An alarm book expressing all fears about AI coming, a kind of typical modern myth. Sure, such texts are quite useful: they are well read and discussed, performing the functions of criticism...
![Image](http://images.realclear.com/219207_5_.png)
But in the case of Barrat, as well as many other authors, his weak philosophical background is making itself felt. He copes well with his task -- frightening readers from the standpoint of common sense. The only trouble is that this common sense does not take into account the newest knowledge of mankind about the nature of one's own consciousness.
Barrat makes two fundamental mistakes. First, it is a false anthropomorphization: not every mind should be like human's. Second, it is a myth of Cartesian evidence: he assumes that the required condition for intellect is self-awareness and space-time localization. In other words, he builds mind by the Descartes model where all ideas are clear and distinct, the mind is locked into res extensa, and we are perfectly aware of what we think.
But from the modern neurophysiology and philosophy point of view, this is a lie from beginning to end. Intellect does not necessarily have to be a thing locked in some space: evidence of this is numerous projects of distributed computing. Intellect does not necessarily have to be self-aware: it can be purely functional. And finally, the fundamental point of this Cartesianism suggests that if AI emerges, we will certainly fix this moment and begin to worry about it. And AI, accordingly, will begin to realize our presence and will become nervous, too.
But, in fact, we won't notice when AI comes. It will exist completely outside the context of our understanding, and it will not have a single reason to tell us what it is. And nothing will change for humans: we will use computers, as we use now but, maybe, more intensively. Social problems are obvious, of course, as robots will replace some of us at work, but there is conceptually no reason to believe that a person and AI generally will notice each other.
Any thoughts? Do you support Stephen Hawking's statement about AI being "the worst event in the history of our civilization"? Or, as well as James Barrat, he exaggerates the danger coming from the artificial intelligence?