"The point is, if the history of technology teaches us anything, it teaches us

that we are extremely bad at predicting ex ante the uses of technology.

This fact has an overlooked, but absolutely vital design corollary;

wherever possible, design the system to run with open content, on open protocols, to be potentially available to the largest possible number of users, and to accept the widest possible range of experimental modifications from users who can themselves determine the development of the technology.

Then sit back and wait to see what emerges.

It may be that your predictions of how the technology will be used, and even your predictions about the potential user group, will be completely wrong."