I will be a little surprised if ems take over fifty years (given business as usual), but I guess ninety is not beyond the bounds of possibility.
Currently, we don't have basic tech for scanning, nor even anything on obvious track (for having both precision and volume speed). Present simulation methods are still in their infancy. From general considerations of technological progress, we can expect that to change eventually, especially if nanotech takes off (for scanning, not simulation). Give it 20-50 years for when tech catches up, and another 10-30 years for reliable simulation of short-term dynamics. Then comes another difficulty: we need to support all brain reconfiguration processes to enable l...
Ian Morris on "why the west rules", which seems to be a provocative title for an interesting book on historical geographical trends and their projection into the future: http://www.youtube.com/watch?v=tvkHiL-H2io. He starts talking about the future at minute 27 and basically concludes that a singularity scenario is one of two possibilities for the 21st century, the other being collapse. Nothing new, but encouraging to see this increasingly in the mainstream.