Brain for Learning

Long-lasting effects of brief sensory experience must be mediated by long-term changes in the strength of connections between neurons in the brain. This phenomenon is known as synaptic plasticity, and the physical location of such change is referred to as the engram. This chapter illustrates how multiple learning and memory systems might be implemented in different anatomical modules of the brain and what role dopamine plays in learning. Most of these neurobiological and behavioral observations can be accounted for by reinforcement learning theory. The goal of reinforcement is to understand how utilities must be altered by experience so that rational choices based on the utility functions can result in the most desirable outcomes through learning.

Conclusion

To better prepare for the future society in which artificial intelligences (AI) will have much more pervasive influence on our lives, a better understanding of the difference between AI and human intelligence is necessary. Human and biological intelligence cannot be separated from the process of self-replication. Therefore, a fundamental gap exists between human intelligence and AI until AI acquires artificial life. Humans’ social and metacognitive intelligence most clearly distinguish human intelligence from nonhuman intelligence. Although advances are likely to improve the functioning of AI, AI will remain a function of human activity. However, if AI can learn to self-replicate and thus become a life form, albeit a man-made one, outcomes become uncertain.

Social Intelligence and Altruism

According to the social intelligence hypothesis, the unusual enlargement of primate brains, including the human brain, was driven by the complexity of social decision-making primates face in their societies. Social decision-making is fundamentally more complex due to the recursive nature of social reasoning. This chapter begins with the review of game theory and illustrates how game theory has transformed neuroscience research on social decision-making. Some of the topics covered include the supposed death of game theory, altruism and its dark side, cooperation, the theory of the mind, the prisoner’s dilemma, the recursive mind, and the social brain.

Levels of Intelligence

Intelligence is the ability to find solutions to complex problems a life faces in a complex and uncertain environment. This cannot be captured by a standardized numerical score, such as IQ, that focuses on a narrow range of cognitive capabilities, such as working memory or verbal fluency. Intelligence also cannot be separated from the most essential property of all life forms, which is self-replication. This chapter briefly reviews the evolutionary history and diversity of intelligence from single-cell organisms to humans. The chapter includes a definition of intelligence and a discussion of how the nervous system works, the simple behavior of reflexes and the limitations of reflexes, connectome (the comprehensive map of all the connections in an animal’s nervous system), the multiple controllers for muscles, and the social nature of many behaviors. The chapter also includes a case study on eye movements.

Artificial Intelligence

Compared to the human brain, current artificial intelligence technology is limited in that its goals are determined by human developers and users. Similarly, despite their superficial similarities, modern-day computers and human brains have many differences. Building blocks of human brain that are functionally equivalent to transistors, functional units of digital computers, have not been identified, and we do not know whether hardware and software are separable in the human brain. This chapter uses Mars rovers as a case study to illustrate the autonomy of intelligent robots, because machines dependent on human intelligence is not genuinely intelligent.

Self-Replicating Machines

Intelligence cannot be separated from the agent, as can be best illustrated by numerous parasites that alter the behavior of their hosts. This is because intelligence is a function of life whose essential feature is self-replication. The behavior of a host that promotes self-replication of parasites at the cost of its own replication is intelligent for the parasite, but not for the host. This chapters reviews important biological concepts necessary to understand biological intelligence, including the origin of life, biochemistry of gene replication and protein synthesis, and brain evolution. The author argues that because brain functions can be modified by experience, genes do not fully control the brain. Nonetheless, the brain is not completely free from the influence of genes.

Brain and Genes

During evolution, more complex structures and functions emerge as a result of division of labor and specialization. Often, this leads to a conflict between a principal who determines the nature of division of labor and an agent who implements the overall goal set by the principal. The principal–agent relationship is a useful construct to understand the relationship between the genes and the brain. This chapter covers the key assumptions of the principal–agent theory and how this can be validated in the case of gene–brain relationship. The author argues that because the brain must learn solutions to problems that genes cannot handle, learning is central to intelligence.

Brain and Decision-Making

The brain is an organ of decision-making, and its studies requires multidisciplinary approaches, including neuroscience, economics, and psychology, three areas that have jointly formed the basis of neuroeconomics. This chapters reviews the utility theory to illustrate why it is important to start with a formal theory of decision-making to understand the biological mechanisms of decision-making, including the function of the brain. In addition, it discusses the limits of the classical utility theory and how its concepts are related to the common notion of happiness. Finally, it reviews key findings from neuroscience about how the brain makes decisions.

Intelligence and Self

The chapter considers self-knowledge or self-insight. The concept of self is an inevitable consequence of recursive social reasoning, but it is bound to cause logical paradox due to its self-reference. Broadly speaking, self is an example of metacognition, namely, a consequence of cognition applied to evaluate other cognitive processes, which includes the feeling of knowing and other abilities to select the optimal decision-making strategies. As the number and complexity of different learning strategies increase, this also produces undesirable side effects, including negative emotions, such as disappointment and regret, as well as potential failures of metacognition, which might manifest as mental illnesses.

Why Learning?

Once the genes delegate the responsibility of decision-making to the brain, the most important function of the brain is to develop successful decision-making strategies by incorporating new information about the animal’s environment. The complexity of this process increased during evolution, and in mammals, including humans, the brain utilizes multiple learning strategies to produce the most appropriate motor responses. After illustrating this using response and place learning, this chapter reviews the history of research on animal learning, including a potential conflict between different learning strategies. In particular, the author addresses the important role of classical conditioning and instrumental conditioning in learning.