A Brief History of Intelligence: Evolution, AI, and the Five Breakthroughs That Made Our Brains

by Max Solomon Bennett

Indeed, neocortex size of primates is correlated not only with social-group size but also with social savviness. An outgrowth of this arms race seems to have been a blossoming of many human social instincts, both the good (friendships, reciprocity, reconciliation, trust, sharing) and the bad (tribalism, nepotism, deception). While many aspects of these behavioral changes did not require any particularly clever new brain systems, there was indeed an intellectual feat underlying this politicking: the ability to engage in theory of mind.

Most evidence suggests that despite the dramatic expansion in size, the brain of our primate ancestor, and of primates today, was largely the same as that of the early mammals. A bigger hindbrain, a bigger basal ganglia, a bigger neocortex, but still all the same regions connected in all the same fundamental ways.

So might it be the case that the surprising smarts of primates, with all their theory of mind, politicking, and trickery, was a consequence of nothing more than brain scaling?

Although most of the brain of the early primates was merely a scaled-up mammal brain, there were, in fact, certain areas of neocortex that were truly new.

The first is the granular prefrontal cortex (gPFC), which was a new addition to the frontal cortex.* This newer granular prefrontal cortex wraps around the much older agranular prefrontal cortex (aPFC). The second new area of neocortex, which I will call the primate sensory cortex (PSC), is an amalgamation of several new areas of sensory cortex that emerged in primates.* The gPFC and PSC are extremely interconnected with each other, making up their own new network of frontal and sensory neocortical regions.

It is their input and output connectivity that renders them new; it is what these areas construct a generative model of that unlocked fundamentally new cognitive abilities.

the granular prefrontal cortex plays a key role in your ability to project yourself—your intentions, feelings, thoughts, personality, and knowledge—into your rendered simulations, whether they are about the past or some imagined future.

if we want true humanlike AI systems, theory of mind will undeniably be an essential component of that system. In fact, theory of mind might be the most essential aspect of building a successful future with superintelligent AI systems.

Unlike the intellectual abilities that had emerged in the breakthroughs prior, theory of mind was not born from the need to survive the dangers of hungry predators or inaccessible prey, but instead from the subtler and far more cutting dangers of politics. Politics was the origin story of Breakthrough #4, but it is far from the entire story. As we will see in the next two chapters, theory of mind in early primates was repurposed for two other new abilities.

It was long assumed that tool use was uniquely human, but tool use has now been found across many primates. Monkeys and apes not only use sticks to fish termites; they also use rocks to break open nuts, grass to floss, moss for sponges, clubs to smash beehives, and even twigs to clean their ears. In the years since Goodall studied these chimps, tool use has been found all over the animal kingdom. Elephants pick up branches with their trunks to swat flies and scratch themselves. Mongooses use anvils to break open nuts. Crows use sticks to spear larvae. Octopuses gather large shells to make shields. Wrasses fish have been found to use rocks to break open clams to get to the inner food. But primate tool use is more sophisticated than tool use in other animals.

groups of chimpanzees often exhibit over twenty different tool-using behaviors. Further, with the possible exception of birds and elephants, only primates have been shown to actively manufacture their tools.

If the driver of brain evolution in early primates was a politicking arms race, why would primates be uniquely good tool users? If the new brain regions of primates were “designed” to enable theory of mind, then from where do the unique tool-using skills of primates emerge?

In other words, the neurons in the premotor and motor areas of a monkey’s neocortex—those that control a monkey’s own movements—not only activated when they performed those specific fine motor skills, but also when they merely watched others perform them. Rizzolatti called these “mirror neurons.” Over the subsequent twenty years, Rizzolatti’s mirror neurons have been found in numerous behaviors (grasping, placing, holding, finger movements, chewing, lip smacking, sticking one’s tongue out), across multiple areas of the brain (premotor cortex, parietal lobe, motor cortex), and across numerous species of primates.

people with impairments in performing specific movements, also show impairments in understanding the intentions of those very same movements in others. The subregions of premotor cortex required for controlling a given set of motor skills are the same subregions required for understanding the intentions of others performing those same motor skills.

These studies demonstrate that the premotor cortex is involved in specifically understanding the sensorimotor aspects of others’ behavior—inferring the strength required to pick up a box or the type of tool someone intended to hold.

The main benefit is that it helps us, as it helped early primates, learn new skills through observation. We already saw in chapter 14 that mentally rehearsing actions improves performance when actually performing actions.

when they observe a chord they do not yet know, their premotor cortex becomes way more activated than when they observe a chord they already know how to play.

Premotor activation is not just correlated with imitation learning; it seems to be, at least in some contexts, necessary for imitation learning.

Transmissibility Beats Ingenuity

I’m going to bet that practically all these skills were acquired by observing others, not by your own independent ingenuity.

Without transmission from others, most chimps never figure tool use out on their own; in fact, a young chimp that doesn’t learn, through observing others, to crack nuts by the age of five will not acquire the skill later in life.

The ability to use tools is less about ingenuity and more about transmissibility. Ingenuity must occur only once if transmissibility occurs frequently; if at least one member of a group figures out how to manufacture and use a termite-catching stick, the entire group can acquire this skill and continuously pass it down throughout generations.

For skills to be transmitted through a population, you don’t need teachers—dutiful observation by novices will do. But active teaching can substantially improve the transmission of skills.

another possibility is that theory of mind enables novices to identify the intent of a complex skill, which makes them highly motivated to keep trying to adopt it. Theory of mind enables a chimp child to realize that the reason it is not getting food with its stick while its mother is getting food is that its mother has a skill it does not yet have. This enables a continuous motivation to acquire the skill, even if it takes a long time to master.

Understanding the intentions of movements is essential for observational learning to work; it enables us to filter out extraneous movements and extract the essence of a skill.

At first, Ross would do most of the driving, then control would be passed to the AI system for a moment, and any mistakes it made would be quickly recovered by Ross. Over time, Ross gave more control to the AI system until it was driving well on its own.

Ross’s AI system was still crashing cars after a million frames of expert data. In contrast, with this new strategy of active teaching, his AI system was driving almost perfectly after only a handful of laps. This is not unlike a teacher chimpanzee correcting the movements of a child learning a new skill.

The second approach to imitation learning in robotics is called “inverse reinforcement learning.” Instead of trying to directly copy the driving decisions that a human makes in response to a picture of the road, what if an AI system first attempts to identify the intent of the human’s driving decisions?

This technique is called “inverse reinforcement learning” because these systems first try to learn the reward function they believe the skilled expert is optimizing for (i.e., their “intent”), and then these systems learn by trial and error, rewarding and punishing themselves using this inferred reward function. An inverse reinforcement learning algorithm starts from an observed behavior and produces its own reward function; whereas in standard reinforcement learning the reward function is hard-coded and not learned.

Both a carnivore and a non-fruit-eating herbivore can survive by hunting or grazing only when they are hungry. But a frugivore must plan its trips in advance before it is hungry. Setting up camp en route to a nearby popular fruit patch the night before requires anticipating the fact that you will be hungry tomorrow if you don’t take preemptive steps tonight to get to the food early.

But such seasonal hoarding is not nearly as cognitively challenging as the daily need to change your plans based on how hungry you will be tomorrow. Further, it isn’t even clear that mice hoard food because they understand that they will be hungry in the future. Indeed, lab mice—although they have never suffered from a cold winter without food—automatically start hoarding food if you simply lower the temperature of their environment, an effect seen only in northern species of mice who have had to evolve to survive winters.

Doris Bischof-Kohler and her husband, Norbert Bischof, proposed a novel hypothesis about what was unique about planning in humans: They hypothesized that while other animals can make plans based on current needs (like how to get to food when they are hungry), only humans can make plans based on future needs (like how to get food for your trip next week, even though you are not hungry right now).

But recent evidence has called this into question.

Fascinatingly, squirrel monkeys learn to select the low treat option, while rats continue to select the high treat option. Squirrel monkeys are capable of resisting the temptation to have treats now, in anticipation of something—water—that they don’t even want yet. In other words, monkeys can make a decision in anticipation of a future need. In contrast, rats were entirely unable to do this—they

This suggests that perhaps Suddendorf’s Bischof-Kohler hypothesis was correct that anticipating a future need is a more difficult form of planning and was correct that some animals should be able to plan but unable to anticipate future needs (such as rats). But it may not be the case that only humans were endowed with this ability. It may instead be the province of many primates.

First, it seems that both theory of mind and anticipating future needs are present, even in a primitive form, in primates, but not in many other mammals, suggesting both abilities emerged around the same time in early primates. Second, people make similar types of mistakes in tasks of theory of mind and of anticipating future needs.

When hungry, you overestimate your own future hunger. The ability to anticipate future needs would have offered numerous benefits to our ancestral frugivores. It would have enabled our ancestors to plan their foraging routes long in advance, thereby ensuring they were the first to get newly ripened fruits.

It laid the foundation for our ability to make long term plans over vast stretches of time.

There are three broad abilities that seem to have emerged in early primates: Theory of mind: inferring intent and knowledge of others Imitation learning: acquiring novel skills through observation Anticipating future needs: taking an action now to satisfy a want in the future, even though I do not want it now

The main change to the brain of early primates, besides its size, was the addition of new areas of neocortex.

It suggests that these new intellectual skills must emerge from some new clever application of the neocortex and not some novel computational trick.

Dunbar argues that the social-brain hypothesis and the ecological-brain hypothesis are two sides of the same coin. The ability to mentalize may have simultaneously unlocked both the ability to successfully forage fruits and to successfully politick. The pressures of both frugivorism and social hierarchies may have converged to produce continual evolutionary pressure to develop and elaborate brain regions—such as the gPFC—for modeling your own mind.

If we were to scrunch the six hundred million years of evolutionary time—from which the first brains emerged until today—into a single calendar year, then we would now find ourselves perched at Christmas Eve, the final seven days of December. Over the next “seven days,” our ancestors will go from foraging fruits to flying Falcon 9 rockets. Let’s find out how.

Darwin believed that “the difference in mind between man and the higher animals, great as it is, is certainly one of degree and not of kind.”

But as the evidence continues to roll in, it seems that Darwin may have been right.

there is no neurological structure found in the human brain that is not also found in the brain of our fellow apes, and evidence suggests that the human brain is literally just a scaled-up primate brain: a bigger neocortex, a bigger basal ganglia, but still containing all the same areas wired in all the same ways.

Was there perhaps no breakthrough at all? This seems to be the most reasonable interpretation—but for one crucial exception. And it is in this singular exception that we see the first hints of what it means to be human.

Our Unique Communication Organisms had been communicating with each other long before early humans uttered their first words.

Human language differs from other forms of animal communication in two ways. First, no other known form of naturally occurring animal communication assigns declarative labels (otherwise known as symbols).

Declarative labeling, on the other hand, is a special feature of human language. A declarative label is one that assigns an object or behavior an arbitrary symbol—“That is a cow,” “That is running,”—without any imperative at all.