In Our Own Image is a book about the prospects of creating artificial intelligence as well as the cultural, economic, historical, philosophical, and political concerns about it by Greek author and scientist George Zarkadakis. The book considers the problem of AI from the perspectives of human evolution, cybernetics, neuroscience, programming, and computing power.
Zarkadakis begins by briefly speaking of his early years and doctoral research, then spends the rest of the introduction outlining what he will discuss in the rest of the book. The book proper is divided into three parts, each with five or six chapters. The first part covers the evolution of the human brain from the primate brain, especially the most recent 40,000 years. The role of language in accelerating human progress is discussed, as well as the effects of totemic thinking, story-telling, philosophical dualism, and theory of mind. The use of metaphor and narrative to understand the world is examined, along with the inaccuracies inherent in them. The invention, uses, and limitations of the Turing test are explored, as are Asimov’s laws of robotics and the role of AI in fictional stories throughout history.
The second part is about the nature of the mind. The differences in approach between dualism versus monism, rationalism versus empiricism, and materialism versus Platonism are discussed. The thought experiment of the philosophical zombie and the possibility of digital immortality are explained. On the matter of why there appears to be no other intelligent life in the cosmos, Zarkadakis shares an interesting hypothesis: science is an unnatural idea at odds with our cognitive architecture, and an intelligent alien species would be unlikely to widely adopt it. This means that the universe is likely full of Platos, as well as Ancient Greeces, Romes, Indias, Chinas, and Mayas, but is perhaps devoid of Aristotles and societies advanced beyond that of humanity in the early eighteenth century. Daniel Dennett’s explanation of consciousness is overviewed, as well as the contributions of a great number of scientists to the field of cognitive psychology. Finally, the field of cybernetics and its offshoots are examined, showing that the hard problem of consciousness is actually solved with ease. The brain-in-a-vat paradigm of consciousness is shown to be insufficient by applying cybernetic theory.
Everything up to this point lays the foundation for understanding the last part of the book. The third part details the history of computers and programming, from ancient theorists to more recent mathematicians, and from punched cards to modern electronics. The limitations of symbolic logic and the implications thereof against AI in conventional computers are explored, and possible solutions in the form of new electronic components and computer architectures are explained. Charles Babbage’s inventions are discussed, as well as the lost potential of their lack of adoption in their own time. The role of computational technology during World War II is considered, along with the results of government spending on computer research at the time. The development of supercomputers, including IBM’s Deep Blue and Watson, is outlined. The ‘Internet of things’ is compared and contrasted with true AI, and the possible societal impact of large-scale automation of jobs is considered. The possibility of evolving rather than creating AI is examined, as are the possible dispositions of an AI; friendly, malevolent, or apathetic. Interestingly, Zarkadakis shows that there is good reason to believe that a strong AI may exhibit autism spectrum disorders. A short epilogue that begins with a summary and then considers possible economic, political, and social implications of strong AI completes the book.
The book is well-researched and impeccably sourced, at least in its core subject matter. That being said, the book struggles to find an audience, as it can be a bit too technical for the average layperson, but does not venture deeply enough into the subjects it covers to interest a professional in AI-related fields. In other words, it is lukewarm where being either cold or hot is best. Zarkadakis also commits some ultracrepidarianism, particularly in the fields of economics and politics. He seems to believe that AI will overcome the limitations described by Hayek’s knowledge problem and Mises’s economic calculation problem, but unless AI can get inside of our heads and know us better than we know ourselves, this is impossible. In politics, he briefly mentions the possibilities of AI leading to anarchism or to neoreactionary-style absolute monarchies with computerized philosopher-kings, but does not give these possibilities the amount of consideration that they warrant. Finally, the book contains more typographical errors and grammatical abnormalities than a competent editor should fail to correct, though we may grant Zarkadakis some leeway because English is not his first language.
Overall, In Our Own Image is worth reading for those who already have some knowledge of the subject matter but would like to fill gaps in their understanding, but there is room for improvement and expansion.
Read the entire article at ZerothPosition.com