Now, I'm certainly not a luddite - hell, technology pays my wages. But, I've long been of the same mind as Ray with regards to historic evolution.
Put simply, the most intelligent species on the planet becomes the dominant force. For now, this is us (humans) of course (although not all of us fall into that category).
As it currently stands, if we follow Moore's Law, then by 2020 computers will be as powerful as the human mind. After this point, they will outstrip us (they already do of course in areas such as computation, mathematics, data analysis etc).
Does this mean that at 2020, machines will mount an uprising and go a-stomping around in superiority or indeed farm us á la Matrix. Well, no - at least not yet.
We need to establish what defines sentience and therefore determine if artificial intelligence will ever attain it. It can be argued that AI (invented by man, remember) already has sapience - the ability to act with judgement. Even a basic if…then…else statement indicates that. However, as it currently stands, those judgements are determined by man and those that develop the code commanding the AI. "Sentience is the ability to perceive" according to Wikipedia and also "as the ability to experience suffering". With computers able to perceive the world around them, perceive scent and more, it can be argued that they are edging closer to sentience.
Ray's 2005-penned name for the point where AI becomes a definable species is "the technological singularity" - the next technological-evolutionary jump. Several people before Ray spoke about such an event. In 1965, statistician I. J. Good qas quoted as saying:
"Let an ultraintelligent machine be defined as a machine that can far surpass all the intellectual activities of any man however clever. Since the design of machines is one of these intellectual activities, an ultraintelligent machine could design even better machines; there would then unquestionably be an 'intelligence explosion,' and the intelligence of man would be left far behind. Thus the first ultraintelligent machine is the last invention that man need ever make."
To supercede our intellect, AI needs to be able to reason beyond senses. The philisopher Descartes argued that the human senses are utterly unreliable for perceiving reality. We can only perceive what they are able to translate. From the Matrix:
"Morpheus: What is 'real'? How do you define 'real'? If you're talking about what you can feel, what you can smell, what you can taste and see, then "real" is simply electrical signals interpreted by your brain..."
Descartes originally expressed this by observing that during a dream, we sense everything as if it were real. Thus, our senses can easily be mistaken or fooled. He then goes on to consider wax as an example of our minds being more reliable than our senses. When holding a piece of wax, our senses inform us that it has certain colour, texture, weight, smell, structure etc. However, when melted in a flame, the wax's colour, smell, texture, structure etc can ll change. Our senses perceive a different object, yet our minds are able to understand that it is still wax.
AI will need to be able to understand the world in this way if it is to surpass us. But, I feel it won't be long.
The biggest barrier in my mind is the fact that machines are man-made. Thus they cannot surpass us. However, nanotechnology and artificial neural networkshave the opportunity to change this. Once AI achieves awareness of itself (Terminator 2's SkyNet was said to have done this. The movie might seem like mindless action but for me, it embodies a very real threat) then it will assess the most efficient methods for its own development and begin creation of itself (as AI is a globally-spanning entity, networked as such. Humans, by comparison, do not have a global network of minds; the Internet being the closest thing to it. Ironically, this is an artificial network).
I'm researching more, so this ends here for now. I intend to revisit the subject, but there's just too much information for me to absorb coherently right now.
Additional info: Ray Kurzweil's site about AI and the singularity





Quite pleased with it since I've never used pen before (cross-hatching and all that) and that it only took 30 mins. Her left boob's gone a bit odd though :) Right, back to work…
The original photo is