Narrator: In 1962, the animated sitcom The Jetsons offered a tantalizing glimpse into the future. It accurately predicted video calls, flat-screen TVs, and even robotic vacuum cleaners. Yet, one of its most iconic creations, the empathetic and autonomous robot maid Rosey, remains firmly in the realm of science fiction. Why is it that we can engineer complex technologies but struggle to create an artificial intelligence with the common sense and emotional understanding of a human? This gap between our technological prowess and our grasp of true intelligence lies at the heart of a profound scientific puzzle.

In his book, A Brief History of Intelligence, author and AI entrepreneur Max Bennett argues that to build a mind, we must first understand how our own was built. He proposes that intelligence is not a single, monolithic quality but a series of five major evolutionary breakthroughs, each adding a new layer of capability over hundreds of millions of years. By tracing this epic journey from the first simple brains to the complex mind of a human, the book provides a roadmap for understanding both ourselves and the future of AI.

Narrator: The story of intelligence begins not with complex thought, but with a simple problem: movement. For the first life on Earth, movement was random. But around 600 million years ago, a revolutionary development occurred with the emergence of bilaterians, creatures with a distinct front and back. This new body plan enabled a new skill: steering. To steer, an organism needed a rudimentary brain that could do one fundamental thing: categorize the world into "good" things to move toward and "bad" things to move away from.

This basic principle is elegantly demonstrated by the nematode worm. With a brain of just 302 neurons, a nematode navigates its world by assigning a positive or negative "valence" to stimuli. When it detects an increasing concentration of food particles, it turns toward the source. This isn't a sophisticated understanding of its environment; it's a simple, hardwired rule. This ability to assign valence and steer accordingly was the first breakthrough in intelligence. It established the brain's core function as a decision-making engine, governed by the two sovereign masters that philosopher Jeremy Bentham would later identify: pain and pleasure. This foundational ability to distinguish good from bad set the stage for all subsequent, more complex forms of learning.

Narrator: The second major breakthrough equipped animals with the ability to learn from experience. In the late 1890s, psychologist Edward Thorndike conducted a series of famous experiments that revealed this mechanism. He placed hungry cats inside "puzzle boxes," with a food treat visible just outside. To escape, the cat had to perform a specific action, like pulling a string or pressing a lever. Initially, the cat would thrash about randomly until it accidentally triggered the escape mechanism. With each subsequent trial, the cat became faster at escaping, having learned to associate the specific action with the positive outcome.

Thorndike called this the "law of effect," a principle now known as reinforcement learning. This ability, which emerged with the first vertebrates, allows an animal to reinforce behaviors that lead to rewards and suppress those that lead to punishment. However, this created a new challenge known as the "temporal credit assignment problem." If a reward comes long after a series of actions, how does the brain know which specific action to credit? The vertebrate brain solved this with a system remarkably similar to an AI algorithm called Temporal Difference (TD) learning. Instead of waiting for an actual reward, the brain uses the neuromodulator dopamine to signal changes in predicted future rewards. A surprising cue that predicts a future reward triggers a dopamine spike, reinforcing the action that led to it. This mechanism, linking prediction to action, allowed vertebrates to learn complex sequences of behavior and adapt to a much more dynamic world.

Narrator: While reinforcement learning is powerful, it is still reactive. The third breakthrough, which emerged with the first mammals, was the ability to simulate the world internally. This was made possible by the evolution of the neocortex, a new brain structure that functions as a generative model of reality. Instead of learning only by doing, mammals could learn by imagining.

Psychologist Edward Tolman observed this in the 1930s. He noticed that rats at a fork in a maze would often pause, looking back and forth before choosing a path. He called this "vicarious trial and error." Decades later, neuroscientists confirmed his theory by recording the brain activity of rats in the same situation. They found that as the rat paused, its hippocampus—the brain's map-maker—was rapidly firing sequences of neurons that represented the possible future paths. The rat was literally imagining the consequences of each choice before acting. This ability to simulate allows for planning, episodic memory (reliving the past), and even counterfactual learning—learning from what could have happened. This "imaginarium" gave early mammals a profound survival advantage, allowing them to out-plan their rivals.

Narrator: The fourth breakthrough was the evolution of "theory of mind," the ability to model the minds of others. While many animals live in social groups, primate societies are uniquely political. Dominance is not just about strength but also about alliances and social savvy. To navigate this complex world, primates needed to understand the intentions, beliefs, and desires of others.

A classic story from primatologist Emil Menzel illustrates this perfectly. He showed a subordinate chimpanzee named Belle where he hid food. When the dominant male, Rock, was present, he would simply take all the food. Belle quickly learned to deceive him. First, she tried to sit on the food. When Rock pushed her aside, she learned to wait until he was distracted. Rock, in turn, began pretending to be uninterested, only to rush for the food when Belle made her move. The deception escalated, with Belle eventually trying to lead Rock to incorrect locations. This intricate dance of deception and counter-deception is only possible because both chimpanzees were modeling the other's mind—what they knew, what they wanted, and what they intended to do. This ability, which is far more developed in primates, was enabled by new neocortical areas, like the granular prefrontal cortex, and became the foundation for imitation, teaching, and advanced cooperation.

Narrator: The final breakthrough is the one that truly sets humans apart: language. While apes can learn a rudimentary vocabulary, human language is unique in its use of grammar and declarative labels. This allows humans to do something no other species can: transfer their inner simulations—their thoughts, plans, and memories—to other minds with incredible fidelity. This ability to share imagined worlds is the engine of cumulative culture.

The tragic history of the Aboriginal people of Tasmania provides a stark example of this principle. For thousands of years, they possessed a sophisticated toolkit, including bone tools, nets, and cold-weather clothing. However, when rising sea levels isolated them from mainland Australia around 8,000 years ago, their population dwindled. With fewer minds to hold and transmit knowledge, their technology began to regress. By the 1800s, they had lost the knowledge to create many of their ancestors' most vital tools. This shows that technology and culture are not stored in an individual, but in the collective network of brains connected by language. Language allows ideas to be passed down, combined, and improved across generations, creating a "collective brain" whose knowledge far exceeds that of any single person. This is humanity's superpower.

Narrator: The single most important takeaway from A Brief History of Intelligence is that our own intelligence is not a singular entity but a layered architecture, built piece by piece over eons. Our brains still contain the ancient steering mechanisms of a worm, the reinforcement systems of a fish, and the simulation engine of a shrew, all operating alongside the more recent primate and human innovations. We are not purely rational beings; we are a complex blend of instinct, emotion, and reason.

As we stand on the verge of what Bennett calls the sixth breakthrough—the creation of artificial superintelligence—this evolutionary history offers a crucial lesson. To build a truly intelligent and beneficial AI, we cannot simply focus on logic and computation. We must understand the entire architecture of the mind, including the foundational systems that govern motivation, emotion, and sociality. The ultimate challenge, then, is not just a technical one. It is a deeply human one: to look back at the four-billion-year journey that created us and decide which parts of our own intelligence are worth replicating, and which we should strive to improve.