The way to get started developing programming techniques for amorphous computing is to focus on some paradigm problems. We have already developed simulators that allow us to try out ideas with small numbers of computational elements, but many of the most interesting phenomena emerge only when dealing with huge numbers of elements.