The Innovators

by Walter Isaacson

  With his low-key personality and even lower-key posting, Berners-Lee did not fathom what a profound idea he had unleashed. Any information anywhere. “I spent a lot of time trying to make sure people could put anything on the web,” he said more than two decades later. “I had no idea that people would put literally everything on it.”32 Yes, everything. Enquire Within Upon Everything.

  MARC ANDREESSEN AND MOSAIC

  For people to summon forth sites on the Web, they needed a piece of client software on their own computers that became known as a browser. Berners-Lee wrote one that could both read and edit documents; his hope was that the Web would become a place where users could collaborate. But his browser worked only on NeXT computers, of which there were few, and he had neither the time nor the resources to create other browser versions. So he enlisted a young intern at CERN, an undergraduate named Nicola Pellow who was majoring in math at Leicester Polytechnic, to write the first all-purpose browser for UNIX and Microsoft operating systems. It was rudimentary, but it worked. “It was to be the vehicle that allowed the Web to take its first tentative step on to the world stage, but Pellow was unfazed,” Cailliau recalled. “She was given the task and she simply sat down to do it, little realizing the enormity of what she was about to unleash.”33 Then she went back to Leicester Polytechnic.

  Berners-Lee began urging others to improve on Pellow’s work: “We energetically suggested to everyone everywhere that the creation of browsers would make useful projects.”34 By the fall of 1991 there were a half-dozen experimental versions, and the Web quickly spread to other research centers in Europe.

  That December it made the leap across the Atlantic. Paul Kunz, a particle physicist at the Stanford Linear Accelerator Center, was visiting CERN, and Berners-Lee recruited him to the world of the Web. “He twisted my arm and insisted that I come see him,” according to Kunz, who worried that he was in for a boring demonstration of information management. “But then he showed me something that opened my eyes.”35 It was a Web browser on Berners-Lee’s NeXT calling up information from an IBM machine somewhere else. Kunz brought the software back with him, and http://slacvm.slac.stanford.edu/ became the first Web server in the United States.

  * * *

  The World Wide Web hit orbital velocity in 1993. The year began with fifty Web servers in the world, and by October there were five hundred. One reason was that the primary alternative to the Web for accessing information on the Internet was a sending and fetching protocol developed at the University of Minnesota called Gopher,I and word leaked out that the developers were planning to charge a fee for use of the server software. A more important impetus was the creation of the first easy-to-install Web browser with graphic capabilities, named Mosaic. It was developed at the National Center for Supercomputing Applications (NCSA) at the University of Illinois at Urbana-Champaign, which had been funded by the Gore Act.

  The man, or overgrown kid, most responsible for Mosaic was a gentle but intense undergraduate named Marc Andreessen, a corn-fed six-foot-four jolly giant born in Iowa in 1971 and raised in Wisconsin. Andreessen was a fan of the pioneers of the Internet, and their writings inspired him: “When I got a copy of Vannevar Bush’s ‘As We May Think,’ I said to myself, ‘Yep, there it is! He figured it out!’ Bush envisioned the Internet as fully as you could, given that you didn’t have digital computers. He and Charles Babbage are in the same league.” Another hero was Doug Engelbart. “His lab was node four on the Internet, which was like having the fourth telephone in the world. He had the amazing foresight to understand what the Internet would be before it got built.”36

  When Andreessen saw the Web demonstrated in November 1992, he was blown away. So he enlisted an NCSA staffer, Eric Bina, a first-class programmer, to partner with him in building a more exciting browser. They loved Berners-Lee’s concepts, but they thought CERN’s implementation software was drab and devoid of cool features. “If someone were to build the right browser and server, that would be really interesting,” Andreessen told Bina. “We can run with this and really make it work.”37

  For two months they engaged in a programming binge that rivaled those of Bill Gates and Paul Allen. For three or four days straight they would code around the clock—Andreessen fueled by milk and cookies, Bina by Skittles and Mountain Dew—and then crash for a full day to recover. They were a great team: Bina was a methodical programmer, Andreessen a product-driven visionary.38

  On January 23, 1993, with just a little more fanfare than Berners-Lee had indulged in when launching the Web, marca@ncsa.uiuc.edu announced Mosaic on the www-talk Internet newsgroup. “By the power vested in me by nobody in particular,” Andreessen began, “alpha/beta version 0.5 of NCSA’s Motif-based networked information systems and World Wide Web browser, X Mosaic, is hereby released.” BernersLee, who was initially pleased, posted a response two days later: “Brilliant! Every new browser is sexier than the last.” He added it to the growing list of browsers available for download from info.cern.ch.39

  Mosaic was popular because it could be installed simply and enabled images to be embedded in Web pages. But it became even more popular because Andreessen knew one of the secrets of digital-age entrepreneurs: he fanatically heeded user feedback and spent time on Internet newsgroups soaking up suggestions and complaints. Then he persistently released updated versions. “It was amazing to launch a product and get immediate feedback,” he enthused. “What I got out of that feedback loop was an instant sense of what was working and what wasn’t.”40

  Andreessen’s focus on continual improvement impressed Berners-Lee: “You’d send him a bug report and then two hours later he’d mail you a fix.”41 Years later, as a venture capitalist, Andreessen made a rule of favoring startups whose founders focused on running code and customer service rather than charts and presentations. “The former are the ones who become the trillion-dollar companies,” he said.42

  There was something about Andreessen’s browser, however, that disappointed and then began to annoy Berners-Lee. It was beautiful, even dazzling, but Andreessen’s emphasis was on enabling rich media for publishing eye-catching pages, and Berners-Lee felt that the focus should instead be on providing tools that would facilitate serious collaboration. So in March 1993, after a meeting in Chicago, he drove “across the seemingly interminable cornfields” of central Illinois to visit Andreessen and Bina at NCSA.

  It was not a pleasant session. “All of my earlier meetings with browser developers had been meetings of minds,” Berners-Lee recalled. “But this one had a strange tension to it.” He felt that the Mosaic developers, who had their own public relations staff and were garnering a lot of publicity, were “attempting to portray themselves as the center of Web development and to basically rename the Web as Mosaic.”43 They seemed to be trying to own the Web, he thought, and perhaps profit from it.II

  Andreessen found Berners-Lee’s recollection amusing. “When Tim came, it was more of a state visit than a working session. The Web had already become a brush fire, and he was uncomfortable that he was no longer controlling it.” Berners-Lee’s opposition to embedding images struck him as quaint and purist. “He only wanted text,” Andreessen remembered. “He specifically didn’t want magazines. He had a very pure vision. He basically wanted it used for scientific papers. His view was that images are the first step on the road to hell. And the road to hell is multimedia content and magazines, garishness and games and consumer stuff.” Because he was customer-focused, Andreessen thought that this was academic hogwash. “I’m a Midwestern tinkerer type. If people want images, they get images. Bring it on.”44

  Berners-Lee’s more fundamental criticism was that by focusing on fancy display features, such as multimedia and ornamental fonts, Andreessen was ignoring a capability that should have been in the browser: editing tools that would allow users to interact with and contribute to the content on a Web page. The emphasis on display rather than editing tools nudged the Web into becoming a publishing platform for people who had servers ra
ther than a place for collaboration and shared creativity. “I was disappointed that Marc didn’t put editing tools in Mosaic,” Berners-Lee said. “If there had been more of an attitude of using the Web as a collaborative medium rather than a publishing medium, then I think it would be much more powerful today.”45

  Early versions of Mosaic did have a “collaborate” button, which allowed users to download a document, work on it, and repost it. But the browser was not a full-fledged editor, and Andreessen felt it was impractical to turn it into one. “I was amazed at this near-universal disdain for creating an editor,” complained Berners-Lee. “Without a hypertext editor, people would not have the tools to really use the Web as an intimate collaborative medium. Browsers would let them find and share information, but they could not work together intuitively.”46 To some extent, he was right. Despite the astonishing success of the Web, the world would have been a more interesting place if the Web had been bred as a more collaborative medium.

  Berners-Lee also paid a visit to Ted Nelson, who lived on a houseboat in Sausalito in the shadow of the Golden Gate Bridge. Twenty-five years earlier, Nelson had pioneered the concept of a hypertext network with his proposed Xanadu project. It was a pleasant meeting, but Nelson was annoyed that the Web lacked key elements of Xanadu.47 He believed that a hypertext network should have two-way links, which would require the approval of both the person creating the link and the person whose page was being linked to. Such a system would have the side benefit of enabling micropayments to content producers. “HTML is precisely what we were trying to prevent—ever-breaking links, links going outward only, quotes you can’t follow to their origins, no version management, no rights management,” Nelson later lamented.48

  Had Nelson’s system of two-way links prevailed, it would have been possible to meter the use of links and allow small automatic payments to accrue to those who produced the content that was used. The entire business of publishing and journalism and blogging would have turned out differently. Producers of digital content could have been compensated in an easy, frictionless manner, permitting a variety of revenue models, including ones that did not depend on being beholden solely to advertisers. Instead the Web became a realm where aggregators could make more money than content producers. Journalists at both big media companies and little blogging sites had fewer options for getting paid. As Jaron Lanier, the author of Who Owns the Future?, has argued, “The whole business of using advertising to fund communication on the Internet is inherently self-destructive. If you have universal backlinks, you have a basis for micropayments from somebody’s information that’s useful to somebody else.”49 But a system of two-way links and micropayments would have required some central coordination and made it hard for the Web to spread wildly, so Berners-Lee resisted the idea.

  * * *

  As the Web was taking off in 1993–94, I was the editor of new media for Time Inc., in charge of the magazine company’s Internet strategy. Initially we had made deals with the dial-up online services, such as AOL, CompuServe, and Prodigy. We supplied our content, marketed their services to our subscribers, and moderated chat rooms and bulletin boards that built up communities of members. For that we were able to command between one and two million dollars in annual royalties.

  When the open Internet became an alternative to these proprietary online services, it seemed to offer an opportunity to take control of our own destiny and subscribers. At the April 1994 National Magazine Awards lunch, I had a conversation with Louis Rossetto, the editor and founder of Wired, about which of the emerging Internet protocols and finding tools—Gopher, Archie, FTP, the Web—might be best to use. He suggested that the best option was the Web because of the neat graphic capabilities being built into browsers such as Mosaic. In October 1994 both HotWired and a collection of Time Inc. websites launched.

  At Time Inc. we experimented with using our established brands—Time, People, Life, Fortune, Sports Illustrated—as well as creating a new portal named Pathfinder. We also conjured up new brands, ranging from the Virtual Garden to the Netly News. Initially we planned to charge a small fee or subscription, but Madison Avenue ad buyers were so enthralled by the new medium that they flocked to our building offering to buy the banner ads we had developed for our sites. Thus we and other journalism enterprises decided that it was best to make our content free and garner as many eyeballs as we could for eager advertisers.

  It turned out not to be a sustainable business model.50 The number of websites, and thus the supply of slots for ads, went up exponentially every few months, but the total amount of advertising dollars remained relatively flat. That meant advertising rates eventually tumbled. It was also not an ethically healthy model; it encouraged journalists to cater primarily to the desires of their advertisers rather than the needs of their readers. By then, however, consumers had been conditioned to believe that content should be free. It took two decades to start trying to put that genie back in the bottle.

  In the late 1990s Berners-Lee tried to develop a micropayments system for the Web through the World Wide Web Consortium (W3C), which he headed. The idea was to devise a way to embed in a Web page the information needed to handle a small payment, which would allow different “electronic wallet” services to be created by banks or entrepreneurs. It was never implemented, partly because of the changing complexity of banking regulations. “When we started, the first thing we tried to do was enable small payments to people who posted content,” Andreessen explained. “But we didn’t have the resources at the University of Illinois to implement that. The credit card systems and banking system made it impossible. We tried hard, but it was so painful to deal with those guys. It was cosmically painful.”51

  In 2013 Berners-Lee began reviving some of the activities of the W3C’s Micropayments Markup Working Group. “We are looking at micropayment protocols again,” he said. “It would make the Web a very different place. It might be really enabling. Certainly the ability to pay for a good article or song could support more people who write things or make music.”52 Andreessen said he hoped that Bitcoin,III a digital currency and peer-to-peer payment system created in 2009, might turn out to be a model for better payment systems. “If I had a time machine and could go back to 1993, one thing I’d do for sure would be to build in Bitcoin or some similar form of cryptocurrency.”53

  * * *

  We at Time Inc. and other media companies made one other mistake, I think: we abandoned our focus on creating community after we settled into the Web in the mid-1990s. On our AOL and CompuServe sites, much of our effort had been dedicated to creating communities with our users. One of the early denizens of The WELL, Tom Mandel, was hired to moderate Time’s bulletin boards and emcee our chat rooms. Posting articles from the magazine was secondary to creating a sense of social connection and community among our users. When we migrated to the Web in 1994, we initially tried to replicate that approach. We created bulletin boards and chat groups on Pathfinder and pushed our engineers to replicate AOL’s simple discussion threads.

  But as time went on, we began to pay more attention to publishing our own stories online rather than creating user communities or enabling user-generated content. We and other media companies repurposed our print publications into Web pages to be passively consumed by our readers, and we relegated the discussions to a string of reader comments at the bottom of the page. These were often unmoderated rants and blather that few people, including us, ever read. Unlike the Usenet newsgroups or The WELL or AOL, the focus was not on discussions and communities and content created by users. Instead, the Web became a publishing platform featuring old wine—the type of content you could find in print publications—being poured into new bottles. It was like the early days of television, when the offerings were nothing more than radio shows with pictures. Thus we failed to thrive.

  Fortunately, the street finds its own uses for things, and new forms of media soon arose to take advantage of the new technology. Led by the growth of blogs and wikis, both
of which emerged in the mid-1990s, a revitalized Web 2.0 arose that allowed users to collaborate, interact, form communities, and generate their own content.

  JUSTIN HALL AND HOW WEB LOGS BECAME BLOGS

  As a freshman at Swarthmore College in December 1993, Justin Hall picked up a stray copy of the New York Times in the student lounge and read a story by John Markoff about the Mosaic browser. “Think of it as a map to the buried treasures of the Information Age,” it began. “A new software program available free to companies and individuals is helping even novice computer users find their way around the global Internet, the network of networks that is rich in information but can be baffling to navigate.”54 A willowy computer geek with an impish smile and blond hair flowing over his shoulders, Hall seemed to be a cross between Huck Finn and a Tolkien elf. Having spent his childhood in Chicago dialing into computer bulletin boards, he immediately downloaded the browser and began surfing. “The whole concept blew me away,” he remembered.55

  Hall quickly realized something: “Nearly all of the online publishing efforts were amateur, people who didn’t have anything to say.” So he decided to create a website, using an Apple PowerBook and MacHTTP software he downloaded for free, that would amuse himself and others who shared his cheeky outlook and teenage obsessions. “I could put my writings and words up electronically, make them look pretty, and engage the web with links.”56 He got his site up in mid-January 1994, and a few days later, to his delight, strangers from around the Web began to stumble across it.