Tonight's ICT4D meetup asks the question, "What's Next?" While it's always risky to predict the future, I think the groundwork for the next decade is largely stable.
Mobile is globally today where the Internet was in the US in the early nineties -- if you wanted access, you could find it. It might be difficult, costly, or shared, but it was available. Think of the changes in the Internet over the past decade and a half -- we've moved from a largely text-only interface with gopher, newsgroups and email being the key players to the web, and now the rich, ajax-y web 2.0.
I think a similar growth could happen (and already is, to some extent) in the global mobile market. As access to mobile networks spread, the possibility to have more and exciting applications filter down to even the cheapest handsets becomes more likely.
Imagine a simple survey app that received, presented, and then encoded, compressed and encrypted questions and answers through SMS automatically, instead of the laborious current manual encoding used in election monitoring. That alone, enabled on the basic cell phone platforms would revolutionize data reporting by reducing training time while also improving accuracy, human rights protections (via encryption), and reducing the opportunities for falsified data to be put into the system.