What is Thought? proposes a model that explains how mind is equivalent to execution of a computer program, addressing aspects such as understanding, meaning, creativity, language, reasoning, learning, and consciousness, that is consistent with extensive data from a variety of fields, and that makes empirical predictions. To achieve this one must first address how the workings of a machine, a computer, can have meaning, what meaning is. The summary of the book in one sentence is: meaning is the computational exploitation of the compact underlying structure of the world, and mind is execution of an evolved program, largely encoded in the genome, that is all about meaning.
The discovery of meaning by evolution comes about through a principle called Occam's razor. As William of Occam posed it in the 14th century, Occam's Razor said one should choose simpler explanations in preference to more complex ones. As such it is a powerful philosophical principle that underlies all science and arguably most day-to-day reasoning. Over the last few decades, computer scientists have formalized this principle to explain concept learning. What is Thought? explains and extrapolates the recent computer science literature (to which I contributed) to posit a central organizing principle of thought: Meaning results from finding a compact enough program (a ``simple enough explanation'') behaving effectively in the world; such a program can only be so compact by virtue of code reuse, factoring into interacting modules that capture real concepts and are reused metaphorically.
For a variety of reasons, including arguments based on complexity theory, developmental biology, evolutionary programming, ethology, and simple inspection, this compact Occam program is most naturally seen to be in the DNA, rather than the brain. Learning and reasoning are then fast and almost automatic because they are constrained by the DNA programming to deal only with meaningful quantities. Evolution itself is argued to exploit meaning in related ways, and thus to speed itself up analogously to how it speeds our reasoning.
This picture explains why artificial intelligence (AI) programs have not achieved understanding in the same way humans do. Finding meaning, finding programs that exploit underlying structure, is a very hard computational problem (in a technical sense: computer scientists say it is NP-hard) that requires extensive computation to solve. Humans are not capable of the hard computation that evolution brought to bear or even that computers bring to bear when they train artificial neural nets. Thus human written programs are generally not Occam and so do not understand. To gain insight into what it means to exploit structure and how it can be equivalent to understanding, What is Thought? discusses AI, Computer science, and human approaches to a variety of problems including Chess, Go, and planning problems. Moreover experiments are described in which modular computer programs were evolved to exploit structure in hard planning problems in a human-like fashion. These evolutionary computing experiments employ new principles for evolving cooperation among modules and achieve results that are much superior to previous evolutionary progamming techniques, at least on the problems tested, and give insight into evolution of cooperation more generally.
This theory explains why language is so highly metaphoric (much more so than non-linguists realize). Metaphor is a manifestation of the reuse of computational modules. Words are labels for meaningful computational modules, explaining how they are learned so rapidly. Theories of why evolution took so long to discover language are discussed. Using the abilility to pass along programs through speech, humans have made cumulative progress in constructing, as part of their minds, useful computational modules built on top of the ones supplied by evolution. The difference between human and chimp intelligence is largely in this additional programming, and thus can be regarded as primarily due to better nurturing.
The many aspects of consciousness are naturally and consistently understood in this context. Evolution produces the program of mind to make decisions favoring the interest of the genes. Creatures with such an internal agenda are naturally said to have will-- that is we think about such creatures and ourselves using a computational module attributing will to them. While it is true that the program of mind has many modules computing different things, the unified self can be understood as the client whose interest the whole system is representing: a reification (manifestation) of the interest of the genes. The nature of what we are and are not aware is naturally explained. We are unaware of extensive computations done by our mind to extract semantically meaningful summary information. We are aware of the meaningful information affecting our decisions because it is the decision making portion of the program that reports what we speak and feel. Qualia (the way experiences "feel" to us) have exactly the appropriate nature and meaning that evolution coded in the DNA so that the compact program behaves effectively. This picture thus explicitly and naturally answers "the hard problem" and says why conscious experience feels "like" something, even though it is nothing but execution of certain computer code. The execution of that computer code that we describe as qualia is like what it is like because it evolved to have certain meanings and thus must be programmed in such a way that it has certain qualities.
No previous familiarity with computer science (or other fields) is assumed-- What is Thought? presents a pedagogical survey of the relevant background for its arguments.