ABSTRACT
As with the evolution of computer displays from character terminals to bitmapped graphics, the transition to 3D graphics offers a unique opportunity to rethink the human-computer interface. Much has changed since the inception of the 2D windowing interface, notably the rise of the Internet, with all its data and connectivity to people and machines, and the abundance of CPU power, ever increasing, as predicted by “Moore’s Law.”
By rethinking the boundary between application and data, we can move information out of silos into a common sandbox, where it can be freely morphed and amended in a mash-up of information that better fits the messy business of human sense-making and understanding. By rethinking the user-interface, we can raise the level of human-computer interaction from one-way direct-manipulation to bi-directional visual discourse—making computers smarter to make us smarter.
In this paper we examine an implementation of this philosophy. A specialized 3D virtual world is the collaborative, graphical front end to a highly-expressive data layer, where disparate information is collected into a single place for use by users and software agents acting on their behalf. Usage models depicted in Part I are decomposed into system components in Part II, with Part III looking at some of the many challenges remaining.
