Open-Sourcing Conferences


Two weeks ago, I was on my way to Copenhagen, for a terrific little conference called reboot: http://reboot.dk/reboot7/. Now I'm in San Francisco, and writing from a conference called Supernova. Its theme is decentralization. I got in tune with the theme right away on the morning of the first day when, after driving 400 miles from Santa Barbara, my car was hit by a garbage truck three blocks from the end of the trip.

After we both pulled over, the truck driver got out and clearly felt bad about causing an accident that was obviously his fault. "Relax", I said. "I'm not hurt, and I'm driving a rental car." He seemed relieved. Credit where due: the agency, Budget, swapped my busted Ford Focus for a new Ford Focus (my favorite rental car), once again violating Searls' 4th Law, which says "No matter what car you want to rent, you'll get a Chevy Cavalier": http://www.linuxjournal.com/article/7888. What's more, they also proved that you don't need to buy the insurance the agencies always try to upsell you at the counter. I didn't buy it, they barely looked at my accident report, and they swapped the car with no questions asked.

My breakfast date (for which I was an hour late) was with a couple folks from Cyworld, the giant Korean online service. It was fascinating. Rick Kim, whose title is too long to repeat, told me a little about what's happening on the Net in Korea. In some respects, it's nothing like what we experience here in the U.S., or in Europe. In others, it's remarkably similar.

On the difference side, the carriers (notably telcos like SK, which owns Cyworld) don't believe broadband is a Done Thing, or that 3Mb/sec is a good maximum speed for broadband to the home. I asked Rick about the speed of service at his house. He said he got 88Mb, but that the more typical bandwidth for homes there was 15-25Mb. And going up.

Not long ago I talked with a cable installer in Santa Barbara who told me his company had no intention of increasing bandwidth, whatsoever, for the forseeable future. We're stuck at 3Mb/down and 300Kb/up, for the duration. No Moore's Law here, folks. Move on.

By the way, when we moved to Santa Barbara in 2001, we were getting speeds of 7Mb/down and 3Mb/up. They throttled back the bandwidth after their backbone provider, @Home, went belly-up. Now they use their own backbone, they say. Oh, and they block Port 80 and Port 25. Nice.

So bandwidth is one big difference. Another is the culture in and around online services. To my surprise, the first syllable in Cyworld isn't short for "cyber." It represents "relationship" in Korean. The company sees relationships with customers, and customer relationships with each other, as paramount concerns. Somehow I don't get that from MSN or AOL or Yahoo, much less from the big cable and telco carriers.

What's similar is the infrastructure on which they build their services. It's Linux. To give an idea of how much their Linux infrastructure supports, 90% of all 20-somethings on the Net in Korea are Cyworld users. Among the free services offered is unlimited storage. There are a few limitations, mostly having to do with technical issues around uploads form cellphones and similar real-world issues. But the space is there. More encouraging is seeing "linux" embedded in the email address of the head engineer.

Right now at Supernova, there's a panel ironically focused on attention. Linda Stone, formerly of Apple and Microsoft, just gave a talk about a meme that she delivered to the world a few years ago: "continuous partial attention". She had exactly that from the audience.

Between the last sentence and this one, I was called outside the ballroom by the head of a well-known "social software" company who wants to open-souce their proprietary goods. What's more, he wants to work on an open source framework for *all* social software (blogs, wikis, RSS aggregators, etc.).

Earlier a guy I was talking with had a cool idea for conferences like this one: Set it up like any other conference, with speakers, panels and so on; then announce at the beginning that all the speakers were just bait, and that the whole conference is completely open: anybody can learn anything from anybody. Bring up the house lights, arrange the chairs in circles, roll out the hors d'ourvres.

With no speakers, every attendee's expertise is a "source" for every other attendee. Conversation becomes the most efficient and effective means for moving and growing knowledge throughout whole crowd.

The idea here is a profound corollary to Bill Joy's observation that "most of the smartest people work for somebody else". The balance of smartness in any conference session heavily favors the audience. What's the most efficient and effective way for everybody to share what they know?

We've been lecturing at conferences for the last umpty years. Audiences have been opting out through schmoozing in hallways, hanging out on IRC channels, blogging, IMing and emailing each other. In other words, they're going to other sources of knowledge.

When I play the lecturer role, I try to be so compelling that the audience has no choice but to quit typing and start listening to what I'm saying. But even when I succeed at that, I'm merely improving on the lecture model. I'm giving one man's opinion. If the purpose of a conference is to increase everybody's knowledge, we need to find ways for everybody to be available to everybody. An any-to-any model, rather than a one-to-many model, is naturally far more flexible and efficient.

So, what we're talking about here is ... ta-da: Open sourcing conferences.

Next step: making them smaller and cheaper, with better UIs.