Tech incubator PARC is still looking ahead after 40 years
Institutions, like human beings, often treat their 40th birthdays as occasions for midlife stocktaking. So it’s not surprising that the 40th anniversary celebration at Xerox PARClast week was devoted as much to looking ahead to the future as to looking back at its fabled history.
Xerox’s Palo Alto Research Center was founded in 1970, when the company was still flush with cash from the copier business but uneasy that new technologies might emerge to undercut its dominant position.
Over the next dozen years, the young scientists and engineers who came together on a hilltop overlooking what would soon be dubbed Silicon Valley invented the personal computer, Ethernet, the laser printer and Windows-style computer displays, among many other advances.
These were such dis-
ruptive technologies that Xerox had trouble commercializing them (although it made billions from the laser printer, among other products).
I’ve been a close observer ofPARC for more than a decade. That’s not only because it’s been that long since I wrote “Dealers of Lightning,” a book about the place, but because PARC has been an everevolving lesson in how corporations can and should manage R&D — not only the innovations that play to their core markets, but those they can’t exploit themselves.
It was that latter condition that provided the first lesson from PARC — a negative one. Xerox was unable to shift gears from the copier business to the PC market in part because in the 1970s there was no PC market to serve as a model. IBM eventually succeeded where Xerox failed, but it was a computer company, and even so shunted its PC business off to a small subsidiary, which happened to capture the zeitgeist.
Xerox watched PARC’s first generation of scientists and engineers walk out the door and seed pioneering projects at fledgling companies such as Microsoft (where PARC alumnus Charles Simonyi brought the what-you-see-is-whatyou-get word processing program that became Microsoft Word, in the process making a personal fortune sufficient to pay for two private voyages into space) and Apple (where Larry Tesler helped foster development of the Lisa and Macintosh).
It took years for Xerox to come around to a new business model, through which it would either absorb PARC innovations in its own product lines or allow PARC to shop them to outside customers or spin them off into joint ventures.
That process intensified after 2002, when Xerox converted PARC to a wholly owned subsidiary and encouraged its researchers to get involved in commercializing their work. (The center is now formally known as “PARC Inc.”)
“We brought in business development expertise and integrated it with the laboratory community,” PARC’s current chief executive, Mark Bernstein, told me. “The left brain is now complemented by right-brain thinking about what our technology is really useful for.”
Bernstein acknowledges that in contrast to the founding era, when staffers thought they could follow their instincts wherever they led, some projects are shelved if they can’t attract a commercial partner. “Sometimes we say, ‘That’s a great idea, but we can only take it to this stage for the time being.’ ”
Results from that process were on display last week during a daylong series of demos and panel discussions tied to the anniversary.
One PARC spinoff, Powerset Inc., was formed to make use of natural language technologies to allow users to do Web searches based on phrases, not just keywords. Xerox retained an equity stake and cashed it in when Powerset was bought for more than $100 million in 2008 by Microsoft, which incorporates the technology in its Bing search engine.
Other spinoffs and companies still being incubated are working with PARC innovations in solar technology, water treatment and networking.
PARC researchers say the process of seeking out customers for their work is oddly liberating.
“At first, we had one customer [Xerox] and one goal,” said Danny Bobrow, an artificial intelligence pioneer who got to PARC in 1972 and is still there as a research fellow. “Now, I can look around for sponsorship, and Xerox supports that.”
Yet it’s an article of faith at PARC that the institution is still unique as a wellspring of basic research owned by a corporation.
“Corporate research is past its glory days,” said William J. Spencer, who as PARC’s director in the mid-1980s established the model of spinning off its technologies as independent companies or with outside partners. “You don’t have monopolies like Xerox and AT&T had, so it’s hard to justify setting aside 2, 3 or 5% of revenues for basic research.”
I asked Spencer, who had worked at Bell Labs and Sandia National Laboratory before joining Xerox, how the country could take up the slack left by the refocusing of corporate research.
“I don’t think we’ve addressed that problem,”
‘We brought in business development expertise and integrated it with the laboratory community. The left brain is now complemented by right-brain thinking.’
PARC’s chief executive
he replied. “We’ve relegated basic research to the universities, but they’re beginning to focus on short-term projects too — faculty members are asking, ‘How do I get tenure?’ ”
Yet the landscape of neither corporate revenue nor corporate research is monolithic. Although Hewlett-Packard, to mention one technology giant, cut its research and development spending under the recently departed CEO Mark Hurd to 2.5% of revenue last year from 6% in 2002, IBM is still spending 6% ($5.8 billion devoted to R&D last year).
Some companies today dominate their markets nearly as decisively as Xerox and Ma Bell did back when. They just aren’t making the same commitment to broad research.
In 1970, the year Xerox conceived PARC, it recorded sales of $1.7 billion and net income of $188 million — that’s a profit of $1.1billion in today’s dollars.
For comparison’s sake, Google earned a $6.5-billion profit on $24 billion in sales last year. The company’s R&D spending of $2.8 billion is 12% of its revenue, although there’s a bit of a cheat there — about 25% of that spending is “stockbased compensation” for the researchers, which isn’t the same as a cash outlay.
While Google Labs turns out some terrific ideas, I’m not sure that anyone would claim that its researchers are pointing the way toward an entirely novel market, as did PARC’s founding generation.
It may be unfair to today’s corporate behemoths to paint the Xerox of the ’70s as an altruistic backer of random innovation. The company founded PARC with the directive to develop the “office of the future,” which Xerox hoped to exploit as a core market. The concept just happened to accommodate almost anything that could be conceived of in electrical engineering or the embryonic field of computer science in that era.
The one thing that hasn’t changed about PARC in 40 years is that it trusts that innovation will always find customers, even if not within Xerox. It’s almost always the case that “you can find someone who has an interest in what you’re doing,” Bobrow told me. And if not?
“The wonderful thing about the world,” he said, “is that it’s rich in ideas and there are lot of things to pursue.”
RESEARCHER: Ed Chi wears an eye-tracking device as he looks at Web pages at PARC in 2001.