Breakpoint_Why the Web will Implode, Search will be Obsolete, and Everything Else you Need to Know about Technology is in Your Brain

by Jeff Stibel

  Our intelligence comes from all of those little charges in our head going on and off in a constant stream, eventually leading to actions that define us. We are greater than the sum of our parts: a result of swarm intelligence, a collective consciousness. It’s like a series of dominoes that fall only to reveal a pattern. It’s an unbelievable idea when you think about it.

  II

  Crowdsourcing operates under the assumption that the sum is greater than the parts. Oftentimes this is true, especially in matters where public opinion is important. Advertisers have been using focus groups and surveys for decades—in such instances, the wisdom of the crowd is considered sacred. Complex technical problems have also found solutions through the crowd; open source software like Linux and Apache, for example, is dependent upon crowdsourcing. Crowdsourcing was even used by the British government in 1714 when it held a contest offering £20,000 to whoever could come up with a vacuum-sealed pocket watch (the result was the first marine chronometer).

  In the past ten years, crowdsourcing has expanded into new territories. We’re relying on crowds to an extent that could not have been predicted even a few decades ago. King Louis XV was incredulous that a crowd could write a poem; imagine his reaction if you told him that we created a 22-million entry encyclopedia with no author.

  That encyclopedia, Wikipedia, is the first massive-scale online crowdsourcing platform. More than just a digital reference book, it’s a living entity, simultaneously creating and surviving in its own ecosystem of drafts, edits, and interpretive processes. In 2012, Wikipedia contained over 22 million articles in 285 languages (only 4 million of the articles are in English). The entire process is free: its army of 77,000 contributors and editors are all volunteers; its legion of almost half a billion monthly users pays nothing to use the site. It is bigger than the print versions of Britannica, Cambridge, and Americana combined and has put all three out of business permanently. It is truly a global information amalgamation, and it has no author.

  Wikipedia had a meteoric growth stage. It grew rapidly from its birth in 2001 to age five, with the number of articles rising exponentially as the total number of words approached two billion (in contrast, the entire Encyclopedia Britannica was only about 50 million words). But lately, life hasn’t been easy for this global knowledge network. It is currently growing at less than half the rate of 2006 in terms of new content. Founder Jimmy Wales told the press in 2011 that the site was losing contributors and having trouble attracting new ones.

  Back in 2007, the network hit a breakpoint and growth began to slow. Contributors and edits peaked, as did the number of new articles, at a staggering 60,000 per month. The site had over 10,000 new editors joining every month, but by 2008, Wikipedia was losing 1,500 per month, and that number jumped to negative 15,000 by 2009. The network itself was starting to crack.

  Of course, part of the reason for the decline is surely that there is only so much content to record, and the low-hanging fruit has already been picked. In the early days, a Beatles enthusiast would have been thrilled to contribute information to build the pages of Paul, George, John, and Ringo, and perhaps even a page for Yoko Ono. Now, Wikipedia is so comprehensive that there simply isn’t much historical information left to provide.

  In many ways, this is analogous to the amount of carrying capacity available. Wikipedia peaked and, in doing so, exceeded its carrying capacity. In 2007, the number of editors and new pages hit their breakpoint and began a quick, sharp decline. Since then, the Wikipedia growth curve looks exactly like our breakpoint model.

  Image 7.1: The Breakpoint of Wikipedia

  Despite the concern of Wikipedia’s founder, the decline led to stabilization, just as we have seen with all other networks that naturally hit a breakpoint. In spite of the content decline, usage increased, and Wikipedia remains the sixth most visited website in the world. This is despite the onslaught of Google, Facebook, Twitter, and the many other up-and-comers. Smaller, slower growth in content may just be Wikipedia’s secret weapon.

  Growth is particularly important to a crowdsourced network like Wikipedia. Crowds need critical mass. Ten or even 1,000 articles won’t drive anyone to a site (imagine a dictionary with only 1,000 entries); rapid growth is necessary to become a universal resource. However, there’s a point where maintaining hypergrowth in terms of quantity decreases the usefulness of the site. Wikipedia has already reached this point, and now it must focus on quality, not quantity.

  Yet in 2011, the Wikimedia Foundation released a five-year strategic plan that made clear that their main focus was growth. By 2015 they hope to more than double Wikipedia’s size to 50 million total articles and 200,000 active contributors per month. They even started hiring recruiters to bring people to the free, nonprofit crowdsourced site. What could be more contrary to the concept of a natural crowd than recruiting? To be sure, there is plenty of room for Wikipedia to grow in different languages—there are a mere 77,000 articles in Thai and 11,000 in Yiddish. But to force growth in terms of Wikipedia’s articles and contributors goes against everything we know about the breakpoint of networks.

  Consider the value of a print encyclopedia. Sadly for many, Encyclopaedia Britannica stopped the presses after printing its final edition in 2010, after 244 years of annual publishing. When asked about the threat of Wikipedia, Britannica’s president Jorge Cauz explained that his encyclopedias were very different from Wikipedia, the former set apart by “its prestigious sources, its carefully edited entries and the trust that was tied to the brand.” There is plenty of truth to that statement, but Cauz underestimated the intelligence of the crowd. Wikipedia inches closer toward the Britannica model each time the crowd corrects an error or deletes a page. And of course, Wikipedia has huge advantages in that it doesn’t weigh 129 pounds or cost almost $1,400.

  Since 2006, commentators have been saying that Wikipedia is dying. Mathias Schindler, a board member of Wikipedia in Germany, recently echoed this sentiment: “The number one headline I have been seeing for five years is that Wikipedia is dying.” Really, it is improving in quality and reaching equilibrium, and it must be allowed to do so. Crowdsourcing is great for rapid growth, and perhaps great for spreading anti-government poetry, but not always so fantastic for improving quality.

  This was Cauz’s main insight, but what he may not have fully understood is that size matters first for a crowdsourced network. Wikipedia’s hypergrowth allowed it to temporarily leapfrog the old model of quality by offering something where there was nothing. Compared to Wikipedia, the printed Encyclopaedia Britannica was simply missing too much information. Interestingly, Jorge Cauz has since reinvented Encyclopaedia Britannica for the digital age by navigating these waters. The company’s online encyclopedia is updated continuously and accepts user input that is vetted by expert editors. This has led to record profitability.

  Wikipedia could learn from Encyclopaedia Britannica’s new model. To grow in quality, the crowd must subside. As Carnegie Mellon professor Aniket Kittur put it, “People generally have this idea that the wisdom of crowds is a pixie dust that you sprinkle on a system and magical things happen. Yet the more people you throw at a problem, the more difficulty you are going to have with coordinating those people. It’s too many cooks in the kitchen.” Once you have sufficient growth, pulling back paradoxically drives momentum forward.

  III

  Wikipedia is habitually considered the first crowdsourced reference tool, but another, more prominent example predates it. In the late 1970s, Professor James Murray was tasked by the Oxford University Press with a unique opportunity: to create a dictionary that “by the completeness of its vocabulary, and by the application of the historical method to the life and use of words, might be worthy of the English language and of English scholarship.” The idea was interesting in that most dictionaries at the time largely ignored colloquial and scientific words, and nearly all of them left out the history of words.
>
  Oxford’s proposal presented a number of challenges for Murray. First, there were already a number of English dictionaries available, so a new one would need to be unique. Second, dictionaries were expensive and time consuming to produce. Finally, a dictionary carrying the Oxford name would need to be superior to all others.

  Instead of authoring a new dictionary, a novel strategy was employed: a crowd of colleagues was enlisted to create the entries. Murray’s own role was not to create the dictionary; instead, he simply became editor at large. By doing it this way, Murray eliminated the costs of having experts and leveraged the diversity of the masses. Over the first few years, Murray received hundreds of thousands of submissions—slips of papers containing a word and a definition—from tens of thousands of volunteer contributors.

  Five years later, the first installment of the Oxford English Dictionary was published. Through the turn of the century, work continued on the dictionary until a complete edition was published. It has since become the preeminent English language dictionary and continues to be the gold standard today.

  The success of this crowdsourced work is amazing considering that it came before Wikipedia and even the World Wide Web. But that statement diminishes the truly incredible historical context of the story: the Oxford English Dictionary wasn’t conceived in the twentieth century, it was started 100 years earlier, in 1884! Crowdsourcing is, in fact, very old.

  There is something more important than age that sets Murray’s dictionary apart from Wikipedia. Whereas Wikipedia’s entries are entirely written and edited by the crowd, Murray remained the lead editor overseeing the Oxford English Dictionary. In many ways, this limited Murray’s creation, but it also gave it a richness of quality that Wikipedia currently lacks.

  Perhaps knowledge networks that involve expert contributions or oversight fare better than ones that rely solely on the wisdom of the crowd. The internet has enabled the proliferation of fan fiction, which consists of stories produced by fans of an existing book, movie, or television show. While written predominantly by a single author, the stories are often posted one chapter or segment at a time, allowing other fans to comment or make suggestions for what follows. Think of it as group editing versus group creation. The popular Fifty Shades of Grey trilogy started as Twilight fan fiction, and presumably received so much praise that the author was compelled to rewrite the story and incorporate the crowd’s feedback into what became bestselling novels in their own right.

  Some crowdsourcing networks are taking a high-quality approach even in their growth phases. Academic Room, developed by Harvard University, allows professors to upload files of lectures to be vetted by their peers. A professor uploads a lecture, thousands of people comment, and the lecture is retooled and revamped over time. This process is not unlike academia itself, in terms of how the peer review process works, although the use of the web creates possibilities that few scholars could have previously imagined.

  Quora is a question-and-answer website founded by two former Facebook employees. People ask questions, other people answer questions, and users vote the answers up or down. This is a similar model to Yahoo! Answers and Answers.com, but Quora has made a concerted effort toward high quality. The company, which has been around since June 2010, intentionally started very small and then gradually allowed more people to answer questions. They invited experts to opine, and that method has driven quality over quantity. Ask a question about acting and Ashton Kutcher may answer; ask about business and Mark Zuckerberg may jump in; ask about economics and you may hear from former Harvard president and Secretary of the Treasury Larry Summers. All of these people have posted answers on Quora. The site is in hypergrowth phase in terms of users but not content—they have far fewer answers than most sites, but the quality is so good that Quora is where many people head for the most difficult questions. They are leapfrogging the breakpoint, willing to forgo the size and scale of Wikipedia in order to reach equilibrium faster. No surprise, the team reportedly passed up a $1 billion purchase offer.

  There is no question that experts provide knowledge that no crowd can deliver, but the crowd’s wisdom should not be underestimated. In The Wisdom of Crowds, author James Surowiecki outlined countless situations in which crowds were far better than experts: choosing stocks, deciding the weight of objects, sports betting, even choosing who will be elected president. What sets individual expertise apart from group intelligence is a different type of wisdom. The key driver of expertise is deep experience and knowledge; for crowds, it is diversity. Understanding this is the trick to successfully navigating a breakpoint in a crowdsourced business.

  In many ways, it is easier to manage a crowdsourced business through a breakpoint than it is with other networks. After regular networks surpass carrying capacity, it’s not always easy to reduce the user base to a manageable number. But with crowds, you have the option to introduce an expert. Quora has utilized this hybrid model brilliantly. They allow crowds to answer a question, but an expert’s answer carries added weight and can dictate the direction of the conversation. As Quora grows over time, experts will likely take a more active role in weeding out unnecessary noise. As Wikipedia’s growth began to accelerate, it too leveraged this concept by giving additional powers to more experienced editors. These editors can restrict edits on pages, remove content unilaterally, and even ban other editors from the site. There are inevitably complaints about this practice, but the net effect has been to improve quality and slow unfettered growth post breakpoint.

  IV

  Crowdsourcing comes in all shapes and sizes online. Just look at the ones that match people needing work done with people seeking employment. Sometimes referred to as “cloud labor,” these companies include Elance, oDesk, Guru, CloudCrowd, and Amazon’s Mechanical Turk. Even in the deepest recession, it is easy to find jobs on these sites, and many of them can be performed from your own home. While considered crowdsourced, these companies are really just online matchmakers. They are powerful online alternatives to traditional outsourcing, with low transaction fees and the ability to hire workers for even the smallest of jobs. Cloud labor sites allow people to find work and employers to find cost-effective labor.

  Since 2007, oDesk has grown by more than 100 percent per year, reaching into over 40 countries. In 2012, there were more than 10,000 new job opportunities each month, totaling $1 billion worth of work. But this isn’t like Monster or CareerBuilder; on oDesk, the jobs are single tasks. You might make a call, stuff a few envelopes for the holiday, or assist with a dinner reservation. For that work, you might receive $.25 or $100 and perhaps the opportunity to work again. Whether you are looking for a job in customer service, copywriting, programming, or horse grooming, oDesk has it online.

  This fills a critical void in our global economy. Businesses and individuals often need help with tasks that are not large enough for full- or even part-time help. The myriad crowd labor companies address that need. The crowd helps make the labor markets more efficient.

  Crowd labor can also be effectively utilized with contests, as has been done since the eighteenth-century British watch competition. Clearly, contests are not new, nor are they purely an internet phenomenon. However, the internet is making it easier for all individuals, companies, and governments to participate.

  Netflix offered a million-dollar prize to anyone who could help improve their recommendation engine. It worked: hundreds of programmers, none of whom was actually employed by Netflix, competed to build the best software that could determine what movie you might want to watch with your boyfriend on a rainy Tuesday in July. Similarly, in 2013 GE will award $500,000 in prize money for the team that can create the most effective algorithm for easing plane congestion at major airports. Even small companies are taking advantage of this concept as part of their underlying business model. 99designs, for example, allows clients to submit a brief for a graphic design project, determine prize money for the winner, and the
n wait for submissions. The average project gets over 100 submissions, and the client pays only for the design he chooses.

  The wisdom of the crowd has long been a precious commodity for big business marketers. But the internet has enabled almost every small business to get opinions from the people who matter most. “Social care,” as it’s been dubbed, refers to providing customer service through Twitter, Facebook, a blog, or a company’s website. Sparked.com goes one step further, helping companies create advisory boards of their best customers. Sparked calls itself the heart of the social business revolution, and while that title may be superfluous, social business is certainly a revolution and an indispensable tool for any customer-facing business.

  V

  In 2012, the world was shocked by a video of Karen Klein, a 68-year-old bus monitor, being bullied to tears by four middle school boys. One man was so moved by the video that he set up an account on crowdfunding site Indiegogo, asking for a total of $5,000 to send Ms. Klein on vacation. Instead, over $700,000 was donated, allowing her to not only go on vacation, but also to retire and set up a foundation to combat bullying.

  We’ve always known that people can be counted on to help others in need—just consider the $64 million donated for victims of Japan’s earthquakes, the $145 million given for Hurricane Sandy victims, the $4 billion contributed in the aftermath of Hurricane Katrina. Donation-based crowdfunding via the web adds a few extra dimensions to charitable giving. People want to help and also want to feel part of a community, a network of like-minded souls, if you will. Calling the Red Cross and offering one’s credit card number fulfills the first need but not the second.