Artificial Intelligence - the way forward?

I've been reading Ray Kurzweil's The Singularity is Near in which he discusses the possibility of creating intelligence, consciousness, body parts - in fact almost anything artificially. He is a great believer in the power of future machines and their creators to be able to reverse engineer the human brain, model its processes and then go far beyond the capabilities of our current brain power.

I have to say his arguments are persuasive and exciting. Whether we will actually get there in my lifetime is a big guess. But since his book was written in 2004 I suspect a lot has moved forward already and next I need to search the Internet for new developments - the Internet being one of those things that point to the possible exponential growth of other technologies.

Should there be any restrictions on our developments? Do we want machines that are more powerful (not just in brute strength - we are obviously well past that, but in intelligence, creativity, replication)? What are the advantages and the dangers? Kurzweil considers all these and is definitely of a positivist school - science can and will provide the answers and we need to chase after them. Have a look at his website www.KurzweilAI.net and see if you agree.