A walk in the Parc

Headnote:

With a new generation of breakthroughs on tap, Xerox's legendary lab aims to be an IS pioneer once again

(Photograph Omitted)

Captioned as: Xerox Parc director John Seely Brown: "Technology has finally gotten powerful enough; maybe it's high time it gets the hell out of our way"

But in one of the biggest blunders in corporate history, Xerox Corp. gave away its user interface technology to Apple Computer, Inc. Xerox also let other killer technologies slip away to other companies in the Valley. Meanwhile, its attempts to market a computer flopped.

Now, after a long, quiet period, the legendary Xerox Parc is again in a position to shape the future of computing, some observers say. "During the 1980s, it was conventional to say that Parc was no longer relevant, but I believe that is false today," says Eric Schmidt, chairman and CEO of Novell, Inc. "The new generation of researchers is doing innovative work on the foundation of the original founders of Parc."

Schmidt cites Xerox Parc's "hyperbolic tree" - a quasi-threedimensional browsing and display technology that offers comprehensive views of complex data - as an example of what might emerge as the next-generation GUI (see story below). "What's the No. I problem on the 'net?" Schmidt asks. "You can't find things. These guys are ahead of everybody else."

Despite having invented much of it, Xerox now disparages today's GUIs as "the WIMP paradigm" Windows, Icons, Menus, Point and click. Those GUIs wimp out against huge, complex data hierarchies such as the World Wide Web, Xerox Parc says.

BRANCHING OUT

The hyperbolic tree is among the first in a planned family of patented components that Xerox Parc calls "wide widgets." Wide widgets supplement the traditional point-andclick with a broader see-and-go approach in which users can take stock of complex stores of information and easily focus on areas of interest before clicking on an item.

The recent Xerox spin-off InXight licenses the widgets to application developers - such as Microsoft Corp., Oracle Corp. and the major Internet search engine companies - to use in their user interfaces.

Wide widgets borrow from techniques in science and modeling. "We saw that we could build office software, not just scientific visualization things that are big and fast and hard to understand like clouds and proteins," Rao says. "Documents on a network are also big and fast and hard to understand."

Xerox Parc still sounds the mantra of usability and simplicity, as it did 25 years ago when it brought together the WIMP components.

"The biggest problem in IT today is that technology has finally gotten powerful enough; maybe it's high time it gets the hell out of our way," says John Seely Brown, director of Xerox Parc and chief scientist at Xerox. "You want to be able to reach right through the technology, as if it's not there, right to the work you want to do."

According to Brown, 90% of the human brain is devoted to processing sensory information and 20% to cognitive or thinking tasks. In computers, that ratio is reversed - to the detriment of users, he says. Tools such as the hyperbolic browser leverage the senses, he says.

At Xerox Parc, inventing tools to aid usability isn't just an exercise in computer science. Getting technology to disappear "requires a new set of eyeglasses," Brown says. "We've learned from business process reengineering that descriptions of how work gets done are almost always dead wrong."

For that reason, Xerox Parc employs four anthropologists whose job is to study -- often in mindnumbing detail - how workers work. In a landmark study for San Jose International Airport in California several years ago, Xerox Parc and Steelcase Corp. interviewed and videotaped ground operations workers for two years. That scrutiny led to the discovery that a critical online scheduling system couldn't be viewed in isolation. To improve the system required an understanding of feeds from telephones, radio and paper documents and how those media were used.

"The Parc people - and particularly project leader Lucy Suchman - did the most wonderful job of teasing apart [office work practices] and got more insight into them than anyone in the world at that point," says Larry Keeley, president of Doblin Group, a Chicago-based design consultancy that worked on the airport project. "Xerox products are materially more effective and easy to understand because of the way they do operability analysis and the way they understand the social and cultural affects of new technology," Keeley says.

Suchman, an anthropologist who heads Xerox Parc's Work Practice & Technology area, urges information technology researchers to bring a combination of "deep understanding and imagination" to their studies of work practices. "People in research really underestimate the amount of innovation going on in the workplace," Suchman says. "The world is much more interesting than we ever imagined."

NEW METHODS

For a law firm document-retrieval system, Suchman studied how a lawyer located documents in his file cabinet. "It's a perfectly banal activity, and you might think you can rely on your own experience," she says. But she found that the lawyer often located documents not by reading titles but by noting physical features. The resulting system uses electronic folders that display thumbnail images of documents' first pages that can't be read but that look like their physical counterparts.

Xerox calls itself "the document company," but it defines "document" much more broadly than its core businesses in copying and printing would suggest. The document is "a carrier of information around which knowledge gets created, captured, consolidated and communicated," Brown says. Any technology that might support those functions is fair game for Xerox Parc's $60 million research budget. "We do everything from atoms to culture here, from inventing fundamentally new types of display technology to engaging anthropologists to look at where value really gets created in the office," Brown says. Asked which of his many projects might have an especially large impact on the workplace, Brown cites Xerox Parc's work in "smart matter," where computer science and material science converge.

In one smart-matter project, microscopic sensors and actuators are distributed along a metal beam. As the beam is stressed, computers feed instructions to counteract the stress forces to the actuators. Such an arrangement might keep a building intact during an earthquake.

Xerox will never again give away its technology, Brown vows. "In the I970s, creating patents was an afterthought, and defending patents was an after-afterthought," he says. "But now we take intellectual property deadly seriously."

Rao acknowledges a certain ambivalence about that strict posture. "It's a little bit of a conundrum. If the genie had been bottled up and held tight, there wouldn't be all that stuff now. So from that perspective, it's been a glamorous success," he says. But if companies never reaped financial rewards from their research, they would cease to fund it, he adds.

Asked how to reconcile the goals of wide dissemination of technology with protection of shareholder interests, Rao says, "We have to provide business value, and bring pricing down. We have to have five carrots in the hand and the patent stick in the back pocket."

Author Affiliation: