Quantum computers are fine. What we also need to focus on are different kinds/models of computing on a traditional computer. The industry is very very young and there are so many untested, postponed or unfinished ideas that were left in limbo (pun kinda intended) because of the market.<p>What I would really love to see are some of these computing models brought back to life. Alan Kay's idea of computing looks very interesting albeit very scattered; Exokernels is another interesting thought, which could lead up to very interesting distributed computing model (like an irc, but c stands for computing);
Transactional Memory is another. There are a lot of them, and I think they are best served by an organisation in lieu with Bell Labs or Xerox Research..