Made for the distributed world

I just came across Dinnerbuzz (catching up on my RSS after vacation; saw it via You’re It). Though the execution is iffy at best, the concept is close to what I’m talking about in creating new information services for the distributed world. Here’s the deal:

When you post a review of a restaurant on your own blog, you tag it and Dinnerbuzz picks up the link and aggregates it with other links to posts about that restaurant, other posts with those tags, and other posts in those cities. So when it comes time to eat, you can come in and find what locals are saying about a restaurant or you can search for “outdoor” “Mexican” joints in “New York.” Further, you’ll be able to get RSS feeds so you can get an alert whenever someone writes about a great new vindaloo in your neighborhood.

In old-centralized-marketplace-think, you’d try to get all those people to write restaurant reviews on your big-media site. And the question is: Why the hell should they? What do they get out of it? And in the old world, you tried to get people to read the reviews on your site when they knew there were reviews on tons of other sites out there as well and it’s a pain to find them all.

In new-decentralized-distributed-think, you recognize that people will write about what they want to write about where they want to write about it and if you’re smart, you’ll find ways to take advantage of all that great information and aggregate it and and aggregate audience around it, sending traffic out to all those writers on the edge because readers know they can come to you find find it all.

To make this work, you need to get people to tag their posts and you need a critical mass of them so that people can start to agree (e.g., “byob” instead of “dry”) on the right tags as happens on Flickr and Del.icio.us. But people will do that if they see that people are finding what they right because they tag and also if they start using the service themselves to find restaurants and so, in this gift economy, they realize that you need to give to get.

The example I’ve often used about how tags will work best in a distributed world is jobs: You tag your resume anywhere on the internet and a specialized successor to Google (who may, indeed, use Google’s API to get raw data) finds jobs and matches them with job seekers without forcing anyone to pick one centralized marketplace or another. I’ve also said this will work in hyperlocal: I don’t want to write an entire blog about my town, but I would tag the occasional post to be aggregated into a community of them — because I’d want to read that collection myself.

This is a model for the future of media. There is tons of great stuff to be had out there; it’s impossible to find and keep up with it all; search won’t do the trick; tags and feeds will help. The key is not to collect the content and traffic — the old, centralized media way — but instead to collect enough information about that good stuff to help people find it when they want and to help support the people who create it all.

: OOPS: Well, it appears I was projecting what I wanted Dinnerbuzz to be. I misread one description of it. As I see the service now, I have to go there to add tags to it with a link to my post.

It would be better if I could just put the tags on my post (Technorati tags) or on Del.icio.us (with a for:dinnerbuzz tag) or simply add the posts and ping them and that would travel to Dinnerbuzz automatically. Those would be the better, more distributed ways to accomplish this.

I also find it terribly frustrating that I can’t find the way to get from a Dinnerbuzz listing to the actual posts!

Or I’m wrong again….

Well, at least in my imagination, I see potential here….

Made for the distributed world

Made for the distributed world

: I just came across Dinnerbuzz (catching up on my RSS after vacation; saw it via You’re It). Though the execution is iffy at best, the concept is close to what I’m talking about in creating new information services for the distributed world. Here’s the deal:

When you post a review of a restaurant on your own blog, you tag it and Dinnerbuzz picks up the link and aggregates it with other links to posts about that restaurant, other posts with those tags, and other posts in those cities. So when it comes time to eat, you can come in and find what locals are saying about a restaurant or you can search for “outdoor” “Mexican” joints in “New York.” Further, you’ll be able to get RSS feeds so you can get an alert whenever someone writes about a great new vindaloo in your neighborhood.

In old-centralized-marketplace-think, you’d try to get all those people to write restaurant reviews on your big-media site. And the question is: Why the hell should they? What do they get out of it? And in the old world, you tried to get people to read the reviews on your site when they knew there were reviews on tons of other sites out there as well and it’s a pain to find them all.

In new-decentralized-distributed-think, you recognize that people will write about what they want to write about where they want to write about it and if you’re smart, you’ll find ways to take advantage of all that great information and aggregate it and and aggregate audience around it, sending traffic out to all those writers on the edge because readers know they can come to you find find it all.

To make this work, you need to get people to tag their posts and you need a critical mass of them so that people can start to agree (e.g., “byob” instead of “dry”) on the right tags as happens on Flickr and Del.icio.us. But people will do that if they see that people are finding what they right because they tag and also if they start using the service themselves to find restaurants and so, in this gift economy, they realize that you need to give to get.

The example I’ve often used about how tags will work best in a distributed world is jobs: You tag your resume anywhere on the internet and a specialized successor to Google (who may, indeed, use Google’s API to get raw data) finds jobs and matches them with job seekers without forcing anyone to pick one centralized marketplace or another. I’ve also said this will work in hyperlocal: I don’t want to write an entire blog about my town, but I would tag the occasional post to be aggregated into a community of them — because I’d want to read that collection myself.

This is a model for the future of media. There is tons of great stuff to be had out there; it’s impossible to find and keep up with it all; search won’t do the trick; tags and feeds will help. The key is not to collect the content and traffic — the old, centralized media way — but instead to collect enough information about that good stuff to help people find it when they want and to help support the people who create it all.

: OOPS: Well, it appears I was projecting what I wanted Dinnerbuzz to be. I misread one description of it. As I see the service now, I have to go there to add tags to it with a link to my post.

It would be better if I could just put the tags on my post (Technorati tags) or on Del.icio.us (with a for:dinnerbuzz tag) or simply add the posts and ping them and that would travel to Dinnerbuzz automatically. Those would be the better, more distributed ways to accomplish this.

I also find it terribly frustrating that I can’t find the way to get from a Dinnerbuzz listing to the actual posts!

Or I’m wrong again….

Well, at least in my imagination, I see potential here….

  • http://oodja.blogspot.com Jersey Exile

    Jeff,
    Something very similar to this has been evolving on the “Deep Web” of proprietary academic databases and indexes, especially as technology advances and publisher mergers have made searching laterally across these databases easier and easier. The aggregate/feed model has become the research method of choice for an increasing percentage of scholars and scientists, and definitely of more use (at least right now) than the still-nascent Google Scholar.
    It’s interesting to see how these principles will transform the “‘Shallow’ Web” as well!

  • http://people.etango.com/~markm/ Mark Mascolino

    Sadly, there isn’t a link for the original post because the Post Submission screen doesn’t let you enter a URL for your post. All this site really is comes down to a centralized resturant review site that implements the folksonomy tagging structure. No distributedness at all.
    Your mentioning the for:dinnerbuzz as a del.ico.us tagging scheme is a good one. I was thinking of creating a recipe finding site along the same lines and couldn’t think of a good tag name. for:deliciousrecipes would hit the spot (although its prehaps a bit long).

  • http://www.focusedperformance.com/blogger.html Frank Patrick

    I don’t see the ability to write a post to Dinnerbuzz from anywhere but from within the Dinnerbuzz site itself.

  • http://www.downes.ca Stephen Downes

    “In new-decentralized-distributed-think, you recognize that people will write about what they want to write about where they want to write about it and if you’re smart, you’ll find ways to take advantage of all that great information and aggregate it and and aggregate audience around it, sending traffic out to all those writers on the edge because readers know they can come to you find find it all.”
    Right. Like this:
    http://www.downes.ca/cgi-bin/xml/topics.cgi
    “To make this work, you need to get people to tag their posts…”
    Wrong. Too many people don’t tag. people who do tag don’t agree on tags.
    You need to analyze the posts for instances of patterns. For example, I define a topic (aka ‘tag’) as a regular expression, then scan for matches.
    — Stephen

  • http://www.corante.com/getreal Stowe Boyd

    I riff on your post at Get Real, and make a case for an Open Tag model.