People will form small teams of three people. One person on each team will be the "avatar". The avatar will be equipped with bud-earphones. Through the use of sound localizing synthesis, he will be able to hear the voices of his teammates - one from over his left shoulder, one from over his right shoulder. The other two team members will be in another room. One of the other team members will be the "searcher." The searcher's responsibility is to search the Web, or other sources, for information that the avatar might need. The searcher verbally informst the avater of these results. The third team member is the "artist." The artist's responsibility is to create text, images, drawings, or anything else that may be visually needed by the avatar. This displayed information is projected into the room where the avatar is located.

There will be two or more such teams at the same time. The teams will cooperate to solve some task. Only the avatars can directly see each other; they form a small work group of "enhanced participants." They will each have information at their fingertips (thanks to their searcher) and they will each be able to call up visual aids (thanks to their artist).

In a more advanced variation, the artist makes sounds appear to the avatars as though coming from different places or objects in the room. Localized sound synthesis and tracking of the avatars' positions within the room are required to effect this.

Over the course of several weeks, we will construct a story, to be told within a single very high resolution photograph of an office: a desk and bulletin board. Clues are contained in objects within this large photograph, such as writing in open yearbooks, matchbooks with scrawled phone numbers snapshots of people seen together, post-it notes tacked to the bulletin board, entries in a wall or desk calendar, coffee mugs, etc.

The photograph will be digitized to a very high resolution (25000 × 15000 pixel) image. A program to zoom and pan through this image in real-time can be used to "read" the story.

The goal here is to learn how to construct a narrative visually.

This is a variation on Charles Eames' falling-ball chimes machine (in which a ball, which tumbles down through the mechanism, hits different chimes in succession, thereby playing a melody). Instead of the ball following only one path, we will allow it to get to "branching points", where it can fall left or right. Depending upon the ball's path, different things happen in the course of playing the piece. Chords may be inverted, or arpeggiated, or a blue note may be struck.

To physically build such an instrument would be a major undertaking, and it would be very difficult to tune, and to find out the proper variations while you're building it. So we will build a software simulator, which lets users construct different variations of a stochastic Eames chimes machine, and to hear interactively the sorts of tunes their variation will play.

A community of people will use persistent applet software that you write, to build something collectively. It can be a cathedral, rooms in a house, a non-linear screenplay, a race-track, a virtual tinkertoy construction, a giant scrabble board, or a puppet show.

You need to think of what you would like to empower people to build. It should be something fun, interesting, and compelling, that will bring out people's talents and interests as they play with it.