The early chapters, which I found the least interesting, put forward Kurzweil's theory of how the cerebral neocortex works. The concept of pattern recognition is central to this. Although Kurzweil does say a little about other brain areas, including the thalamus, hippocampus, and cerebellum, as well as the structures responsible for feelings of pleasure and pain, his main focus, not surprisingly, is on the neocortex, which he sees as the distinctive mammalian innovation that supervises and controlls all the more 'primitive' apparatus beneath.
This hierarchical view is an ancient idea, going back to Plato, that was given a neurophysiological twist by Paul MacLean in the last century. MacLean's concept of the 'triune brain' was taken up at a popular level by a number of writers, notably Arthur Koestler, who thought it explained human irrational behaviour. The model perhaps suffers from being excessively 'cerebral'. As Antonio Damasio has pointed out in Decartes' Error, the brain does not exist in isolation but is part of the body as a whole. He uses the metaphor of a landscape to describe this. The viscera (heart, lungs, gut) and the muscles are the components of this landscape, and a 'feeling' is a momentary view of part of that landscape. These feelings are totally essential to the quality of being human.
For me, the later chapters, in which Kurzweil explores the nature of consciousness in relation to artificial intelligence, are the most interesting part of the book. He presents a succinct history of the development of computers which I found useful. He is prominent in the field of artificial intelligence and his views on its future are therefore well worth reading.
He is almost certainly correct in saying both that apparently conscious computers will be built (they will pass the Turing test) and that we will quickly accept them as conscious. Many of us already swear at inanimate objects such as present-day computers when they don't do what we expect (and think of Basil Fawlty cursing and kicking his car when it failed to start), so we would be easily persuaded to relate to these supercomputers as conscious individuals.
But that doesn't answer what the philosopher David Chalmers has famously called the 'hard question', which is how physical processes in a brain (or a computer?) can give rise to subjective experiences of colour, sound, heat and cold, and so on (qualia). Kurzweil's position is that if a machine exhibits behaviour that is indistinguishable from that of other beings we regard as conscious, such as humans, that is as much as we can legitimately ask for. But is it? The explanatory gap between physical events and consciousness still remains for many of us.
Perhaps the hard question is insoluble, at least for us. Some philosophers think so. Kurzweil makes a passing reference to Colin McGinn but fails to mention McGinn's basic view, which is that the human brain is simply incapable of understanding consciousness. In his book The Mysterious Flame McGinn speculates that to understand consciousness we would need to have superhuman mental powers, which might perhaps be achieved by genetic engineering. No doubt Kurzweil's supercomputers would also do the job, but presumably we, with our limited intelligence, would be unable to understand the solution.
16 August 2014