How to Create a Mind: The Secret of Human Thought Revealed

Our old brain—the one we had before we were mammals—has not disappeared. Indeed it still provides much of our motivation in seeking gratification and avoiding danger. These goals are modulated, however, by our neocortex, which dominates the human brain in both mass and activity.

Animals used to live and survive without a neocortex, and indeed all nonmammalian animals continue to do so today. We can view the human neocortex as the great sublimator—thus our primitive motivation to avoid a large predator may be transformed by the neocortex today into completing an assignment to impress our boss; the great hunt may become writing a book on, say, the mind; and pursuing reproduction may become gaining public recognition or decorating your apartment. (Well, this last motivation is not always so hidden.)

The neocortex is likewise good at helping us solve problems because it can accurately model the world, reflecting its true hierarchical nature. But it is the old brain that presents us with those problems. Of course, like any clever bureaucracy, the neocortex often deals with the problems it is assigned by redefining them. On that note, let’s review the information processing in the old brain.

The Sensory Pathway

Although we experience the illusion of receiving high-resolution images from our eyes, what the optic nerve actually sends to the brain is just a series of outlines and clues about points of interest in our visual field. We then essentially hallucinate the world from cortical memories that interpret a series of movies with very low data rates that arrive in parallel channels. In a study published in Nature, Frank S. Werblin, professor of molecular and cell biology at the University of California at Berkeley, and doctoral student Boton Roska, MD, showed that the optic nerve carries ten to twelve output channels, each of which carries only a small amount of information about a given scene.² One group of what are called ganglion cells sends information only about edges (changes in contrast). Another group detects only large areas of uniform color, whereas a third group is sensitive only to the backgrounds behind figures of interest.

The visual pathway in the brain.

“Even though we think we see the world so fully, what we are receiving is really just hints, edges in space and time,” says Werblin. “These 12 pictures of the world constitute all the information we will ever have about what’s out there, and from these 12 pictures, which are so sparse, we reconstruct the richness of the visual world. I’m curious how nature selected these 12 simple movies and how it can be that they are sufficient to provide us with all the information we seem to need.”

This data reduction is what in the AI field we call “sparse coding.” We have found in creating artificial systems that throwing most of the input information away and retaining only the most salient details provides superior results. Otherwise the limited ability to process information in a neocortex (biological or otherwise) gets overwhelmed.

Seven of the twelve low-data-rate “movies” sent by the optic nerve to the brain.

The processing of auditory information from the human cochlea through the subcortical regions and then through the early stages of the neocortex has been meticulously modeled by Lloyd Watts and his research team at Audience, Inc.³ They have developed research technology that extracts 600 different frequency bands (60 per octave) from sound. This comes much closer to the estimate of 3,000 bands extracted by the human cochlea (compared with commercial speech recognition, which uses only 16 to 32 bands). Using two microphones and its detailed (and high–spectral resolution) model of auditory processing, Audience has created a commercial technology (with somewhat lower spectral resolution than its research system) that effectively removes background noise from conversations. This is now being used in many popular cell phones and is an impressive example of a commercial product based on an understanding of how the human auditory perceptual system is able to focus on one sound source of interest.

The auditory pathway in the brain.

Inputs from the body (estimated at hundreds of megabits per second), including that of nerves from the skin, muscles, organs, and other areas, stream into the upper spinal cord. These messages involve more than just communication about touch; in addition they carry information about temperature, acid levels (for example, lactic acid in muscles), the movement of food through the gastrointestinal tract, and many other signals. This data is processed through the brain stem and midbrain. Key cells called lamina 1 neurons create a map of the body, representing its current state, not unlike the displays used by flight controllers to track airplanes. From here the sensory data heads to a mysterious region called the thalamus, which brings us to our next topic.

A simplified model of auditory processing in both the subcortical areas (areas prior to the neocortex) and the neocortex, created by Audience, Inc. Figure adapted from L. Watts, “Reverse-Engineering the Human Auditory Pathway,” in J. Liu et al. (eds.), WCCI 2012 (Berlin: Springer-Verlag, 2012), p. 49.

The Thalamus

From the midbrain, sensory information then flows through a nut-sized region called the posterior ventromedial nucleus (VMpo) of the thalamus, which computes complex reactions to bodily states such as “this tastes terrible,” “what a stench,” or “that light touch is stimulating.” The increasingly processed information ends up at two regions of the neocortex called the insula. These structures, the size of small fingers, are located on the left and right sides of the neocortex. Dr. Arthur Craig of the Barrow Neurological Institute in Phoenix describes the VMpo and the two insula regions as “a system that represents the material me.”⁴

The sensory-touch pathway in the brain.

Among its other functions, the thalamus is considered a gateway for preprocessed sensory information to enter the neocortex. In addition to the tactile information flowing through the VMpo, processed information from the optic nerve (which, as noted above, has already been substantially transformed) is sent to a region of the thalamus called the lateral geniculate nucleus, which then sends it on to the V1 region of the neocortex. Information from the auditory sense is passed through the medial geniculate nucleus of the thalamus en route to the early auditory regions of the neocortex. All of our sensory data (except, apparently, for the olfactory system, which uses the olfactory bulb instead) passes through specific regions of the thalamus.

The most significant role of the thalamus, however, is its continual communication with the neocortex. The pattern recognizers in the neocortex send tentative results to the thalamus and receive responses principally using both excitatory and inhibitory reciprocal signals from layer VI of each recognizer. Keep in mind that these are not wireless messages, so that there needs to be an extraordinary amount of actual wiring (in the form of axons) running between all regions of the neocortex and the thalamus. Consider the vast amount of real estate (in terms of the physical mass of connections required) for the hundreds of millions of pattern recognizers in the neocortex to be constantly checking in with the thalamus.⁵

So what are the hundreds of millions of neocortical pattern recognizers talking to the thalamus about? It is apparently an important conversation, because profound damage to the main region of the thalamus bilaterally can lead to prolonged unconsciousness. A person with a damaged thalamus may still have activity in his neocortex, in that the self-triggering thinking by association can still work. But directed thinking—the kind that will get us out of bed, into our car, and sitting at our desk at work—does not function without a thalamus. In a famous case, twenty-one-year-old Karen Ann Quinlan suffered a heart attack and respiratory failure and remained in an unresponsive, apparently vegetative state for ten years. When she died, her autopsy revealed that her neocortex was normal but her thalamus had been destroyed.

In order to play its key role in our ability to direct attention, the thalamus relies on the structured knowledge contained in the neocortex. It can step through a list (stored in the neocortex), enabling us to follow a train of thought or follow a plan of action. We are apparently able to keep up to about four items in our working memory at a time, two per hemisphere according to recent research by neuroscientists at the MIT Picower Institute for Learning and Memory.⁶ The issue of whether the thalamus is in charge of the neocortex or vice versa is far from clear, but we are unable to function without both.

The Hippocampus

Each brain hemisphere contains a hippocampus, a small region that looks like a sea horse tucked in the medial temporal lobe. Its primary function is to remember novel events. Since sensory information flows through the neocortex, it is up to the neocortex to determine that an experience is novel in order to present it to the hippocampus. It does so either by failing to recognize a particular set of features (for example, a new face) or by realizing that an otherwise familiar situation now has unique attributes (such as your spouse’s wearing a fake mustache).

The hippocampus is capable of remembering these situations, although it appears to do so primarily through pointers into the neocortex. So memories in the hippocampus are also stored as lower-level patterns that were earlier recognized and stored in the neocortex. For animals without a neocortex to modulate sensory experiences, the hippocampus will simply remember the information from the senses, although this will have undergone sensory preprocessing (for example, the transformations performed by the optic nerve).

Although the hippocampus makes use of the neocortex (if a particular brain has one) as its scratch pad, its memory (of pointers into the neocortex) is not inherently hierarchical. Animals without a neocortex can accordingly remember things using their hippocampus, but their recollections will not be hierarchical.

The capacity of the hippocampus is limited, so its memory is short-term. It will transfer a particular sequence of patterns from its short-term memory to the long-term hierarchical memory of the neocortex by playing this memory sequence to the neocortex over and over again. We need, therefore, a hippocampus in order to learn new memories and skills (although strictly motor skills appear to use a different mechanism). Someone with damage to both copies of her hippocampus will retain her existing memories but will not be able to form new ones.

University of Southern California neuroscientist Theodore Berger and his colleagues modeled the hippocampus of a rat and have successfully experimented with implanting an artificial one. In a study reported in 2011, the USC scientists blocked particular learned behaviors in rats with drugs. Using an artificial hippocampus, the rats were able to quickly relearn the behavior. “Flip the switch on, and the rats remember. Flip it off and the rats forget,” Berger wrote, referring to his ability to control the neural implant remotely. In another experiment the scientists allowed their artificial hippocampus to work alongside the rats’ natural one. The result was that the ability of the rats to learn new behaviors strengthened. “These integrated experimental modeling studies show for the first time,” Berger explained, “that…a neural prosthesis capable of real-time identification and manipulation of the encoding process can restore and even enhance cognitive mnemonic processes.”⁷ The hippocampus is one of the first regions damaged by Alzheimer’s, so one goal of this research is to develop a neural implant for humans that will mitigate this first phase of damage from the disease.

The Cerebellum

There are two approaches you can use to catch a fly ball. You could solve the complex simultaneous differential equations controlling the ball’s movement as well as further equations governing your own particular angle in viewing the ball, and then compute even more equations on how to move your body, arm, and hand to be in the right place at the right time.

This is not the approach that your brain adopts. It basically simplifies the problem by collapsing a lot of equations into a simple trend model, considering the trends of where the ball appears to be in your field of vision and how quickly it is moving within it. It does the same thing with your hand, making essentially linear predictions of the ball’s apparent position in your field of view and that of your hand. The goal, of course, is to make sure they meet at the same point in space and time. If the ball appears to be dropping too quickly and your hand appears to be moving too slowly, your brain will direct your hand to move more quickly, so that the trends will coincide. This “Gordian knot” solution to what would otherwise be an intractable mathematical problem is called basis functions, and they are carried out by the cerebellum, a bean-shaped and appropriately baseball-sized region that sits on the brain stem.⁸

The cerebellum is an old-brain region that once controlled virtually all hominid movements. It still contains half of the neurons in the brain, although most are relatively small ones, so the region constitutes only about 10 percent of the weight of the brain. The cerebellum likewise represents another instance of massive repetition in the design of the brain. There is relatively little information about its design in the genome, as its structure is a pattern of several neurons that is repeated billions of times. As with the neocortex, there is uniformity across its structure.⁹

Most of the function of controlling our muscles has been taken over by the neocortex, using the same pattern recognition algorithms that it uses for perception and cognition. In the case of movement, we can more appropriately refer to the neocortex’s function as pattern implementation. The neocortex does make use of the memory in the cerebellum to record delicate scripts of movements—for example, your signature and certain flourishes in artistic expression such as music and dance. Studies of the role of the cerebellum during the learning of handwriting by children reveal that the Purkinje cells of the cerebellum actually sample the sequence of movements, with each one sensitive to a specific sample.¹⁰ Because most of our movement is now controlled by the neocortex, many people can manage with a relatively modest obvious disability even with significant damage to the cerebellum, except that their movements may become less graceful.

The neocortex can also call upon the cerebellum to use its ability to compute real-time basis functions to anticipate what the results of actions would be that we are considering but have not yet carried out (and may never carry out), as well as the actions or possible actions of others. It is another example of the innate built-in linear predictors in the brain.

Substantial progress has been made in simulating the cerebellum with respect to the ability to respond dynamically to sensory cues using the basis functions I discussed above, in both bottom-up simulations (based on biochemical models) and top-down simulations (based on mathematical models of how each repeating unit in the cerebellum operates).¹¹

Pleasure and Fear

If the neocortex is good at solving problems, then what is the main problem we are trying to solve? The problem that evolution has always tried to solve is survival of the species. That translates into the survival of the individual, and each of us uses his or her own neocortex to interpret that in myriad ways. In order to survive, animals need to procure their next meal while at the same time avoiding becoming someone else’s meal. They also need to reproduce. The earliest brains evolved pleasure and fear systems that rewarded the fulfillment of these fundamental needs along with basic behaviors that facilitated them. As environments and competing species gradually changed, biological evolution made corresponding alterations. With the advent of hierarchical thinking, the satisfaction of critical drives became more complex, as it was now subject to the vast complex of ideas within ideas. But despite its considerable modulation by the neocortex, the old brain is still alive and well and still motivating us with pleasure and fear.

One region that is associated with pleasure is the nucleus accumbens. In famous experiments conducted in the 1950s, rats that were able to directly stimulate this small region (by pushing a lever that activated implanted electrodes) preferred doing so to anything else, including having sex or eating, ultimately exhausting and starving themselves to death.¹² In humans, other regions are also involved in pleasure, such as the ventral pallidum and, of course, the neocortex itself.

Pleasure is also regulated by chemicals such as dopamine and serotonin. It is beyond the scope of this book to discuss these systems in detail, but it is important to recognize that we have inherited these mechanisms from our premammalian cousins. It is the job of our neocortex to enable us to be the master of pleasure and fear and not their slave. To the extent that we are often subject to addictive behaviors, the neocortex is not always successful in this endeavor. Dopamine in particular is a neurotransmitter involved in the experience of pleasure. If anything good happens to us—winning the lottery, gaining the recognition of our peers, getting a hug from a loved one, or even subtle achievements such as getting a friend to laugh at a joke—we experience a release of dopamine. Sometimes we, like the rats who died overstimulating their nucleus accumbens, use a shortcut to achieve these bursts of pleasure, which is not always a good idea.

Gambling, for example, can release dopamine, at least when you win, but this is dependent on its inherent lack of predictability. Gambling may work for the purpose of releasing dopamine for a while, but given that the odds are intentionally stacked against you (otherwise the business model of a casino wouldn’t work), it can become ruinous as a regular strategy. Similar dangers are associated with any addictive behavior. A particular genetic mutation of the dopamine-receptor D2 gene causes especially strong feelings of pleasure from initial experiences with addictive substances and behaviors, but as is well known (but not always well heeded), the ability of these substances to produce pleasure on subsequent use gradually declines. Another genetic mutation results in people’s not receiving normal levels of dopamine release from everyday accomplishments, which can also lead to seeking enhanced early experiences with addictive activities. The minority of the population that has these genetic proclivities to addiction creates an enormous social and medical problem. Even those who manage to avoid severely addictive behaviors struggle with balancing the rewards of dopamine release with the consequences of the behaviors that release them.

Serotonin is a neurotransmitter that plays a major role in the regulation of mood. In higher levels it is associated with feelings of well-being and contentment. Serotonin has other functions, including modulating synaptic strength, appetite, sleep, sexual desire, and digestion. Antidepression drugs such as selective serotonin reuptake inhibitors (which tend to increase serotonin levels available to receptors) tend to have far-reaching effects, not all of them desirable (such as suppressing libido). Unlike actions in the neocortex, where recognition of patterns and activations of axons affect only a small number of neocortical circuits at a time, these substances affect large regions of the brain or even the entire nervous system.

Each hemisphere of the human brain has an amygdala, which consists of an almond-shaped region comprising several small lobes. The amygdala is also part of the old brain and is involved in processing a number of types of emotional responses, the most notable of which is fear. In premammalian animals, certain preprogrammed stimuli representing danger feed directly into the amygdala, which in turn triggers the “fight or flight” mechanism. In humans the amygdala now depends on perceptions of danger to be transmitted by the neocortex. A negative comment by your boss, for example, might trigger such a response by generating the fear of losing your job (or maybe not, if you have confidence in a plan B). Once the amygdala does decide that danger is ahead, an ancient sequence of events occurs. The amygdala signals the pituitary gland to release a hormone called ACTH (adrenocorticotropin). This in turn triggers the stress hormone cortisol from the adrenal glands, which results in more energy being provided to your muscles and nervous system. The adrenal glands also produce adrenaline and noradrenaline, which suppress your digestive, immune, and reproductive systems (figuring that these are not high-priority processes in an emergency). Levels of blood pressure, blood sugar, cholesterol, and fibrinogen (which speeds blood clotting) all rise. Heart rate and respiration go up. Even your pupils dilate so that you have better visual acuity of your enemy or your escape route. This is all very useful if a real danger such as a predator suddenly crosses your path. It is well known that in today’s world, the chronic activation of this fight-or-flight mechanism can lead to permanent health damage in terms of hypertension, high cholesterol levels, and other problems.

The system of global neurotransmitter levels, such as serotonin, and hormone levels, such as dopamine, is intricate, and we could spend the rest of this book on the issue (as a great many books have done), but it is worth pointing out that the bandwidth of information (the rate of information processing) in this system is very low compared with the bandwidth of the neocortex. There are only a limited number of substances involved and the levels of these chemicals tend to change slowly and are relatively universal across the brain, as compared with the neocortex, which is composed of hundreds of trillions of connections that can change quickly.

It is fair to say that our emotional experiences take place in both the old and the new brains. Thinking takes place in the new brain (the neocortex), but feeling takes place in both. Any emulation of human behavior will therefore need to model both. However, if it is just human cognitive intelligence that we are after, the neocortex is sufficient. We can replace the old brain with the more direct motivation of a nonbiological neocortex to achieve the goals that we assign to it. For example, in the case of Watson, the goal was simply stated: Come up with correct answers to Jeopardy! queries (albeit these were further modulated by a program that understood Jeopardy! wagering). In the case of the new Watson system being jointly developed by Nuance and IBM for medical knowledge, the goal is to help treat human disease. Future systems can have goals such as actually curing disease and alleviating poverty. A lot of the pleasure-fear struggle is already obsolete for humans, as the old brain evolved long before even primitive human society got started; indeed most of it is reptilian.

There is a continual struggle in the human brain as to whether the old or the new brain is in charge. The old brain tries to set the agenda with its control of pleasure and fear experiences, whereas the new brain is continually trying to understand the relatively primitive algorithms of the old brain and seeking to manipulate it to its own agenda. Keep in mind that the amygdala is unable to evaluate danger on its own—in the human brain it relies on the neocortex to make those judgments. Is that person a friend or a foe, a lover or a threat? Only the neocortex can decide.

To the extent that we are not directly engaged in mortal combat and hunting for food, we have succeeded in at least partially sublimating our ancient drives to more creative endeavors. On that note, we’ll discuss creativity and love in the next chapter.