Raph Koster leads off, saying that "3D is a red herring," and that most of the people in virtual worlds at this point are playing games.
Rueben Steiger from "millions of us" says that "people like 3D environments," that we "think in 3D" and thus the 3D internet is inevitable. The numbers now aren't important, he claims.
Clay Shirky says that the aspect to look at is game vs non-game. Games have led the way in adoption of virtual worlds. We're not headed towards general purpose virtual worlds, but rather towards specific implementations for specific purposes, and games are a prime example of that.
Raph points out that 3D, 2D, text is not the issue. ("It's all bullshit. It's irrelevant.") We've had text-based vws for a long time. If there isn't stuff TO DO, people go away. If there is stuff TO DO, they stay. If there isn't something useful TO DO, that's unique to that environment, it won't work. (Chat won't keep people in VWs, for example.)
(This seems to me to be very accurate for adults, perhaps not as much so for kids.)
Rueben claims that SL is a game, because life is a game. It's just a really really broad game. The inflection point for user adoption is around 3.5 hours...before that, attrition is high, after it's low. Funneling people to compelling content doesn't do it. Having them meet people does.
Clay says "I've never in my life bet against the users." The users are getting left behind by SL. The attrition both short and long term is significant. He says that games are cognitively special, and that game-like situations cannot be trivially imported into work (or the reverse).
Rueben says that registered users isn't the important statistic, it's number of hours of system use...which has been going up.
Raph calls bullshit on all them. Hours of use are measures of devotion, but not measures of adoption. Need to ask "of recurring users what percentage are coming back month after month?" There are industry standard methods for measuring this. To industry insiders, the LL numbers look wrong.
An audience member says that he wishes the panel would quit talking about Second Life.
Rueben proposes a grid with two axes--social to entertainment, and 2D to 3D.
Clay notes that we have seen over and over again that communication cannot be a perfect substitute for transportation.
Raph does a quick history of social virtual worlds. Too hard to encapsulate. Says that the 3D worlds folks seem to think they're going to "swallow the web"--and he thinks they're wrong.
Some discussion about federated vs scattered identity. Identity in a utility form is not the same as identity in an avatar or personality form.
Raph points out that virtual worlds and virtual reality have almost no overlap in practice. VW designers are not informed by VR research. What matters fundamentally is emotion, not representation; communication, not representation. Making representation better isn't the point.
An audience member asks if Facebook is a VW--Raph says that the dividing line is whether there's a modeling of "space"--is there north, west, east, south in Facebook?
Clay: screens are not the path to simulating being in a room. "Jaron Lanier is the Charles Babbage of our generation."
Time's running out, and I've left a ton out. This was a five-star panel. Smart, funny, articulate people who really know the field. One of the reason my notes are sketchy is because I was really engaged in listening...