Stefan Savage, speaking in a segment on March 13’s On The Media, asked:

The question I like to ask people is, what are you going to do to the highway system to reduce crime. And when you put it that way, it sounds absolutely ridiculous, because while criminals do use the highway, no rational person is suggesting that if only we could change the transportation architecture that crime would go away.

Savage was speaking on the matter of internet security, and his comment was a counterpoint to a number of commentators who suggested the only way to secure the internet would be to replace the internet. This notion that we need a smarter internet has been around for a while, but its proponents have forgotten that the basic dumbness of the internet is the foundation of its success.

Mike Neuenschwander, for one, was ecstatic that the On The Media segment didn’t “slide into a futile discussion on the merits of world peace,” and followed Savage’s point with considerable discussion about the difference between the network and the social structure of trust. (In contemplating a recent NY Times story on this subject, Computing Community Consortium also quoted Savage on this point. The Coolest part: Savage commented to explain more.)

Near the end of the piece, Jonathan Zittrain explains why attempts to impose more limitations on the internet are so dangerous to the future viability of the internet:

so much of the code we now think of as central and crucial and cool and revolutionary is code for which, when most rational people first see it, their reaction is, what’s the point?

Zittrain offers Twitter as an example, but Ray Tomlinson offers an even better one. According to the legend, the man who invented email told his friend “Don’t tell anyone! This isn’t what we’re supposed to be working on,” as he first demonstrated the application that would eventually become the internet’s first killer app.

All yesterday and this morning I’ve been seeing tweets about SWIFT, so I finally googled it to see what it was about. The service promises to help organize conferences in some new 2.0 way, but it looks to be about as preposterous a social network as WalMart’s aborted 2006 attempt at copying MySpace.

There are some real lessons here, however, about how to court the early adopters that are essential to making an application that depends on user activity successful:

• The license matters. Nobody wants to invest the effort it takes to get the ball rolling just so somebody else can profit from it. Wikipedia’s GFDL was essential, SWIFT’s license policy is already strangling the site. Don’t call them “users” if they’re really the ones building the site, and don’t expect them to work for free if you’re going to make a killing off of it.
• Details matter. Would you trust a company that can’t figure out permalinks in WordPress (the site’s blog is powered by WordPress) and has URLs like http://www.imswiftblog.com/?page_id=698 to be technically capable of protecting your privacy or solving the problems that matter to you? What does the placeholder text in the last page of this FAQ suggest?
• Your friends matter. A supposed social software app that associates with a conference that looks like astroturf has already lost my trust.
• If you find yourself explaining what value your service offers to the very people who you hope will be early adopters, you’ve already lost them.

Anybody want to bet me a drink that SWIFT won’t break the 1,000 users mark before the end of CIL? I doubt anybody using the CiL wiki would take me up on that.

Cameron Moll (via Ryan Eby) wants “weight” customer ratings to reflect how two products of the same rating might have wildly different numbers of reviews.

At first glance I agree with him, but after a moment of thought, I begin to wonder if I want the ratings weighted by the number of reviews, or the number of reviews I “trust.”

Amazon keeps huge amounts of data about all its customers. So how hard could it be to correlate my purchasing behavior with the purchasing behaviors of the reviewers along with the details of which reviews I’ve previously checked as “helpful.”

reviews, review, rating, ratings, product reviews, user contributed reviews, amazon reviews, trust, weighting, recommendations