The Standards Blog

Whither the Knol? Google Takes a New Experiment Live

Semantic & NextGen Web


Google Knol

Back in December of last year, Google posted a brief announcement of a new experiment in online publishing. At first blush it seemed to represent a challenge to the Wikipedia - but with a few differences. Google summarized the concept as follows:

Earlier this week, we started inviting a selected group of people to try a new, free tool that we are calling "knol", which stands for a unit of knowledge. Our goal is to encourage people who know a particular subject to write an authoritative article about it. The tool is still in development and this is just the first phase of testing. For now, using it is by invitation only. But we wanted to share with everyone the basic premises and goals behind this project.

Then the project dropped out of sight, while the chosen authors contributed initial content, and while Google decided whether to green light the project for ongoing support and public participation

This Wednesday, Google lifted the password curtain on its infant knol site, and issued a new announcement.  In some respects, the description of the knol game plan (and even the words) are identical to what we read in the original blog entry.  In others, they are different, apparently reflecting lessons learned and author feedback received during the intervening seven months.  And, of course, there is now the nascent site itself to browse and watch evolve as well. 

What you'll see when you visit turns out to be quite different from the Wikipedia - at least for now.  Given that until yesterday it was neither available to public feedback nor open to volunteer authors desiring to launch their own knols, what it looks like today will almost certainly be very different from what it looks like a year from now, or perhaps even in month from now.  With the door now open to anyone that wants to walk in and the freedom to go in any direction - not to mention ad revenues to be reaped and shared with authors - what we will see will be akin to the Apple App Store for the everyman and woman - not just for developers, but for anyone who can type.

Here's my take on what to expect, how the knol experiment may evolve, and why I think it matters - a lot.

Google KnolWhat makes a knol:  As originally announced, anyone can start a knol.  And as usual, Google has made it very easy for anyone to get started.  The most obvious questions about what you can do and how you can do it are answered at an Introduction to Knol page where you can also read the questions of/responses to other aspiring knol authors. 

There are a number of significant differences from the Wikipedia that invite a sort of Cathedral and Bazaar comparison, contrasting the unitary model of the Wikipedia, with a single set of rules, a stylistically coherent end result, and a strictly non-profit operational model, and the more liberally administered knol site, where anyone can write about anything any way they like, and even make a little money while they're at it if they so desire (or not).  Thus, while the Wikipedia has created a product with proven value, associated in part with its consistency, the knol site will provide a test bed within which many approaches may be taken, with some proving to be more popular than others (perhaps even as between disparate topic areas - say medical information vs. popular music).

Here are some of the key features of the knol project that should be of interest to curious authors:

  • You must sign your work (taking both credit as well as responsibility for it)
  • You can launch a knol with other authors if you wish
  • You can allow visitors to edit your knol (and moderate the changes if you wish - this is the default setting) or restrict it to your changes alone
  • You can allow ads to display, or keep it commercial free
  • You can launch your own knol on a topic that someone has already adopted
  • You can choose the copyright mode for your knol (e.g.,  Creative Commons - the default setting - or All Rights Reserved)
  • You can allow visitors to comment on your knol

An intuitive interface allows you to choose the settings to your liking, as described here.  You'll also see an active discussion there of what new authors like and don't like about the set up, in which you can participate.

How will knols evolve:  Although I have suggested that we can expect to see a wide variation of styles, Google appears to be hoping otherwise, while not (apparently) reserving the right to compel that result.  As you'll see at the Best Practices: Writing Good Knols page (reproduced in full at the end of this blog entry for archival purposes), Google is hoping that the knol project will attract very high quality, "authoritative" articles, while at the same time inviting authors to "voice your opinion" - at best a difficult tightrope to walk, even for experienced authors, and at worst a pair of contradictory expectations for an unedited site. 

By launching the site with a range of thorough, well researched and written articles by knowledgeable authors, Google is clearly hoping to set the tone.  That said, I performed ten searches on broad and obvious topics (e.g., IBM, NASCAR, Great Britain, Global Warming) and didn't get a single hit.  What that means is that if the knol site becomes broad enough to be useful it will take an enormous amount of participation, with the results swamping the starter set of well-styled contributions.  And it will remain to be seen whether edgy or scholarly entries attract more visitors, and higher revenues.  If the former proves to be more popular than the latter, Google's hope for a successor to the Encyclopedia Britannica may slide closer to a competitor to People Magazine.

It will also be interesting to see whether Google decides to allow knols to adopt the types of additional features that blogs and social networking sites are constantly evolving, or whether Google will want to the knol site to maintain a more "respectable" image.  Even within that constraint, however, perhaps an author should be able to load the widget that allows them to display the covers of the books they are reading.  How better to allow readers to gauge whether an author's viewpoint is biased?

But why stop there?  Should a knol writer be able to decorate her own knollish cubicle as she wishes?  Early indications are favorable: Google has cut a deal with The New Yorker magazine that will permit each author to display one New Yorker cartoon per entry, providing an urbane and sophisticated aura to each knol entry that chooses to opt in.  Should authors be able to go beyond such Google-enable options, even adding revenue generating widgets from other sources, perhaps making a little extra money for themselves (and Google) as well?  Who knows.  We'll see.

What's in it for Google:  So why would Google want to launch the knol project, and why in the way that it has chosen to do so?  According to Cedric Dupont, the knol project manager, "The ultimate goal is: we want to improve search."  Well, I expect that the potential to expand ad revenues might factor in there as well.

But that's OK, and even (as I'll discuss below) preferable.  I find the knol concept to be a refreshing affirmation that the Google ethos ("don't be evil," at least as somewhat subjectively defined by Larry Page and, especially, Sergey Brin) and innovation is alive and well in the Googleplex.  Google is also to be commended on allowing all authors to choose the copyright mode for their entries, meaning that it has not locked in its authors to the project at all.  Theoretically, someone could make an open offer in the future to all knol authors that have reserved their ownership rights to move their content wholesale to another site.  And much of the Creative Commons material would already be available, depending on the intended destination.

Happily, avoiding evil seems to be working very nicely, thank you, for Google's shareholders as well as for Google visitors and (now) potential authors.  Every engineer continues to be required to spend 20% of her time on projects of their own choosing, resulting in an ongoing torrent of ideas competing for attention, productive criticism, and perhaps adoption.  At the same time, the knol project demonstrates that those that hope to be allocated the necessary resources to go live are still allowed to honor goals beyond maximizing ad clicks.  Ultimately, I believe that's good for everyone, as such a mammoth project will need tens of thousands of participants.  Allowing them the freedom to choose whether to contribute their time purely for non-economic satisfaction or to earn a few dollars on the side harms no one and honors the motivations of all.

What does it mean for Wikipedia:  First and most obviously, the knol project has to attract thousands and thousands of entries - something that may never happen.  With a seven year head start and over 8.2 million entries in over 200 languages (some 2 million are in English), the knol project seems unlikely to ever catch up, if entry numbers are the measure, even if it is successful.  Moreover, while Google appears to be hopeful that the knol project will distinguish itself through higher quality and accountability, that is a reputation yet to be earned.  It may well be that community spirit and the "many eyes make all [factual errors] shallow" reality of the Wikipedia will deliver higher quality than knol owners that are free to ignore suggestions offered by better informed readers.

Ultimately, I expect that the Wikipedia and the knol project are more likely to evolve into sufficiently different resources that each will attract its own audience.  That would be ideal, as it would better support through providing dual sources of information that can be compared. The truly interesting experiment will be to see whether the appeal of having sole control of content and perhaps an extremely modest profit opportunity will outweigh the psychic satisfaction of being part of a very large and (outwardly) anonymous community that has achieved enormous success.  The results of that experiment may prove to be uneven, with the profit motive proving to be more compelling, and with authors competing to write the definitive entry on male erectile dysfunction (imagine the ad revenues!), while those interested in all things lepidopteral continue to gravitate to the Wikipedia.

Why the knol project matters:  I noted above that I think that the knol project is important, but did not suggest why.  Here's my concern.

Part of the hugely good news about the Internet and the Web is that both, through a combination of deliberate decision and happy accidents became established on a free-access basis.  The bad news, however, is that there is less and less profit motivation to do more and more of the things that used to pay the wages of millions of people in businesses such as the media and software development.  The "before" picture of fifteen years ago is a world with many types of profit making, wage paying businesses plus a multi-billion dollar advertising industry.  The "after" picture of today increasingly resembles a world with only advertising revenue and a wealth of free products and services that used to represent separate revenue streams.

The bottom line is that you can't eliminate billions upon billions of dollars of revenues and support the same range of products, services, innovation and in-depth research with only ad revenues to pay the bills, even taking into account the substantial savings achieved through shared open source development.  I firmly believe that we are in an as-yet unappreciated downward spiral in quality and choice that will eventually need to be addressed and reversed.

That downward spiral is most obvious in the media, as respected newspapers axe their reporting, editorial, and foreign correspondent budgets.  I don't expect my RSS aggregator to ever deliver to me a set of morning feeds that equal my morning New York Times for breadth, depth or sophistication of analysis.  And that worries me.

That's where projects such as the knol come in.  What I truly hope is not that the knol will replace the Wikipedia, but that it will provide a test bed within which multiple business models will evolve that will inspire the creation, aggregation, and convenient presentation of high quality content, hopefully on a profitable basis for those that contribute them. 

My ultimate dream would be to see the evolution of a knol equivalent of a newspaper, with well respected correspondents from all over the world contributing news, opinions and analysis on a daily basis - and getting paid a living wage for it.  If the Google knol encyclopedia project could show the way to new resources like that, I'll be a very happy man.



For further blog entries on the Future of the Web, click here

sign up for a free subscription to Standards Today today!

* * * * * * * * * * * * * * * * * *

Overview Knols are meant to be authoritative articles about a specific topic. In order to write a successful knol, think about what readers would want to read when searching online for the topic you are writing about. All good writing takes time, we hope you will find that Knol tools help you in the process. Here are a few tips and tricks on how to write good knols.

Things to do

  • Voice your opinion. Your name is behind your knol, and it should reflect your unique point of view.
  • Be organized in your writing.  Consider your audience and how best to present information clearly. For example, use headings to divide sections and use lists and formatting to allow users to quickly scan material.
  • Be succinct, but comprehensive on your topic of choice.
  • Provide references, and display your credentials. Readers will want to know who you are and gain context on the knols you are writing.
  • If you know of good resources on your topic, link to them from within your article to make it easy for your readers to learn more.
  • Ask for reviews. A fresh perspective from someone else always improves your writing. Learn about reviews in knol.
  • Review user comments regularly and update your knol with feedback.
  • Follow the rules, please. For the specifics, check out our Content Policy.

Things to avoid

  • Don't write a blog. Knols are meant to be standalone articles on a topic of your choosing. Knol is not optimized for diary-type writing.
  • Don't write teaser articles. Knols should be comprehensive pieces of information on a topic. Readers are like you, and don't like getting incomplete information.

Searchability Knols are indexed in search engines like any other website. Search engines will generally give more importance to your knol title and subtitle, so try to choose a relevant and descriptive title and sub-title.
Credentials One important way readers evaluate the quality of a knol is by the author's credentials. You are not required to specify your name and credentials, but you may wish to create a bio for maximum user visibility.



I think your concerns about the 'loss' of revenue-generating jobs are misplaced.

1. I've found that most journalists, don't have a clue about the technical, political, or international events they write about.  Far too many 'journalists' (especially those from the NYT & Wash Post) are simply parroting entries from the AP, Reuters, or some other source.  Most of the time, the so-called investigation done by the 'journalists' is slip-shod at best - full of omissions, misstating factual information and drawing invalid conclusions from the facts they do manage to quote correctly, etc.  I would be relieved if most of these 'professional reporters' were replaced with internet blog sites that accept comments from the people directly affected by the news - people that have correct information and that understand the facts and implications of the situations better than a 'professional journalist' and that allows people to discuss the article from various perspectives to arrive at the truth.  Instead, we have to accept a 'journalist'  that shows up for 5-15 minutes, asks inappropriate questions, then tries to fit the answers into the journalist's pre-conceived opinion and world-view prior to writing an article that is edited for political correctness before being published as fact.

The ability of the average person to provide valuable contributions to a news story should not be underestimated - Groklaw can be considered a prototype if you concentrate on the recent 'request to pick your brain' articles.  At Groklaw, techies contribute technical info, lawyers contribute legal perspective, those with experience in certain areas contribute that first-hand experience.  At the NYT & Wash Post, 'journalists' generally transcript AP or Reuters stories and periodically contact a small set of 'analysts' to get quotes they can use to bolster their word-count.  There appears to be no first-hand experience being reported by journalists, nor doe there appear to be attempts at unbiased reporting.  I long ago stopped reading anything printed by the NYT or Wash Post because I find their reporting not only technically incorrect, but also politically biased.

As to software houses, I think that rather than worry about the economic impact of MS and their proprietary ecosystem going out of businesss, consider the advantanges to the world's IT industry if each programmer that is paid by the 'pay-for-software' business model were actually employed by industry players - database authors, insurance industry, countless SMEs, financial institutions, etc - each 'industry player' hiring a few in-house developers that use and contribute back to FOSS projects.  I think there would still be a market for the pay-for-software model, but it would be for niche products like games rather than for 'bread-and-butter' projects like operating systems or Office Suites, security products, anti-virus products, network management products and protocols, etc.

I believe that it would be MUCH cheaper and far more efficient for the world to have open-source projects/products supported and developed by the industry that actually uses that product rather than have software whose feature set and required operating system are defined in a 3-rd party executive boardroom whose primary goal is to maximize the monetary value of software that currently does not currently meet all requirements of their customers, cannot timely meet the changing requirements of their customers, and will probably always lag the true needs of their customers by several years.  Additionally, by having the development of the product/project done by actualy industry-hired programmers, the industry members can determine which operating systems their in-hous employees with concentrate on - whether that be the financal industry, manufacturing, aerospace or retail.

Today there's a lot of venture capital being invested into software products that truly have no market because neither the venture capitalists nor the startup have done a valid survey of the industry.  This type of waste of resource would not happen with the FOSS development model described above (although the proprietary software niche market would still be subject to this particular business risk).

Microsoft and their supporters keep arguing that if FOSS succeeds, free-market economies will collapse and economic chaos will ensue.  This is BS.  Instead of concentrating software development in proprietary software houses that often have inept project managers and/or project management procedures, much of the software in use today would be better and would innovate faster if it were decentralized among all the *USERS* of the software and if each programmer involved in the project were working with and encountering the problems of that industry on a daily basis.  The intimate knowledge of the industry's needs would then be built into the design of the software and the sofware would be that much more useful and productive.

Thanks for the long and thoughtful comment; I'm sorry for being so long in acknowledging it.

While I'm personlly not as down on the best of the media (and particularly the NYT), I don't disagree with much of what you say on each point.  Having been interviewed many times, and also followed some stories closely enough to know how well they have been reported, I know that (as anywhere else) there are very good reporters, very careless reporters, and very lazy reporters.  Unfortunately, the financial pressures the media is under is making it a lot tougher to be a good reporter.

There are also good things about both the Groklaw model and the NYT model.  The former, as you rightly point out, can dive very deep, draw on a lot of resources, and develop a discussion at great length - but the model tends to work best within a variably narrow or wide focus, depending on the site.  The NYT model can try to cover the world from many dimensions over a long period of time, and select out a daily range of news that can help people stay on top of a variety of issues - something that would be very, very time consuming to acquire through site visits.  The Economist can do the same - with a viewpoint that is not hidden, and therefore can be appreciated.  BusinessWeek can cover a sector deeply.  And so on.

What I will be watching with great interest is whether the best of both can be combined.  I think that would be the best of all possible solutions - although I would still greatly miss the experience of a physical paper and a morning cup of coffee by the water, or at a cafe, or on a plane, if only for tactile and nostalgic reasons.

  -  Andy