When faced with the breadth and majesty of books like The Singularity is Near, Engines of Creation and Hacking Matter it is difficult not to feel inspired and optimistic about the future. There is a particular brand of techno-progressive idealogue who can weave images in the mind as effectively as any of the technologies they predict.
There is a problem. Ray Kurzweil calls it "the argument from incredulity". The picture of a plentiful future peopled by synthetic superhumans is so compelling it registers on our internal TGTBTM (Too Good To Be True Meter) as a being unreasonably optimistic.
Most rational people who don't live within the bleeding edge of pattern-recognition technologies simply refuse to accept that the clanking, smoking, bad-for-the-environment hulks of machined metal that are what most people still associate with the word "machine" could ever turn into anything as sublimely effective and versatile as a human being.
It is interesting to look back at how technological changes have been predicted to occur and how they actually did occur. Take the widely-predicted advent of powered flight in the late Victorian era. We see those wonderfully fanciful drawings of individuals flitting around in one-man ornithopters. The artist that comes to mind is Heath Robinson, although his were probably actually drawn in the early twentieth century.
One of the great fears of the Victorians, as exemplified by the writing of H.G. Wells, was the concept of "war in the air" (the other great fear being that of the advent of powered cavalry). That WITA became a terrible reality through the Blitz of London within the lifetimes of many of those already adults during the Victorian era suggests that our current fears of GM viruses, dangerous artificial intelligences and rampant nanotechnology scenarios may not be as far-fetched as we could imagine.
I disagree with Kurzweil's sweeping "law of accelerating returns". The most compelling point he makes in support of this is that over the lifetime of the universe, complexity has tended to increase in a manner which strongly resembles an exponential graph.
However when we get down to the nitty-gritty level of technological development over the years and decades of a human lifetime "progress" (an amorphous term in this debate) seems to happen in fits and starts, and is strongly influenced by political and economic factors.
That a young (perhaps 20 year-old) Victorian in 1900 growing up in a world without heavier than air flight could live to see the Apollo and the Moon landings (Bertrand Russell (1872-1970) springs to mind) is an extraordinary thing.
Another point to consider before dismissing Kurzweil entirely is the nature of technological change. The counterpart of the technology of aircraft in the early twentieth century is the development of computers in the late twentieth and early twenty-first century. Computers tend to affect things in ways that were difficult to comprehend beforehand. Highly efficient information management and versatile and decentralised communication generate a lot of side-effects that would not be immediately obvious to someone who was not aware of computers.
Spime and conversion and the gradual move from centralised manufacturing towards decentralised CAD/CAM machines will allow the effects of abundant, effective computing to move from the world of pure information to the real world the rest of us inhabit.
At the moment our world is still very much run by meat. Humans are required for their versatility and imagination, if for little else. I think that of all Ray Kurzweil's predictions, the most likely to turn out to be correct is his belief that there will be a strong convergence between human and machine intelligence with the result of an (even more) profound change in the way the world works.