> 3. Multi-threaded programming will fall out of favor by 2012. Not really. We are already in the multicore age, and it will only get moreticore. What that pre

> 3. Multi-threaded programming will fall out of favor by 2012. Not really. We are already in the multicore age, and it will only get moreticore. What that prediction missed was that Moore's Law finally broke down for single-threaded performance. Single cores got only about 3x faster from 2004 to today, rather than continuing the historical trend with four more doublings. Moore's Law is alive and well but now requires parallel architecture to exploit. But you can't blame an armchair pundit for missing this one. The 3 GHz clock speed barrier was hardly something a layman could have predicted; it's an artifact of extremely sophisticated manufacturing processes and infinitesimally detailed subatomic physics. A layman couldn't have said whether the physical limits of clock speed would be reached at 1 GHz or 100 GHz and in 2005 or 2025. > 6. A new internet community-hangout will appear. One that you and I will frequent. This succeeded way more than anyone thought. More than "you and I", our moms and grandparents even hang out there. It is Facebook.