[WikiEN-l] 6 months later: Knol update

[Gwern Branwen]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512

On Tue, Jan 20, 2009 at 10:56 AM, Anthony  wrote:
> On Tue, Jan 20, 2009 at 10:30 AM, Gwern Branwen  wrote:
>>
>> As will surprise none of the Knol nay-sayers here (in which number I
>> believe I can count myself), Knol hasn't done too great.
>
> Compared to what?  I can't imagine Knol is much worse than Wikipedia when it
> was 6 months old.  Knol just published its 100,000th article.  When
> Wikipedia was 5 months old, it said on the main page "We've got over 6,000
> pages already. We want to make over 100,000."  The Wayback machine then
> skips ahead 5 more months, by which point Wikipedia brags "We started in
> January 2001 and already have over 13,000 articles. We want to make over
> 100,000, so let's get to work"
>
> To be sure, Knol has a lot of very serious problems with it.  But it's only
> 6 months old.  The concept is far from finalized.  6 months into Jimmy
> Wales' encyclopedia dream he was still working on Nupedia.
>
> Here's the Wikipedia on George Bush 19 months into Wikipedia:
> http://web.archive.org/web/20020817062610/www.wikipedia.com/wiki/George_Bush
>
> How long will it take you to find a better article on Barack Obama in Knol?

With Wikipedia by this point, the basic concept of collaboration had
been proved. With Knol, we see only the divisiveness of the payments
system*, and a few isolated authors striving on their own.

More to the point: yes, we should expect more of Knol than of
Wikipedia at similar stages! Knol has, by virtue of its position in
time, *numerous* advantages over Nupedia/Wikipedia. We should expect a
lot more.

It has:

1) A clear license regime. Thanks to 8 years of Creative Commons, the
choice is not limited to just the GFDL (with its many problems).
2) 8 years of hardware advances, or approximately 5 iterations of Moore's law.
3) 8 years of wiki development, demonstrating dead ends, the good
ideas, & what remains to be improved. Imagine if Knol had to start
with the state of the art in 2001. It would be truly gruesome. (Anyone
looked at the very old Wikipedias in Nostalgia, or old usemod wikis
like Ward's? They're hideous and unusable! They make me quite grateful
for 2009 MediaWiki with all its modern conveniences.)
4) The backing of a commercial juggernaut. Quite aside from Knol's
hosting being a) very good; and b) not the Knol devs' concern,
Google's backing offers an array of advantages, from certainty to
excellent software development resources**, such as:
5) Massive publicity. To be facetious, at launch Knol had infinitely
more publicity than did Wikipedia.
6) A public educated to read wikis, and to use them. How many people
could Wikipedia hope to draw on at day 1 - that cared even a little
about Free content, that knew what a wiki was, that wouldn't dismiss
it as hopeless, and had an editing familiarity with wikis? Darned few.
We had to constantly evangelize and educate people about wikis, and by
dint of unremitting effort create the English Wikipedia and make it
interesting and valuable enough that people would contribute who
didn't fulfill any of those criteria. En was the existence proof that
large-scale wikis were possible and valuable. Knol, on the other hand,
can draw immediately on that pool of people Wikipedia created.
7) A model targeted directly at people unhappy with Wikipedia. Are you
an expert tired of 'anti-expertism' on Wikipedia? Why try to get along
with those bumpkins when you could have your own article completely to
yourself on Google Knol (and get paid for't)? Wikipedia appealed to
those unhappy with Nupedia. Nupedia when Wikipedia launched was a lot
smaller than Wikipedia was when Knol launched. I think this pool of
possible contributors was thus also much larger for Knol than it was
for Wikipedia.

etc. etc.

* If I weren't so lazy, this is where I'd cite some of the studies
showing paying some contributors to FLOSS projects reduces
volunteerism.
** Not to denigrate the efforts of Magnus and Tim and all the other
MediaWiki developers over the years, but one simply expects more of
full-time developers experienced with the famous Google infrastructure
and supposedly at Google's standards of excellence.

- --
gwern
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)

iEYEAREKAAYFAkl3RIsACgkQvpDo5Pfl1oJWXwCfVL0x328tmey5rElGolu44PQj
2RcAoIhofvJUWXmj0HoeJwbl+kPSftIB
=kEJG
-----END PGP SIGNATURE-----