Working with language
This is more of a proposal of an explorative working topic than a sketch of a project.
I was getting interested into the impracticabilities that were introduced by the technical aspects of the medium that we used to speak to each other.
Designing a medium of communication in between two spaces by constraints in language, could be one topic of work. In particular, when using speech recognition software, a small and insufficient dictionary could come in place to recognize just some hundred words out of a pre-given story. Those words could then be used to trigger audio and video events, or even mechanical devices.
There is some speech recognition software API, that I managed to run on my computer
Sphinx4 (java based) (this is also available in C++)
As a starting point, I experimented with a small chat program that I manipulated. It now displays google keywords instead of the text that the user enters.
Please feel free to try it: Suggestive chat