ML: As far as you can tell, what is the web really about?
What's it good for? What's it bad for?
Shirky: The web is about making it easy to move information
from one part of the planet to another, on demand, at low cost. And
that's it. That's the revolution.
Everything the web has turned out to be good for relies on these
characteristics, but no one knows in advance what the next good use
will be. Amazon showed us that the web could be good for selling
books. Hotmail showed us that it could be good for email.
Travelocity showed us that it could be good for travel planning.
Right now, a handful of people working late into the night are
getting ready to launch something else the Web will be good for, and
although we don't know what it is now, it will rely on these 'ease
of use/on demand/low cost' qualities, and it will seem obvious in
Everything the web has turned out to be bad for doesn't use these
qualities. Contrary to all the hype about "convergence", the Web is
bad for broadcasting the same thing to many people at the same
time. Broadcast TV is good for that, and the web is lousy at it,
so there is no simple merger of TV and the web coming anytime
If you could make one rule about the web that everyone would from now on
have to follow, what would it be?
I would make a rule that said that no one would ever be allowed to
make any rules that everyone would have to follow.
The usability guru Jakob Nielsen and I have a serious disagreement
on this issue which played out in a couple of ACM publications this
year [An Open Letter to Jakob Nielsen].
For Jakob, web pages are primarily software interfaces, and as
everybody has noticed, 90% of web pages are crap. Like any good
engineer, this infuriates Jakob, and more than anything else, he
would like to get his fingers around the collective neck of people
who design web pages and shake some sense into them.
I start from the opposite direction, with the premise that if the
web has succeeded where every other attempt at a usable global
network has failed, it must be doing something right. I believe the
two most important things it has done right is to open itself to
amateur content, and to impose absolutely no external standards of
"good" design. Many of the Web's most important sites -- Yahoo,
eBay, slashdot -- started from amateurs with an idea of what they
wanted and a willingness to keep experimenting til they got it
This flood of content has led to a market for quality, where
standards are not imposed externally (with the imperfect execution
that always implies) but are rather embraced by the people who have
the biggest incentive to please their users, often leading us in
surprising new directions for design. This market for quality has
actually made web design improve faster than standards-driven
attempts to do the same thing, as we saw when AOL began to abandon
its own design tools in favor of HTML.
The Web is marvelous, as is. Attempts to improve it through central
planning will destroy the very freedoms that have gotten us this far.
How would you sum up the current state of advertising on the web?
Web advertising is sitting uneasily between the mass market model
and the direct mail model, in an arena where it doesn't have to be
an either/or choice.
Web advertising has swung back and forth between the two poles of
traditional advertising without yet settling in on its own
imperatives. Before the Web, we had fast media -- radio and TV --
which could reach the consumer quickly, and we had direct media --
mail and inserts -- which let the consumer respond to an
offer. However the fast medium wasn't interactive, and the
interactive medium wasn't fast, so neither the brand people nor the
direct people can transfer what they know directly to the web.
When there are enough people who can think in terms of brand message
and direct mail accountability, at internet speed, then we'll know
what web advertising is really like, but so far we've only just seen
glimmers of what that will be.
Ultimately what role should advertising play on the web?
Advertising should connect people with new products and new
opportunities. The fact that it is doing so on the internet as well
as on TV doesn't change that basic mission.
No one alive today will see the end of the upheaval caused by the
Internet, so I don't think there is an "ultimate" role for
advertising, because like the net itself, it will always be adapting
itself to new circumstances.
What do you think advertising on the web will look like a few years
It will be more appropriate to the medium.
TV currently sets our basic assumptions about advertising, and a lot
of people are waiting for the Web to become more like TV. This
expectation is keeping people from noticing that TV is quickly
becoming more like the Web; we are currently witnessing the death
rattle of "mass media", where you could buy the attention of 1
household in 4 with a single TV show like Seinfeld. Those days are
gone, never to return. Like the Web, TV is becoming an all-niche
medium, with no TV show reaching even 1 household in 5 and no show
from the current season reaching even 1 household in 8.
To reach a large number of people in an all-niche world, you need 10
messages, not just one, and those 10 messages need to reinforce one
another while also being appropriate to each niche.
I think media buying agencies in particular are going to be
radically transformed, so rather than vertical expertise -- "We know
TV, and we can reach whoever watches hit shows" -- we will see
horizontal expertise -- "We know urban sophisticates, and we can use
not just TV but also banners ads, billboards and blimps to reach
What's the most important web trend getting the least amount of
attention right now?
The application of Open Source principles -- the idea that software
improves faster if anyone anywhere can modify or imporve it without
restriction -- to areas other than software design.
The web is so brilliant at aggregating input from a group of people
spread out across the world that any problem that can be attacked by
a "divide and conquer" strategy -- divide the problems out among
members of the group and solve them one at a time -- has a much
better chance at capturing the intelligence of a group than
centrally managed projects do.
This strategy has already given us the world's most popular
scripting language (Perl), the world's most popular web server
(Apache) and the world's fastest growing operating system (Linux),
and now its starting to be applied to other problems. Lawrence
Lessig at Harvard is looking into using Open Source principles for
legal systems, there is a movement afoot to put human "source code"
-- decoded human DNA -- into the public domain, and there are people
working on Open Source textbooks, art pieces, and anything else
where collaboration might augment or even replace central planning.
What's the dumbest thing you've seen anyone try to do on the web?
Making money off of content.
People selling content before the Web always sold the content but
charged for the package -- the paper of the magazine, the plastic of
the CD -- so when the Web came along they didn't understand the new
rules. When the production bottleneck for physical things disappears
completely, when I can write something in the morning and publish it
on the web that afternoon, then the laws of supply and demand take
over and the price falls to zero. With two exceptions -- porn and
financial data -- all attempts to charge the consumer for content
have failed. Slate, Prodigy, MSN "Channels", all wiped out because
there's no competeing with the web, and on the web, there are too
many alternatives for anyone to be able to raise their price above
What's the smartest thing?
Making money off of content.
The web is simply too good at matching people with content for this
to fail. Every society that has ever existed has found a way to
reward its creators.
The current situation is grim, as I outlined above, but that just
means that the problem is bigger than we first understood. Despite
the lack of physical bottlenecks, there are still human bottlenecks
-- everybody cannot read everything. Eventually, in the flood of
content released by the web, some way of sorting the good from the
mediocre will arise, and people will pay for that service. Its not
clear if they will be paying money directly, as for a book, or
whether they will be paying attention, attention which the
advertisers will then buy, but over the next few years, someone will
figure out a solution to the problem of rewarding creators for their
What is it about the web that you most enjoy personally?
That feeling of "Aha!" when I come across someone else's way of
thinking that turns a lightbulb on in my head. I get this much more
often from essays people write to explain things to their peers,
essays like "Lessons From Lucasfilm's Habitat" and "The Cathedral
and the Bazaar", than I do from any traditional media outlet.
What is it about the web that you most worry about?
ICANN, the group responsible for taking over the orderly running of
the Internet address database. The Internet is so marvelously
distributed that it seems almost impervious to takeover, so we've
been lulled into thinking that its always going to be this way.
There is one chokehold: the master record of internet addresses,
which is what keeps traffic flowing in an orderly fashion from point
A to point B. Ownership and management of that database is hotly
contested right now, and in all the horse trading between the
current owners (Network Solutions, Inc.) and the future owners
(ICANN), the whole fabric of the net could be compromised.
Its a geeky, complicated, unsexy issue, but it matters a damn sight
more than amazon's quarterly numbers to the future of all of us
using the net right now.
What are a few of your favorite web sites and why?
slashdot.org - The geek's Wall St. Journal, and web design done by
people for whom the web is a native tounge. The design is
marvelously spare, owes nothing to the 'print graphics ported to
the web' tradition, and works beautifully.
freshmeat.net - If slashdot is the Wall St. Journal, freshmeat is
the stock exchange itself. Another example of a web interface by
people who have not been sidetracked by the print tradition.
mcsweeneys.net: My dream site - no pictures, just text. One of
the only web 'zines that lives up to the promise of the medium, and
one of the only sites I read for the quality of the writing.