First. We have to think small. As small as we can, and use the ideas to solve the problems we have--so that we can become more efficient. There's nothing worse then putting in a lot of effort into something for a result which is uncertain--and getting nothing.
The features I mentioned, I think...