Subject: Re: near/medium future digital media economics
From: "Ben Tilly" <>
Date: Fri, 19 May 2006 07:07:51 -0700

 Fri, 19 May 2006 07:07:51 -0700
On 5/18/06, Thomas Lord <> wrote:
> Ben Tilly wrote:
> > Given the known power laws governing the distribution of articles of
> > interest for researchers, it is extremely unlikely that Metcalfe's Law
> > applies in Tom's setting.
> >
> No, not so.  I took note of that example in yr paper but their isn't time
> to reply to everything preemptively :-)

That's never stopped you before...

> Traditional journals are a small number of broadcast channels.  There
> aren't very many of them.   They don't comprise a bidirectional network,
> to the first approximation, let alone one anchored on participatory content
> development.

Moving your content from a traditional journal to participatory
content development doesn't change whether I'm interested in what
you're working on.  Secondly I'd differ on how bidirectional journals
are.  (At the least, every journal article includes contact
information for the author and people *do* use that.)

> > For collaborative content there is also a time dimension - a network
> > that has been producing content with twice as many people for twice as
> > long has 4 times as much content.
> >
> > However size is not the only factor.  While collaborative content is
> > big and will get bigger, there are limits to what you can do with it.
> > For instance wikipedia has a demonstrated history of accidentally
> > *discouraging* experts from getting involved because said experts
> > discover fairly quickly that participating means constantly trying to
> > explain the basics of their subjects to people without much background
> > or interest in aquiring said background.
> And that's a direct result of one of the ways that Wikipedia *fails* to
> be a true network.   It's a better approximation than many things, sure,
> but it has limitations of scale because (like journals) it is ultimately a
> centralized, scarce, broadcast channel.   It just happens that bandwidth
> on that channel is open to the public to fight over.   At small scale, and
> before it was too widely noticed, it emulated a network (before there
> was much to fight over).

Sorry Tom, but I believe that that's intrinsic to the nature of the
beast.  Given that some content is more interesting to more people
than other content, a decentralized network will centralize.  This has
been documented to happen even for something as fundamentally
decentralized as blogs.  (There is a handful of blogs that gets an
extreme amount of attention.)  If you make the content collaborative,
then competition will emerge over any bits of content that are
interesting to lots of people.

If you go to extremes to make it difficult to centralize, eventually
someone like Google will come along and make it easy to find the
people that you want to centralize around.

> Now, if said experts could more easily go momentarily sandbox
> themselves, work
> stuff out, and put up a face -- then we'd be back in network territory.

This is not how opposing camps have *ever* settled their differences.
Unless you change human nature, this won't be how they do it in the
future either, no matter how cool a communication technology they

Also note that people participating in collaborative content are hardly experts.