Evolutionary Intelligence

It is often observed that technology is neutral in itself and can correspondingly be used for constructive or destructive purposes depending on human motivation. The theme of this well-informed book is 'a fundamental revolution in the human use of computational intelligence,' which the author sees as comparable to the invention of language, writing, agriculture, and steam and electric power (p.122 ff.). The tools we are now building can potentially enhance our capacity to think at the digital interface between humans and machines as a form of co-evolution – hence the author's term evolutionary intelligence (EI) as 'the use of technology to enhance our capacities when we speak or listen' without outsourcing agency to the machine. They are affordances, and their role is that of adviser while the overall purpose is 'systematic compensation for the well-known biases in evolved human cognition,' to which a separate chapter is devoted.

The information we need for optimal decision-making is not always available, and a number of developments in augmented reality are reviewed as potential sources of relevant information – though also of overload. How much do we actually need to know? There is an interesting chart detailing seven mechanisms of compensatory evolutionary intelligence (p. 58) where common cognitive shortcomings can be corrected. A separate chapter is devoted to potential downsides, including artificially intelligent crime, misrepresentation, privacy issues and cognitive atrophy. The author sees the best overall policy as transparency, but I think he underestimates the prevalence of secrecy under the guise of national security, along with existing censorship regimes deployed in the name of misinformation, disinformation, and even malinformation; also the close relationship between military intelligence (e.g. DARPA) and the development of the Internet as a means of citizen surveillance. He does warn of corporate and government system capture, which I think is already going on as an understandable exercise of power. On a more general level, the analysis is predominantly left hemisphere and cognitive with little attention paid to feelings and none to emotions – however, it is a positive contribution to our understanding of the human– machine interface.