly conscious. That includes people, although not all people. You might remember Terri Schiavo. It was very controversial whether she was actually conscious or not. She was clearly alive, but because of anoxia and the damage sustained by her brain, she couldn't communicate with the outside world, and medical and scientific opinion held that she wasn't conscious. Yet she was moaning on occasion and making reflex-like movements. So there can be controversial cases even with people. What about a newborn infant? What about a fetus? Even there it's not totally clear. As you said, most people would agree that cats, dogs, mice and elephants are conscious, but what about non-mammals? So for now, let's just stick with some simple examples that we can actually work with in a clinic or a lab. Let's understand the neural basis of consciousness in them. In the fullness of time, we can then look at the squid and the octopus, which are very complex. Also bees, birds, worms. And in the future, we'll be able to answer a question like bacteria.

It's fascinating to hear you say that at some point, we'll have to consider the question of whether bacteria are conscious. The assumption is that you need a mind to be conscious and bacteria don't have minds.

Koch: That's correct. Therefore, bacteria are probably not conscious. However, right now it's a guessing game. Most people say bugs are just little machines, but if you look at bees, they're amazingly complex. They can recognize single faces. And how they find a nesting site when they swarm in the spring is very complex behavior. How do we know it doesn't feel like something to be a bee? It's not that bees will have as elaborate minds as we have, but we cannot rule out at this time that it doesn't feel like something to be a bee, or that they consciously smell the gold nectar of certain flowers.

You are searching for the neural correlates of consciousness. The challenge is not simply to find the hot spots in the brain that are active during certain mental activities, but actually to explain how neurons and synapses and chemical circuits generate conscious experiences. This is more than mapping correlations between brain activities and what people say they're thinking about, which we can already see to some degree with fMRI scans. Isn't the real challenge to explain how specific neural correlates are wired together to generate conscious experiences?

Koch: The challenge is to see why this activity that's in your visual brain gives rise to a picture of a cup of coffee in front of me, and other neural activity in the olfactory part of my brain gives rise to the smell of freshly brewed coffee. In both cases, neurons are firing. They look the same. They release a bunch of chemicals like sodium and potassium. Somewhere else, a similar neural activity gives rise to pain. But in another part of the brain, the cerebellum, similar neural activity doesn't give rise to anything. The cerebellum, which is like a little brain at the back of the cortex, contains three-quarters of all the cells in your brain, roughly 69 billion. If you lose the cerebellum, you won't be a ballet dancer or a rock climber anymore. You'll walk with a funny gait. You'll talk very strangely. But your consciousness is only mildly impaired, if at all. So that implies that some types of neural activity are privileged. Certain parts of the brain seem to have a much closer association with consciousness than others. We have to understand what type of activity, what type of circuit, gives rise to conscious sensation.

I know a lot of neuroscientists get very excited about fMRI, but it sounds like the current state of technology is actually a fairly blunt instrument for trying to figure out what's going on in the brain.

Koch: It's a wonderful instrument because it allows us to peer inside your brain safely, but it's also very blunt. Remember, each cubic millimeter of the brain has a hundred thousand cells. So if I take the volume of brain that would be the size of a grain of rice, it contains about a half a million different nerve cells, and maybe seven miles of cable. It's very, very complicated. Yet in fMRI, all you can see clearly is this one box as a single point. But we know this one point contains half a million neurons, some of which may fire, but most of which don't fire. Some might fire less, some might fire more or fire in a very complicated pattern. But fMRI can't see any of that. It can just see the part of the brain that's active or the part that's suppressed. So fMRI is a wonderful tool, but it is very, very crude.

We are only now beginning to realize the complexity of it all. Over the last ten years, we realized there might be a thousand different cell types. They all have highly specialized jobs to do. They look different. They have different chemicals. They use different neurotransmitters. They have different genes encoding for them. They are wired up in a very distinct manner. So it's a little bit like having 86 billion tiles, but they come in sets of a thousand different colors and shapes. They all fit in very specific ways. And now, the big challenge is trying to unravel all of that and see the amazingly specific way that tile number 55 only talks to tile 255 and 972. And it inhibits tile number 17 and 31, and it doesn't talk to any of the other ones. That's what it's beginning to look like. So how does anxiety or sadness or the color blue emerge out of all of that?

There are plenty of very smart scientists and philosophers who say we will never crack the fundamental mystery of how matter turns into mental experience. For instance, the philosopher David Chalmers has talked about the "hard problem of consciousness." He says subjective experience is fundamentally different from what biology and physics can tell us. And he believes that science will never be able to bridge this divide between mind and matter. Do you disagree?

Koch: Well, if you look at the historical record of philosophers, it's pretty disastrous. This is little acknowledged. Lots of people from 150 years ago said the same thing about life: "You shall never understand what life is. It requires a special force, élan vital." It didn't turn out to be true. The laws of physics describe life. Roughly 200 years ago, famous philosophers said we'll never know what stars are made of. That was 20 years before they discovered spectroscopy, and of course they realized you can analyze the elements out of which stars are made. So I'm profoundly skeptical when philosophers tell us, once again, what we'll never know. Science has a spectacular record of understanding the universe. Yes, right now it's a hard problem. For its size, the brain is by far the most complex system in the known universe. There's no guarantee that we'll understand it. Our cognitive apparatus just might not be up to it, but in principle I don't see any reason why we should be unable to understand it. Just because some philosopher doesn't get it doesn't mean we shall never know this. It's ridiculous. But a lot of people are very happy about that message because, for various reasons, they don't really want to understand things in the way science does.

David Chalmers has also suggested that consciousness may be a fundamental property in itself, not reducible to the laws of physics. What do you think of this idea?

Koch: This is an old idea, which I'm quite sympathetic to. Just like physicists in the 16th century realized that magnets have this weird property of attracting other magnets or pieces of metal. This is an invisible force. We realized our description of the universe wasn't complete and we had to include magnetic fields. To understand the universe, we say there's space, time, matter, and energy. It may well be that to fully understand consciousness, there is something equally fundamental: experience. I think a particular type of complex system, which of course needs to be defined, will have conscious experience. That's just the way the universe is, just like there's matter. In this hypothesis, we live in a universe that has this fundamental property of consciousness.

Isn't this called "panpsychism," the idea that all matter is sentient to some degree?

Koch: They're related but not strictly equivalent. From this hypothesis, it's easy to say that all matter has some consciousness associated with it, but very complicated pieces of matter like your brain, or, to a lesser extent the brain of a dog, or to a much lesser extent the brain of a bee, are complex systems with various degrees of consciousness. You could say even the worm and the fly have a little bit of consciousness, although it might be much less consciousness than you have in your deep sleep, which is not a lot.

I'm guessing that you're pretty unusual in suggesting that consciousness might be a fundamental property of nature, along with the laws of physics. Do many other neuroscientists think this?

Koch: No. Most biologists don't think about it. It makes them feel very uncomfortable thinking about these issues. But I think that's silly. We have to explain where consciousness comes from. There are two possibilities. One is as an emergent property. Small brains don't have it. Large brains have it. It's like the wetness of water. One molecule of H2O isn't wet. Two molecules aren't wet. But you put a whole bunch together, like in my cup of coffee here, and you get a wet liquid. Water is the quintessential emergent property. There are many other emergent properties in the world around us -- for example, democracy. One or two people don't have democracy, but in certain types of organization you get this emergent phenomenon called democracy. The claim of consciousness is of that ilk. I used to believe that, but I've been studying it now for 25 years, and consciousness is so strange. It's different from anything else because it has a subjective aspect. I don't think it's an emergent property. A much more elegant theory is that certain types of complex systems are conscious, and what type of consciousness they have depends exactly on the complexity of the systems.

I was surprised to see your book invoke Pierre Teilhard de Chardin, the Jesuit priest and paleontologist who believed the universe is becoming more conscious as it gets more complex. Most scientists write off Teilhard as a religious apologist.

Koch: Most scientists don't even know about him. He had this idea about evolution where he argued that from very simple micro molecules to single cell organisms to multi-cell organisms to simple animals to complex animals to us is the emergence of complexity. He observed that the universe was getting more and more complex, and he postulated this would continue. Essentially, he postulated something like the Internet. He called it the "noosphere" -- the sphere of knowledge that covers the entire planet and is heavily interconnected. He died in 1955, long before any of this emerged, and he postulated that human society would evolve into a very complicated entity that would become self-conscious. He thought this would happen on other planets and throughout the entire universe, and the universe in some weird state would become self-conscious. It's all totally speculative, but I do like some of these ideas. I see a universe that's conducive to the formation of stable molecules and to life. And I do believe complexity is associated with consciousness. Therefore, we seem to live in a universe that's particularly conducive to the emergence of consciousness. That's why I call myself a "romantic reductionist."