Christopher Allen is the founder of Alacrity Ventures. Ray Ozzie works as CEO of Groove Networks Inc. Esther Dyson is chairman of EDventure Holdings, as well as an editor at large for CNet Networks. And Mena Trott serves as CEO for Six Apart. What follows is a partial transcript of their Supernova panel discussion:
Christopher Allen: How should the tools evolve to take advantage of the subtleties of human behavior? We want two people to use one machine. We want people to share data. That's all technical. We need to make our software and processes more adaptive to how groups work. We dont have good models of how groups work. One thing we have is the Dunbar number. Robin Dunbar found that the optimal group size for the human primate was about 150. For hunters and gatherers, that was sort of a relative max. And even though we use the Dunbar number, 150 is a large group. It might work for survival-oriented groups, but for non-survival groups, that number can be a bit smaller. In fact, the most effective size is about 60. We may have gone beyond our normative size. We've gone beyond our capacity to trust.
Esther Dyson: There's been a lot of buzz about social networking and social networking tools. Human behavior is a lot more fun to talk about. In some sense, you need human rules and protocols, and software need to support them. More choice should make us happier, but often we don't choose what makes us happier.
Most of these social networking tools won't survive as businesses. Almost any function will be part of some other product. They may survive as functions within business, but they might not survive as such. Things like LinkedIn go way beyond the Dunbar number. It's akin to friend inflation. You get a kind of fake niceness. They need to lose their buzz and go away, and people need to focus on their true networks and circles of friends.
How do you make the tools better? If you take the calculator and the corporate financial system, the thing that was in the middle was the spreadsheet. It gave individuals the opportunity to create formulas. There was a grammar. Imagine something like that for social interaction. In this spectrum, email has no power at all, and then you have complex corporate workflow applications.
Kevin Werbach: Esther wants a social spreadsheet. Mena, do you have it?
Mena Trott: Um. Esther mentioned niceness. At the other end, you have extreme animosity. At Moveable Type, we recently revisited our fee structure. It wasn't the most pleasant experience. But the people it affected, Moveable Type users, reacted like you react when you didn't get enough warning. After making the announcement, we turned on Trackback and had 800-900 trackbacks that ranged from "I'm not sure about this" to "I can't believe Mena and Ben would betray me like this." People also said, "I can't believe they left trackback on. Did they forget to turn it off?"
We needed the feedback. If we had open comments, it would have been worse because it would breed anonymity. The accountability of a Weblog is that it's your online personality. If you're going to write a nasty one liner, you don't want it mixed in with their content. We got really long, thoughtful essays. Really useful feedback. And now we're closer to getting it right. As we go forward and the tools evolve, Weblogs will evolve from something you publish to thousands and thousands to smaller groups. In TypePad, our hosting service, 30% are restricted, and 10% are private. We have families talking to each other.
Kevin Werbach: Is there a break in the continuum where we can say this is social software and this isn't? Or is all software social?
Allen: Let's go back to the '80s. The biggest software categories were spreadsheets and presentations. The real power of the spreadsheet is that you can persuade someone. You could say that there's very little software that's truly single-user software.
Dyson: You could say they persuaded people, or you could say they beat people into submission. The space shuttle report blamed, perhaps unfairly, PowerPoint. These can be used for communication, and they can be used for obfuscation.
Social software can help clarify things, but they can also be damaging. When I was working for Ziff-Davis years ago and first published an org chart, there was pandemonium. Or perhaps you think, "I wrote to my boyfriend 10 times, and he only wrote me three times." It can be damaging to have all of these reminders. Social software can make things visible more than we really want.
Trott: In the case of the trackbacks, these weren't referrals, people explicitly posted something saying, "We know you're going to read this." That's not publishing, that's communication. People knew they had a way to reach us. That affected how their audience read them, and it affected us.
Allen: It's much easier to make a negative comment than to make a positive comment. Trackback as a medium enforces you to have a higher-level form of discourse. There's more incentive to explain yourself than an anonymous email. Some of these media encourage positive interactions.
Ozzie: Our tool is default private rather than default public, and there are all sorts of fascinating dynamics we've documented about the fabric of conversations. The nature of the network is a meritocracy rather than what you see in most organizations. It matches business practice rather than the organization as constituted.
Update: Allen's presentation slides are also available.