Ray Kurzweil is one of the world’s leading technologists, especially in the area of artificial intelligence (AI). His credentials in the field are about as impressive as it gets. For some time, he has been promoting discussion about what he calls the Singularity.
The Singularity is the moment in human history when machine intelligence surpasses human intelligence in a way that leads to an exponential growth in the development of technology and scientific discovery. I’m using “machine intelligence” and “artificial intelligence” in this article as umbrella terms; the achievement of beyond-human intelligence could be accomplished via organic brain augmentation, genetic engineering, or other means. The Singularity will allow such AI’s to develop smarter AI’s, which will develop yet smarter AI’s, and so on. The result of the exponential growth in technology at this point will be change on a scale that is difficult to imagine; it could mean the transcendence of humanity over our own biology. Assuming it goes well. How to ensure that it does go well is a core issue of "The Singularity Debate".
At a recent conference on the topic, Cognitive Science Professor Douglas Hofstader expressed a good deal of skepticism over the claims of Kurzweil and other futurists. He doesn’t rule out the possibility of superior machine intelligence; he merely holds reservations, given the lack of current evidence. He finished with a somewhat chilling quote from futurist Paul Saffo, "If we have super-intelligent robots, the good news is that they will view us as pets; the bad news is they will view us as food."
Kurzweil estimates that we will have sufficiently reverse-engineered the human brain enough to create human-level artificial intelligence by the year 2029. In a response to those critical of wild speculation and ignoring human social factors, he argued that Singularity is at its root about solving our practical problems. Disease, poverty, ecological and other issues can all be solved through a better understanding of them, and creating effective AI is a means to that end. When we have the knowledge, it can be applied to solving our problems.
Some philosophers, notably John Searle, argue that the entire concept of artificial intelligence is a misnomer. Even if we can create a machine that perfectly emulates the brain, it will not be "intelligent" in the sense that we apply that concept to people. Being a person is inherently tied to our biological nature, intelligence as we know it is a property of fundamentally organic beings. Searle, if I recall correctly, goes so far as to argue that "wetware" is a necessary condition for consciousness, so essentially, machines will never be conscious. They cannot have intention, nor understanding.
Along those lines of thought, even if consciousness and/or intelligence does turn out to be an emergent property of a complex neural network as we see in the brain — whether it exists in the medium of wetware or computer hardware and software — if it is void of hormones, sensations and emotions, a machine intelligence would be an insectile, calculating monster. If Kurzweil is right, it might be a monster that can solve all of our problems. Then again, a significantly more intelligent being modelled after the human brain might be less interested in solving the problems of humanity than it would be in the things that the brain evolved over millions of years to be interested in: sex and power. Hmmmm.
