David Brin has a nice post summarizing his vision for "what we need from science" :

Given the daunting range of problems and opportunities that we face, I'd have to say that our most urgent scientific and technological need is to develop better methods for problem-solving.

Some pieces to the puzzle are already getting attention. Governments and big institutions are developing ways to combine sensor meshes and data mining with powerful analytic and projection tools. But this emphasis on centralized or professional-level anticipation ignores the other half of the solution -- generating a resilient citizenry. A populace so knowing and capable that all problems get noticed and addressed, quickly, by a billion eyes

Great point!

What Brin understands is that the "collective" and the "individual" are not in opposition but that smart, disorganized individuals can come together in dynamic, ad-hoc networks to solve problems.

Traditionally, we've relied on certain types of static "organization" (typically hierarchical and procedural) to achieve the benefits of acting collectively. These work. But at a high cost. They're often dumb (information flows badly through hierarchies because no one passes bad news up, and the top becomes a bottleneck); inefficient (participants waste their energy politicking against each other); and unpleasant.

In contrast, we want (and can start to seriously imagine being able) to retain our capacity for individualistic action while working together in loosely co-ordinated, ever changing but highly effective groups. The internet has taken us a long way in this direction, but as Brin points out, there's still room for more scientific understanding and other tools to take us further.