The idea of using information bubbles as interfaces emerged in 2002 from two apparently unrelated groups of researchers virtually simultaneously. Groxis and Cloudmark, both in San Francisco, base their data navigation interfaces on the concept that information is really a kind of node, or bubble, and that related bubbles can be nested one inside another. These interfaces also rely on color cues based on the category, importance, or urgency of information to make navigation easier.

Groxis's version of the bubble interface, Grokker, has been available as stand-alone software for several years and was used only by a small community, before its recent release as a Java plug-in [see photo, "Grokking It"].

Give Grokker a hierarchical database and the software carves it into a series of colored balls, with each ball representing one topic. In the P&G search, these balls include news, market research, and brands. Look at a topic and you can zoom in on its contents, which are represented as smaller and smaller balls. For example, a search for P&G products turns up balls for "new products," "paper products," "care products," and "boycotted products." Go up in the hierarchy and you can view many balls and their relative relationships. Search for another topic and the information rearranges itself accordingly. Grokker can represent Web searches, corporate databases, or mixed data collections.

Cloudmark's bubble interface is called Information DNA. This interface grew out of the company's efforts to improve its e-mail antispam software. Cloudmark's engineers discovered that they could find patterns in large e-mail collections that are like DNA—that is, certain patterns represent various kinds of spam, such as fraudulent mortgage offers or pornography, that can be consistently identified. To help with the identification, the engineers created mathematical algorithms, which are applied to e-mail to produce a screen filled with bubbles whose sizes and colors indicate collections of safe information and of spam.

The third problem being tackled by new interfaces is organizing the information you create or collect. This includes photos, videos, and audio files along with traditional text documents—a much more complex task than the designers of the documents, folders, and menus interface tackled back in the 1970s.

The first product to address this challenge directly is EverNote, from EverNote Corp., Sunnyvale, Calif. The company made the PC version of the software available for free download in June and promises a Macintosh version soon. EverNote plans to charge a fee for an expanded version that will synchronize data from PDAs and smart phones along with computers.

The idea behind EverNote is simple: it keeps all your computer files—no matter what type or how they were created—in a single, chronological manuscript holding everything [see photo, "Time Traveler"]. You can search this manuscript by keywords, categories, or other designations. Or, probably most useful, you can look at graphical representations of your files on a timeline, based on when each item was obtained or created.

The graphics on the timeline are small versions of each document in its original format. That is, it is not represented by an icon. A document looks like a note or a letter, a handwritten note looks like handwriting, and a Web site looks like the actual Web site in miniature. People tend to remember more or less when they created or obtained a document—not the exact date, perhaps, but often the month or season. Because users also have a visual memory of that document, the graphical timeline enables them to select the right document quickly even though they can't read the text.

The basic idea behind EverNote is not new: David Gelernter of Yale University, in New Haven, Conn., implemented such a system in Lifestreams in the late 1990s; a version was released for the Apple Newton handheld computer in 1993. Lifestreams was a research project that kept all of an individual's data—notes, photos, manuscripts, Web-collected information—in a single permanent, chronological file. But Lifestreams converted the documents into a standard format, so they lost their visual uniqueness.

As the information being navigated and collected by computers becomes increasingly complex, it may turn out that two dimensions are not enough in which to represent it.

A number of companies have been working on 3-D interfaces for a long time. Such interfaces allow more flexibility in displaying information, permitting the images that represent information to look more natural, letting them rotate in space or overlap with transparency and dimension as clues to their position in space and size (and, therefore, their relationships or importance).

The first of these interfaces are 3-D representations translated to a 2-D display. A good example of this type of interface is Project Looking Glass, from Sun Microsystems Inc., in Santa Clara, Calif. [see photo, "Through the Looking Glass"]. It permits multiple documents to be viewed simultaneously, but instead of placing them as if they are standing straight up on the screen, as is typical in most interfaces today, Project Looking Glass tips the documents back, as if they are lying on a slanted drawing table. As a result, many more documents are visible at the same time than is possible in a traditional presentation. Project Looking Glass's display has perspective: documents toward the bottom of the screen are considered foreground and are larger; the user can push documents back, making them appear farther away and smaller but still visible. Documents can be "turned over," and the user can add notes on the back. This interface has been placed under an open-source license by Sun, and a development community is now gathering around it.

True 3-D interfaces are still very expensive, but they do exist. One example is the Perspecta Spatial 3D System, from Actuality Systems Inc., in Bedford, Mass. The system includes a 51-centimeter dome, which displays full-color and full-motion images that occupy a volume in space. This means the user can look at a 3-D image in 3-D space without special glasses and can interact naturally with the image in real time.

People who use special applications for medical research, radiation oncology, and petroleum exploration are the only ones who currently benefit from such true 3-D interfaces, but these devices may eventually migrate for use by ordinary computer users. Ravin Balakrishnan, a professor of computer science at the University of Toronto, is working on appropriate interfaces for such a world, perhaps based on the mapping of users' hand gestures and their interaction with virtual 3-D volumes.

In the future, user interfaces may go beyond the visual to the tactile. SensAble Technologies Inc., in Woburn, Mass., offers a device that looks like a pen attached to a robotic arm. It allows users to touch and manipulate virtual objects with varying degrees of precision, suiting the application and the user's budget. The idea is for the user to experience the object exactly as if he or she were touching it in the physical world. Force feedback via the haptic interface allows the user to explore virtual objects as if they were physically present and being explored by actual touch. SensAble has research projects with a number of commercial and university laboratories.

Although SensAble's interface is designed for high-end business users, haptic interfaces will quickly enter the consumer realm. In February, at Demo@15, an industry conference in Scottsdale, Ariz., Novint Technologies Inc., of Albuquerque, N.M., showed Falcon, a consumer haptic interface. Novint expects its interface to be used with games soon, at prices as low as US $99.

Such revolutionary new interfaces are steadily moving into users' hands. Some will catch on; most will fade away.

But none of the more exotic ideas discussed are mainstream yet: they are all experiments. The prize may go to the style of interface that gets adopted by a mainstream vendor, like Microsoft, Apple, or Google, and then proliferated to everyone's device. Or a tiny vendor that no one has ever heard of before may introduce an interface so seductive that no one can live without it. The evolution of the computer user interface is mainly still ahead of us.