Although the "singularitarian" sounds a bit cultish, Ray Kurzweil makes a convincing case for the approaching singularity based on sound arguments, data and his credentials (MIT educated and still giving guest lectures @MIT). Although some of his predictions for 2010 made in 2005 are bit too optimistic (computers in clothing, ubiquitous), the computing power predictions are spot on, and even perhaps conservative. This means that while we will have computing power and sofware AI equivalent to the brain by 2029, the actual singularity may happen a little later, perhaps by a decade, due to the slower impact of technology to ALL of society due to economics and politics (although he has addressed that). The singularity may even have the potential to drive change of the economic model from supply-and-demand capitalism to a more resource-based economy due to the abundance (lack of scarcity) of energy and food in the 2030's.
It doesn't however address the dark side of technology too much: the potential of mass produced robots to temporarily cause mass unemployment for a decade or two until personnel skills are updated; self-replicating nanobots that reproduce until they consume all the matter and energy in the universe, and large scale terrorist attack using bio/nano-technology, etc.
Otherwise, an excellent read worth the time and money for it makes you think about the future and decisions regarding your personal life, career path and investments.