What made Xerox PARC so successful?

One of the best books I read last year was Michael Hiltzik’s Dealers of Lightning. It’s about Xerox PARC, the incredibly influential, yet still mostly unknown, research division of Xerox. PARC developed some of personal computing’s most important technologies: Ethernet and modern GUIs are its best-known inventions, but there are dozens of other examples. PARC was also a very important talent incubator — Adobe and Pixar were founded by its alumni, while other employees went on to influential roles at Apple and Microsoft.

While reading, I tried to figure out what made PARC so successful, with an eye to how that success could be replicated today. Here’s what struck me as particularly important.

Contents

  1. Timing
  2. Exposure to outside ideas
  3. Network building
  4. Limited bureaucracy
  5. A social environment
  6. Why didn’t it work?
  7. Meta Reality Labs: the modern PARC?

All quotes are from the book and attributable to Hiltzik, unless otherwise specified.

Timing

PARC was founded in 1969, and I’m not sure it could have been founded at any other time. The late ’60s were “a buyer’s market for high-caliber research talent”: the vast expense of the Vietnam War had cut into government budgets, and a recession “[exerted] the same effect on corporate research”. That meant Xerox didn’t have to fight particularly hard to hire talented people.

Computing was also at “a historic inflection point”, making big leaps forward relatively easy to come by. A 2010s version of PARC might have had equally talented researchers producing equally excellent research, but the research wouldn’t be as revolutionary simply because the easy, game-changing problems had already been solved. (Patrick Collison talks about this a bit here — we’re mostly confronting “wicked problems” these days.)

“Historic inflection points” also attract brilliant people. Take Butler Lampson, who decided to abandon physics for computer science. “He found the task of advancing a science that history’s greatest intellects had been mining for 300 years fundamentally uninteresting. Especially when a brand-new field beckoned in which every new discovery represented a terrific leap forward in human enlightenment.”

Smart people like new things, both for the intellectual challenge and for the increased ease of fame and fortune. So you’ll often find brilliant people working on new, exciting fields where there’s a lot of opportunity for rapid progress — and because those fields attract brilliant people, the progress is indeed rapid. Tyler Cowen and Ezra Klein have noted that an awful lot of smart people are attracted to crypto right now; perhaps that’s a modern example of this kind of talent clustering.

Exposure to outside ideas

Serendipity was behind a lot of PARC’s success. It seems people would regularly hear about ideas by chance — at a conference, at a party, from an invited speaker — and decide to work on them as a result. As David Thornburg said, “we were physically adjacent to Stanford University, so there were visitors dropping in and out of the lab all the time.” This was intentional. Bert Sutherland, who managed one of PARC’s labs, had a policy “to keep [the lab’s] atmosphere enriched via continual contact with the outside world.” Learning what’s going on elsewhere can spark new ideas — and introduce you to potential solutions you’d never have come up with yourself.

PARC was also an important source of inspiration for other organisations, most famously Apple. Steve Jobs and a few other engineers toured PARC, which supposedly inspired many of the key features of Apple’s Lisa and Macintosh. According to Hiltzik, the story’s been warped over the years and Apple didn’t exactly pilfer PARC’s ideas. But the visit was still important: for Bill Atkinson, an Apple engineer, “seeing the overlapping windows on the Alto screen was … more a confidence-builder than a solution to his quandary.” In Atkinson’s words, “knowing it could be done empowered me to invent a way it could be done.”

Network building

While at ARPA, Bob Taylor — who went on to found PARC — realised that research was all about talent. “He would visit his grant recipients several times a year, but not solely to hear the researchers’ obligatory progress reports. He was engaged in something more like community outreach, developing new teams, nurturing up-and-coming young researchers, cultivating an entire new generation of virtuosi.” He also organised conferences, bringing researchers together in an attempt to forge connections (and expose them to new ideas).

This worked, both in encouraging research and in helping Taylor recruit. Because of the years he spent building a network, Taylor knew — and was trusted by — practically everyone. That made PARC a hub for many of the world’s most talented people.

Limited bureaucracy

PARC’s leaders hated bureaucracy, and employees had plenty of freedom to go off and pursue ideas. Its founders’ “experience had taught them that the only way to get the best research was to hire the best researchers they could find and leave them unburdened by directives, instructions, or deadlines.”

Some senior PARC leaders came from ARPA, which appears to have inspired this lack of bureaucracy. ARPA was rare among government agencies, in that it didn’t overburden staff with paperwork. Bob Taylor, founder of PARC and director of ARPA’s Information Processing Techniques Office, said he “had no formal proposals for the ARPANET”, one of the most important projects in human history. Here’s the book’s account of the meeting that greenlit ARPANET:

“After listening politely for a short time, Herzfeld interrupted Taylor’s rambling presentation … All he had was a question. ‘How much money do you need to get it off the ground?’ ‘I’d say about a million dollars or so, just to start getting organized.’ ‘You’ve got it,’ Herzfeld said. ‘That,’ Taylor remembered years later of the meeting at which the Internet was born, ‘was literally a twenty-minute conversation.'”

This lack of bureaucracy only worked because PARC’s managers were humble. The whole idea of a manager “approving” a project in some way implies that the manager knows more than the researcher what’s worthwhile — which, if you’re hiring the world’s best researchers, probably isn’t true. Instead, managers at PARC seem to have often just trusted researchers, accepting that they were smarter than their bosses. PARC manager Harold Hall, who said “I had developed and honed the skill of making myself useful to people whose intellectual gifts dwarfed my own”, epitomised this humility.

A social environment

PARC wasn’t just a workplace. It was a community of people that spent most of their life together. “There were family picnics and a softball team”, a “volleyball net for daily lunchtime matches”, and regular tennis matches, bike-rides and communal meals — all of which served as venues to talk about work while doing something other than work.

Alan Kay, whose lab was “socially … PARC’s most cohesive unit”, was a firm believer in the importance of these events. “He believed strongly that a lab’s success depended on a shared vision,” and spending so much time with other people helps to create that vision.

The sections of the book on in-person socialising significantly changed my views on remote work, and I’m now much less certain that revolutionary work can be done remotely. At the very least, it seems that regular team retreats are a necessity.

Why didn’t it work?

PARC was immensely innovative, but you could argue that it was ultimately a failure. You don’t own a Xerox laptop or use a Xerox phone, after all. Hiltzik offers many reasons for why Xerox failed to commercialise most of PARC’s technology, but one in particular sticks out.

In the ’70s, Xerox’s business was essentially based on microtransactions. They loaned copiers to businesses, and each time a customer copied a page, a meter on the machine “clicked” and Xerox was paid a small fee. Xerox’s salespeople’s commissions came from those fees.

This incentive model would not easily translate to computers: how do you charge customers for each ‘use’ of a computer? And without a per-use fee, there was no obvious revenue stream for the salespeople’s commissions. “To sell computers, Xerox would not only have to build a new kind of machine, but also a new system of compensating and motivating its more than 100,000 sales executives.” That proved difficult.

Not all of PARC’s problems can be blamed on its parent company, though. There were plenty of internal issues too. PARC founder Jack Goldman, for instance, “had little conception of the economics of product development.”

Perhaps more importantly, as PARC aged it developed a sort of sclerosis. Bob Taylor, seemingly out of misplaced confidence, ignored his colleagues’ advice that important things were happening at the Homebrew Computer Club and at Apple. “That’s not going to happen, because we have the smartest people here. I believe if you have the smartest people, you’ll end up ahead,” he reportedly said. Yet it turned out that some very smart people worked elsewhere, and they ended up leapfrogging PARC.

Meta Reality Labs: the modern PARC?

Writing in 1999, Hiltzik offered an explanation for why nothing like PARC existed anymore. One reason was the first thing we looked at: timing. “The science of computing is no longer at the historic inflection point it occupied at the start of the 1970s,” Hiltzik wrote.

That was certainly true in 1999. But today might be different. Most technologists seem fairly sure that we’re on the cusp of a new computing platform. It might not be called the metaverse and it might not be delivered through smartglasses, but the odds are good that we’ll soon be living in a hybrid and omnipresent digital world. This impending shift is much bigger than the PC-to-smartphone one, because it requires entirely new UX paradigms. Countless clunky demos show that we’re a long way from figuring out what those paradigms are.

Yet people are trying to figure it out — most notably, the people at Meta Reality Labs. Its researchers are working on new interaction mechanisms (haptic gloves and brain-computer interfaces), new operating systems, new manufacturing techniques, and all sorts of stuff I won’t even pretend to understand. This research is broad, experimental, and potentially revolutionary, much like PARC’s. In fact, Reality Labs may be the 21st century’s PARC.

The parallels aren’t perfect: for one thing, Reality Labs’ work is more obviously tied to Meta’s goals than PARC’s was to Xerox’s. But I’m not the only one to draw the link between the two. In an interview last year, Reality Labs head Michael Abrash said “[Doug] Engelbart and Xerox PARC are the only time that fundamentally the way we interact with the digital world has ever changed … AR glasses are going to require that to happen.” The next few years will show if, like PARC, Reality Labs can change the world — and if Meta, unlike Xerox, will be able to capitalise on it.

For its part, PARC still exists, albeit in a very different form. And its influence on humanity may not be over: PARC researchers are involved in the Marine Cloud Brightening Project, an effort to explore whether we can use geoengineering to help fight climate change.

You can buy Dealers of Lightning here; I highly recommend reading it.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s