conclusions from the SXSW panel
March 27, 2003
About two weeks ago, I did a panel at SXSW with a couple of friends, and the topic was the future of personal publishing.
There were a handful of points that I keep coming back to in regard to the future of weblogs, and they're all based around the fact that we experience life as a series events, each one registered as an individual idea, filtered through our experience and context. Perhaps not coincidentally, weblogs are based on the publishing of idea-sized entries, with social, political, economic, and cultural context assigned by choice of links and blogrolls.
Though Justin, Ben, Paul and Mena each made great points during the panel, I can't really do justice to articulating their ideas, so you'll have to check out Heath's transcript for what they were trying to communicate. But I thought it might be useful to summarize the points that I intended to get across, a few of which I actually succeeded in articulating.
The first one's no surprise, if you're a regular reader of this site or my sidebar: We're all going to start using desktop software that's designed to handle chunks of data, regardless of source or format. I explained a lot more about this in Introducing the Microcontent Client. Paul disagreed, saying that it makes more sense for the chunks of data to come to you in the applications you're already using, like your mail client or IM client. He almost had me convinced.
The second idea that I really believe in, despite the fact that nearly everyone who heard it thought I was either being crazy or facetious, was that in 2 or 3 years, many of us will be reading 10,000 weblogs. It's a hard statement to justify literally unless you factor in how sofware and platforms are going to evolve.
If you look at the classic Gladwell progression of idea spread, as outlined in The Tipping Point, the tiers at which social systems tend to reach new thresholds occur at groupings of (roughly) a dozen, one hundred, and ten thousand.
Two or three years ago, I read about a dozen weblogs. Now I actively track nearly a hundred, with another hundred that I either check infrequently or read because they're linked from the sites I am already visiting. I also (still!) check in on MetaFilter, which currently has over seventeen thousand members. Now, I'm not actually asserting that I'll click through a 50-page blogroll, reading each update, but I do think that a smart system that knows the people I trust (maybe the same list of 100 that I use now) will be able to follow the network out to the people they trust, and when you get two generations out, you're easily past 10,000 blogs.
So get the posts from people I care about, plus the relevant comments from the 10,000 people who are within 2 degrees of my blog, and you've got my personal MetaFilter, with posts just by people I'm interested in, and comments from only the people who are either trusted by them or me. Only I think I'll read it in a client like NetNewsWire or NewsGator, not the web browser.
It's not that far from today's New York Times, in some ways. The Sunday Timeshas easily got 10,000 contributors in total, if you count all the AP stringers who contribute wire stories and all the editors who pore over the text and all the people who write copy for the ads, etc. Except that I don't get to be in ultimate control of what appears there. With weblogs, I will.
Next up: The really big point I wanted to raise about weblogs, but didn't get to mention yet.
4:06 PM
6 TrackBacks
I finally sussed the functions in PHP that build a tree from a database and lays it out correctly, given only the parent of each record. That was a big fairy tale which is now over. Now I can concentrate... Read More
Came across this article by Anil Dash (through EriK). What I mostly liked was the statement at the beginning: we... Read More
Came across this article by Anil Dash (through EriK). What I mostly liked was the statement at the beginning: we... Read More
I've just spent an hour getting all my recently subscribed-to RSS feeds working. My favourite way of subscribing is to simply hit Subscribe in the Mac Service menu, which pipes the URL to NetNewsWire. If there's a link in the page in the right format, ... Read More
I've just spent an hour getting all my recently subscribed-to RSS feeds working. My favourite way of subscribing is to simply hit Subscribe in the Mac Service menu, which pipes the URL to NetNewsWire. If there's a link in the page in the right format, ... Read More
I need more things to read. As a result, I'm scraping URL's from Matt's portal and running them through my Feed on Feeds meat-grinder. I'll probably be nice and update Ye Olde Sidebar today to reflect the deltas. [Who am I kidding? I'll neve... Read More
16 Comments
Leave a comment
- Earlier: hacking AOL magazines
- Next: secrets of music promotion
"Only I think I'll read it in a client like NetNewsWire or NewsGator, not the web browser."
Why's that?
Come on Meg - you know the answer - because HTML is simply not sufficient for building rich media apps - or any apps at all, besides the most simple and mundane ones.
Right on to Anil for pushing the notion of chunky thinking. I agree that Email and IM are still FINE ways to grok our data - but what happens when the ideas are created and stay in chunky form WHILE AT THE SAME TIME we can create new kinds of apps and services which will - well like you said - become your own personal metafilter.
Though I dig this data/idea game - I'm stil more turned on when we equate idea - to people and what happens when we can start to interact at the social software level - like Clay talks about - while grooving on media, and using our Home LANs and devices.
I think that what you really mean is that in the future people will have the aggregation of 10,000 weblogs available to them.
Many people seem to think you mean you'll read every post on 10,000 weblogs. There's simply not enough time in 24 hours to do this.
It's much more reasonable to predict having easy access to the content of a huge array of weblogs.
Am I close?
I believe that at some critical masss of text/information consumption, our individual brains (and collective culture) will begin to systematically reject the aggregated narratives, opinions, articles and facts that the kind of software apps you're talking about have been funneling into our eyes and minds.
At that point, we will abandon the infrastructure of the web as we know it (and as you love it, Anil) and begin to work toward a text-free world and a text-free web. Desiring to escape the trappings of written language, we will reinvent our cultural memory without symbolic heiroglyphics, relying instead on voice, image, art, soundscapes and full-motion video eyecapture.
Aggregate data, Trackbacks and NewsGator will be tools only of old-world Gurus like you, Mr. Dash. The rest of us will move forward into some uber-web. The future of weblogs? Real-time auto-interest-filtered multi-media brainsharing, communed through 3D contact lens chips and earDolby 10.9.
We have confused text with knowledge.
"Only I think I'll read it in a client like NetNewsWire or NewsGator, not the web browser."
Why's that?
Because the browser isn't built to handle large numbers of sites comprised of microcontent chunks. Browsers haven't evolved to reflect the way people use the web in about 5 years, with the possible exception of the Google toolbar and Safari search box. IE on the Mac tried some work in that area a few years back, with the auction manager, but that was overly specific and too special a case.
I want a web browser with threaded history and smarter favorites, but I don't think that'll come from the browser makers, so I'll look to the client app developers who are innovating these days.
i'm hoping for aggregated bookmarks, which surely must come from google.
something like: a plugin that uses the blog api to synchronize your bookmarks with your website; thereby creating a dmoz scale/breadth continously updated peer directory. an internet-roll.
searching google from my own browser would give increased relevance to any link i've bookmarked, and 2degrees from me have bookmarked.
I definitely think you're on the right path with this line of thought, but the progression of 100 people you trust to 10,000+ is a bit problematic. First is the issue of what you call trust. If you don't have physical time to read 10k blogs, then you need to trust the flitering tools you have at your disposal. I imagine that the tools will evolve, but unless there's some unknown technology in its nascency about to burst forth then we're stuck with what we have now. I really don't think the "recommendation" software available now is worth using. It doesn't work the way I want it to- does it for anyone? The second problem is one of tangency. Just because my friend likes something is no guarantee I will. So how do I filter through the sheer volume of "Blogs I Read" lists to come up with the ones worthwhile?
When you say "many of us will be reading 10,000 weblogs" do you mean the "us" who already read a lot of blogs, or do you mean to imply that lots of people will give up on the Times and move to blogs?
I just wish you would update your RSS so that I could read your *FULL* post in Newzcrawler, not just the first few words. It drives me mad daily. Please? Pretty please? What can we do to convince you? Please?
Brewdog: As for trusting the filtering tools, it's not only a matter of trusting the tool. You could create and tweak the level or trust yourself, using XFML's "strength" property, for example, with which you personally specify how much you trust a source, and even further, lets you equate a topic you use with another they use. Taking content from someone else's blog(or whatever), you will generally have the tool lower the relative "validity" of their info/link by a factor of 1 and that of someone found via them by 2, and so on. Over time, you could find someone whose judgement you are confident enough in to give them a little more weight, such that their posts float to the "top." Subsequently, people they link to will get a little more weight, also.
Of course, I'm completely unaware of any piece of software that makes any sort of automated aggregating/filtering use of XFML, so there's your infancy factor, though the XFML home page does have links to utilities and templates for various tools such as Radio and MovableType. If it gets around enough, maybe someone will take the time to integrate it into an RSS reader.
I foresee a single RichClient PIM app that handles email, blogs, and every other medium of net-based communication (documents, tasks, etc.).
This best as a single app because that gives you
* a single logical approach/place for having agents assigning priorities to incoming traffic
* a single place to look to find all those high-rated items.
http://webseitz.fluxent.com/wiki/UniversalInbox
Call me a loser, but I'm waiting here with baited breath for the big point...
wat een moie website ga zo door
In case ytou hadn't realised, the above is a dutch language spam voor some weblog or other; saw it at other sites too.
Let's think about how something like this might work. A first effort might go something like this:
I use NetNewsWire Lite as my RSS reader. It keeps my list of feeds in an OPML file (an XML outline).
Another program could troll through that, and parse all the RSS feeds or related web pages (since not everyone publishes the full text of their blog in RSS, for the time being we need HTML as a fallback) to extract all the links. Each time a given URL is linked to (perhaps with some smart parsing to get at a blog rather than a blog post), that's a vote for that URL. If a given URL gets enough votes, rising above some threshold I set, it gets added to my OPML file "first-degree suggested reading". Iterate through this process again with all the found links instead of my original OPML file, and you get a list of "second-degree suggested reading".
I'm guessing this is something that a good PHP, Python, or Perl coder could knock out in 2 or 3 days, not 2 or 3 years.
It should come as no surprise that Mark Pilgrim has already done something sort of like this.
http://diveintomark.org/archives/2002/11/20/new_reading.html
What is my own take on this debate? As of now, I don�t have a strong opinion either way.
However, I do know that
1) I get almost all of my news and views from other bloggers or from magazines. I don't read general-purpose newspapers directly any more unless there's an article on a blog or meta-filter site (Google News, Samachar.com etc.,) that cites an article from a particular newspaper.
2) However, while blogging is very addictive, it takes a decent amount of time and effort to bring out a good blog regularly. Most blogs could do with significant editorial and proofreading makeovers. In the absence of any monetary remuneration, I�m not sure how many people will continue to invest the time and energy required to bring out a quality blog.