Interview with Second Life creator Philip Rosedale

Ian Lamont — Fri, 30 Jan 2009 10:11:23 -0500

Philip Rosedale is a virtual world technologist and entrepreneur who launched Second Life in 2002. He currently serves as chairman of parent company Linden Lab and also leads several Second Life development projects. The Industry Standard interviewed Rosedale earlier this month in San Francisco, and talked with him about the state of the industry, Second Life's growth curve, and the potential of for 3D cameras and other cutting-edge technologies to change the way people interact online (see our interview with Linden Lab CEO Mark Kingdon here).

The Industry Standard: Second Life has been around since 2002 or 2003.

Philip Rosedale: Yes. I think 2002 was when we put it first out.

Industry Standard: Now, take the first 1000 users. If you are looking at the community then, and comparing it to now, what are some generalizations that you can say about how it has evolved or how it has changed?

Rosedale: One thing is that the first 1000 users were probably more like each other than any 1000 of the current million or so more active users are like themselves. Do you know what I mean? There is a lot more diversity in use, demographics and behavior in Second Life today than there was, say, at the end of 2003, which is when I think we were at about 1000 who were using it.

If you looked at the end of 2003, you saw the set pioneers, people who were inevitably more technical and more pioneering. They had more time on their hands to invest in the world of Second Life. I think there were probably just a few... If you got them all together in a room -- and I've seen this happen at our user conferences and the community conferences that people held -- they over time look less and less like each other.

The use cases -- you've been involved in reviewing them. By the way, it's so great to see somebody seriously evaluating a product that was itself built on top of Second Life. It is such a delight to me. I like that. That does look good. I mean, that was the first time, at least in my memory that that really happened.

Industry Standard: You're talking about Immersive Workspaces?

Rosedale: Yes. I think that, if you look at the diversity of use cases today, it is a lot broader. In the beginning, and I think there is a path that all new technological mediums follow. I think, the path is always the same in terms of who the users are in the beginning, middle and end. In the beginning, you always have novelty, art, socialization and early entrepreneurship driving the use of the medium. I think that was true for the Web. I think it was true for email. I think it was true for instant messaging. I think it's true for text messaging. As people begin to use that a lot more.

Industry Standard: For text messaging, art?

Rosedale: Yes, I think that people goof off. I guess what I'm saying is you don't use text messaging yet as a sort of reliable form of business communication. "Ian, are you here on time yet?" You know what I mean? It's still when you get a text message it's that weird feeling. You feel like it's kind of social, or sexy, or it's offbeat somehow. Artistic, novel, you know?

It's fun to send people text messages in meetings or just at times when you think they wouldn't be expecting it. And I think that's art. You're basically just playing with, you're making novel use of the new medium.

And I think Second Life in the very beginning was a lot like that. Everybody was like, "Huh. Well, can I make myself into a skeleton?" You know? Well, let's see, can you put the pieces together to do that? Or, "can I just wander around and talk to people?" Socialization. I mean, wow, there's just these other people.

And I think that new mediums always go through that, the first stage being that. Then the second stage, I think as people start to understand or trust the medium a little bit more, and I think this is somewhere around kind of where we are today with Second Life is you get like students and educators because they don't have CIOs to convince. They don't have as much of a barrier to widespread adoption.

So, you see educators and students kind of jumping in and saying I can use this a bit more practically. So, students were emailing their class notes around to each other in 1992. Nobody else was using email in 1992, but students were. Especially here in California, it was quite common.
The UC system had, I just remember that for me was the very first Internet experience, was really long before I really actively started programming on and understanding like TCP/IP. The very first thing was I remember I was in college and we had these terminals, rooms with terminals. Where we'd go get online resources for classes.

But then, somebody was like, no, no, no, if you go out "you can do this -- email Chris @something.edu." Remember? And you know you type in a line and it's like that's weird, that's bizarre, I can send a message. So, I think that the educators are the next ones to kind of discover the medium.

And then about the same time after that you have more robust early entrepreneurship. So, like people making money in Second Life, which is really starting to happen now. And then like groupware. Like, you know, businesses using it. But, they always use the new technology initially as kind of a group accelerator because you can usually do that under the radar of the IT system. You can do it in a covert way.

So that would be like the way businesses use text messaging today. If you look at how we use text messaging internally as a company, we don't have an official text messaging policy, but we just for the first time put text addresses like phone numbers into our personnel tool. Obviously, we're pretty avant guard as a company, but you're seeing people text messaging.

But, it's a way to make work a little bit better or more efficient and that's Immersive Workspaces. Right? I mean, actually Immersive Workspaces I think is a product maybe is a little bit more positioned as a bit more of a top-down thing.

But, if you've got somebody who's like, "Dude, let's not use Skype because there's five of us, we can sit around and look at a Web page too in Second Life. And the sound sounds just as good as Skype only better because it's 3D."

So, you have this little advantage that it delivers to a work group. So, I think we're in the middle right now of the beginning, or we're somewhere on the curve, of work group collaborative adoption and education adoption. So, that's like stage two or stage two and three or something.

And then the one after that is going to be more formal widespread business use, where it becomes more like a competitive imperative. So, that would be like when businesses really started adopting Windows NT broadly in the infrastructure. You just kind of move up to this point where ... Or Oracle or something, right? The CTOs or the CIOs were saying, "Yeah, we really need to use this database system because it's giving us competitive efficiencies."

So, I think that stage is totally yet to come for virtual worlds. We're not there yet at all. And then the one at the very end of the line, and then you've commented on this in your blog I think too. The one at the very end of the line is the one that everybody fantasizes about in the very beginning which is what I would call ecommerce, which is the final use of the new technology medium as a way of directly selling to the end user consumer of goods and services. So, that's Amazon.com in its full glory, which it's only been in for three or four years where it's really been at scale. It's been a significant commerce player in the United States.

That only happened after the Internet had matured all the way. And if you look at the earliest use of Amazon, it was that early entrepreneurship stage one, stage two thing that I was talking about, which is Jeff [Bezos] and somebody saying, "well, wait a minute now, it wouldn't be too hard to sell a book on there." You wouldn't need to see that much about it.

And that's like some of the things people are doing today in Second Life kind of looks like when Amazon was launching. So, I think you go through these stages.

Industry Standard: Let's talk about stage three when it becomes something that's pushed out to all the users and the company. What needs to happen for Second Life for that to really occur? I mean besides the fact that people are telling each other about it and people are becoming aware of it. What needs to happen?

Rosedale: Well, gosh, a lot of things need to happen. I mean, like account creation, like the way you sort of get an account set up to go into Second Life. That experience needs to become standardized and normalized in a way that works for businesses as well as it does for end user people that are signing up today.

Obviously, as you're probably well aware from your own hands-on experiences, the orientation process, the first hour. And I'm sure Mark talked about this too, but there's this utility function in Second Life, or actually with anything. Where when things are still in the early adopter phase or late early adopter or early majority or whatever, they're in this problematic space where sometimes the cost of getting acclimated in the new media exceeds the benefit associated with the particular experience. So, like have you ever gone to a live music event in Second Life?

Industry Standard: No, I haven't.

Rosedale: So, it's one of the coolest experiences, I think personally. I mean, it's really amazing, right? You go into somebody's nightclub, it's clearly a venue. There are signs up. They've talked about how this or that performer is going to be there tonight.

The person is standing up on stage, they're singing and playing live, they've got an avatar. There's all these people from around the world obviously sitting next to you. You're chatting with them which doesn't distract, so you can chat with the whole room, but you're still listening to this performer. The performer's shouting out stuff, you know? It's just a totally surreal, empowered experience. It's just amazing, right? It's going to be a big deal in the future.

So, why isn't it that [San Francisco music club] Bottom of the Hill is going out of business because basically everybody just does this in Second Life now, live performance? Well, the reason for it is simple. It's not that the economics for the musician aren't better in Second Life, I actually think they are. They're a lot better.

I mean, we've got a bunch of musicians that work here, and we laugh about it. Because why on Earth would you go try to get a gig at the Red Devil Lounge on Polk Street in this, one of the world's good music cities? This is a good place to be as a musician.

But, there's no way it makes sense for you to do that as compared to performing in Second Life. Because in Second Life you still are getting 25-45 year old people listening, but 65 percent of them are international.

I mean that's just a huge. You can't argue with that. You really as a budding musician you want to hit a large audience so that they'll get the word out. And the fan out of a group of people in Second Life just crushes what's going to happen anywhere in San Francisco.

So, there's an economic benefit to the performer, but there's not yet a utility function for the attendee. I mean the problem is, right, if I told you "some band that you kind of like is playing at Second Life tonight," are you really going to go see it?

Industry Standard: It depends if I had the time.

Rosedale: That's the problem, right? I mean, how many hours are you willing to spend prepping for an experience which is listening to a band? How many? I mean, 30 minutes. You call or go online to get the tickets, call your friends, meet. So, we have to get the experience of logging in to Second Life and getting to that event, after you got like an email or an IM or something from a friend, down to 15 to 30 minutes.

And today, the statistical estimate that we kind of throw out there is several hours. So, for you to really get acclimated and get to a live music event is several hours away. So, even if you total the hours, so you're probably not going to do it yet.

So, going back to your question, we need to collapse the orientation experience on learning the interface down to a 30-minute timeframe. And we're not there yet.

Industry Standard: Is that a question of improving the way that you're educating people or actually changing the interface?

Rosedale: I think it's both. I mean, I think that the experience of which things you're walked through and what the essential elements of what the initial user experience are, which is a narrative content sort of thing. I think, we definitely have to change that. There is abundant evidence that people are trying different experiments. You look at them and say, "Hmm, one of these things is going to finally go."

But, I also think that the basic UI of the software also needs to change. It has too many pixels. People are not smart enough to absorb the amount of information Second Life throws at you. It's just daunting.

Industry Standard: What do you mean by too many pixels?

Rosedale: What I mean by "too many pixels" is that there are too many letters, numbers, buttons, and symbols on the screen at any given point in time for you to really understand. And they're all kind of demanding your attention -- your [Linden] dollar balance, your inventory window, all the buttons on the bottom bar, chat and text that are visible in the window, that's asking something of you, blue pop ups that are coming up. All of that stuff taken as a whole at any given moment is pretty overwhelming.

So, I think that -- and we're very much engaged in this -- there's a design process here and we're whittling that down from ... I mean, it's kind of like going from, if you remember the first MP3 players, they had a lot of buttons and stuff. But, at some point, everybody realized "Well, this one you don't really need and this one you can hide until you have to do this one." So, we can kind of get it down to a volume control and track selector. You know, simpler. So I think that there are just some basic work like that that we have to do.

And it isn't easy because you can look at all the other products that are out there and some of them, even though they have a different reason for being than Second Life does, like games or like [Google] Lively or IMVU or Gaia [Online] or [3D] environments that have a sort of similar final amount of complexity in the interface, like what you can be able to do.

You can see people doing experiments on this. And the only thing I would say is that, from watching all those products and ours too is that it's not simple. You know what I mean? It's not like a no brainer to just distill the complexity of the interface into two or three simple steps.

One of the reasons that virtual worlds are growing linearly and not like hyper linear, like Facebook or something, is that it's just pretty hard to solve these design and engineering problems.

And the other thing is that there's kind of a "meet in the middle" thing, I think, where I think - and I was alluding to this earlier - I think some people look at virtual worlds and say, "there's way too much going on here." They look at Second Life and they say, "99 percent of this people don't need." I think that's dead wrong.

I think that if you build -- because I think the experiments, like [Google] Lively as an example, there are experiments of plenty where people have said, "You know what? I don't think it's about creating content. I think it's just about exploring some rooms that your friends build from catalogs of 100 objects."

The UI for that is so much simpler. There are only 100 couches to choose from. Awesome. Right now, I can put them in a little poster view of 10x10. When you hover them I can skin them. I mean, I can do all sorts of cool stuff with that interface. Great! When I want to move the couch around the room, since I know there's only a type couch, type chair, type stool, I can build code around that and custom build. I mean, I move that couch and it slides around and it's so cool how it automatically orients itself.

You look at that and you think, "That's so much simpler than Second Life." That will run. That will get people right in there. They'll start using. They know.

Industry Standard: Now, I'm going to bring up an example from something we discussed earlier, and that's Webkinz. Webkinz has this type of format where you can choose only certain -- I mean, they have an inventory. You can't customize anything.

Rosedale: I've never seen Webkinz, but I bet it's like Club Penguin.

Industry Standard: I think it is.

Rosedale: It's very similar. Club Penguin has a big inventory of stuff kids can buy.

Industry Standard: Right. And despite the fact that it's limited, kids aren't really creating, although they do some interesting things on their own. It's proved to be a runaway success. And I've seen other virtual worlds like There.com, as an example, where it's very much simplified. At the other end of the spectrum, you have World of Warcraft, which has probably one of the most complicated UIs I've ever seen, yet it's also extremely popular.

Rosedale: Well, I think there are different experiments in different areas. If you look at There, it has less capability than Second Life and it has not quite proportionately, but it also has less users. It's less successful. There are things that There does that are unique and cool and they've gone after different verticals and stuff, and I'm not commenting on that. I'm just saying in the mainstream, it looks like a slightly less capable version -- not slightly, substantively so it's a good comparison -- less capable version of Second Life in terms of what kind of content you can build that is also substantially less interesting as looked at as both the mass scale of the adoption and the demographic. The demographic is tighter, narrower, less diverse. Our demographic is bigger and more adverse. So, that just looks like nothing new being gained there.

If you look at Webkinz and Club Penguin, what I think is happening there is a different problem, which is what I said about my son, which is the half life of use is short, but the number of prospective users is gigantic. And so it's like a wave that's gaining momentum and size and it's rolling its way all the way through. Every kid in the world will use Webkinz once. But, wait three months and see if your child is still totally sort of following along.

... The other thing, I think, which is something that I talk about a lot and think about a lot, is that I really do believe that once the usability problems and the initial experience problems are brought down, there's going to be an inversion of how we think about virtual worlds, where we realize that virtual worlds are potentially easier to do things in, especially for people that have less - I'm not going to say "education," but are less Web educated, that have less text...

Industry Standard: Less Web savvy.

Rosedale: Yeah, less Web savvy. I think that the less Web savvy people, which are still the great majority of people on planet Earth, will actually go to virtual worlds first, because they're simpler.

Industry Standard: Really?

Rosedale: Yup. I think there are a couple reasons for that. I know that's a bold claim. It's one of the things when I get up at conferences and I'd say, "Look, I'm telling you, in 10 years time..." If you look today at Web router traffic, if you look at a tier 1 router today and you say, "What's going through it?" It's about 70 percent Web traffic and then the rest is email. So, you basically have the Web as the dominant use of information technology on the global network. That's what we do, we use the Web.

The second thing we do is send email back and forth. And those are basically in round numbers and everything else, is smaller - obviously, still in many cases, big businesses -- but that's the whole thing. That's the frequency (muddled).

I believe that in a decade's time, the number one spot will be virtual world traffic. It will basically be the act of moving around in and consuming content in a virtual world, in the same manner that we see the 70 percent traffic on the Web today.

I think the Web will be -- this is the mystery of super-linear growth conditions -- the Web will be a lot bigger than it is today. I'm not saying that the Web will be smaller; I'm just saying that in the same way that we've seen all of these exponential curves on new products, I believe that virtual worlds will ultimately get on an exponential [curve] where they attract users at a rate greater than the Web. So, I think there's going to be this inversion where virtual worlds are going to be number one, the Web will be number two, and email will be number three, in 10 years time.

The reason for that is that - think about it - it's a couple of things. The main one is that human beings generally try... I look at all life experiences as being: information consumption, creation, or sharing. It's like "The Matrix," right? You can say that it's all digital. All we're talking about, back at that camera, our conversation is basically just digital stuff being exchanged. So, what we're doing is sharing information back and forth. When you surf the Web, that's what you're doing. You're generally consuming information.

If you look at human beings and you look at their day to day activities, they generally enjoy and optimize for the creation, sharing and consumption of information in the presence of others. It is just generally the case that we're not alone. We spend less of our time alone -- not none of our time alone, we certainly spend a lot of time alone -- but if you just look at the average human, they spend a majority of their time in the presence of others.

The Web and the Internet, to date, do not afford us the feature where we do things in the presence of others. And, "in the presence of others" has a very, very, very, specific meaning that relates to how the brain works, which is within the one-second delay range, where something I do is visible to you and you can respond to it in real time, which for the brain is in hundreds of milliseconds. So, Facebook does not fall into that category, is what I'm saying. Facebook is incredibly powerful and important, but it is not the same thing. You don't have presence. When you go on Facebook, it's not like walking into a bar. You don't log onto Facebook and go, "Ooh, all of these people around me!"

Industry Standard: Well, you can see at the bottom of the screen how many other people are online at the time.

Rosedale: But, you don't feel them. You can't talk to them. You can't say, "Hey, you guys! Who's on here today?" or, "Who's near me?" But, various technologies, and Second Life is one of them, create the ability to have presence, to have the presence of other living people around you in the experience that you're having with information. So, shopping is actually a good example of that. We do a lot of shopping online now, but we still do a majority of shopping in places that are populated in real time, by other people. There are reasons for that. Sometimes it's our friends, but sometimes it's not.

We also use the presence of other people as a cuing mechanism and a guidance mechanism in our day to day, walking to the market. There's a tremendous amount of stuff -- asking for directions, as has been joked: the difference between men and women. All of those type of activities, we take advantage of other living humans to get from information to information. The Web doesn't give us the ability to do that.

Now, it may be that somebody beats Second Life by coming with a kind of Web layer that makes everything live, but the example I always give is: when you're on Amazon.com, shopping for a camera, wouldn't it be nice...? There are 1,000 people at the same page as you, even at the same subpage, even at the same camera. There are five people looking at the same camera, right now. Wouldn't it be incredibly cool if you could talk to them, even just texting -- which I'd find as being adequately present if you can go, "Hey! Is there anyone there?" Enter. Yes, that's fast enough to be "presence." So, if you could talk to those people, you could refine your search. You could say, "Has anybody bought one of these cameras?"

There's stuff that you'd just naturally do that you tend to do even faster online than you do in reality. You won't walk up to and tap people on the shoulder -- some people will, tap people in the store. Online, almost everybody will tap somebody on the shoulder and ask a question. So, you have this accelerant of people being there with you.

The other reason why I say it will be bigger than the Web is just that the interface, once we do solve the design problems, you can get some brilliant 80 year old who's never seen the Web and put them in front of a computer with Second Life, and once they get over that initial, admittedly, "crawling over broken glass" hurdle of figuring out the interface, they're done. Right? Because they know how to walk, how to talk to people, how to follow a street, how to read a map, all of those things are things that they grew up doing.

But, if you put them in front of Google.com and say, "Oh, yeah. It's all there, just start typing," that's like a PhD that we've all gotten in the last decade. What we're missing, or what we gloss over, is the semantic complexity of how we navigate those Web pages, how we type in that text to Google, and how we follow those links and that we understand the difference between Craigslist and LinkedIn. That's a semantic complexity that some guy in India, who's a grand master chess player, but has never used the Internet -- I just don't know that he's going to go through the training process of multiple years of learning what all of those semantics are.

Industry Standard: On the official Linden Lab blog, there is -- I think probably your CTO -- a comment about tiered storage for Second Life. There are [3D] objects that need to be called up a lot. Maybe they are newer objects or things that are used a lot. Things that might not be used as much, if they are old objects that people have created. Can you talk a little bit about that?

Rosedale: In general, obviously, we let people build really as much as they want and store it in their own inventory. Then, what's resident in the world and viewable by other people is a subset of that data. I think what is resident in-world is on the scale of, maybe a little less than a hundred terabytes of data? By the way, that is just a staggering amount of data.

That is one of the things that makes Second Life so special. The amount of stuff that is actually in people's potential inventory is two or three times that. I think I should look that ratio up for you. I think the ratio is even larger. People's willingness and desire to create content is pretty much unfettered. I think that one of the things that Second Life shows about human behavior is that we are all a lot more creative than our current real world environment gives us opportunity to be or credit for.

In entertainment, there is a conventional maxim. When you talk to people who start start up companies, they're trying to do things that entertain people, there is generally this assumption that has gone into, I think, many, many companies over the last couple of decades. They have said "well, a very, very vanishingly small percentage of people are interested in being creative as a form of entertainment." We would say that being creative in some manner is actually itself entertaining.
The conventional thinking has been that almost no one has that bizarre property. It is much rarer than being left handed. It is extremely rare, and the majority of people want to passively consume content. The Web, in its sort of original bibliographic form, television and books are examples of that world. We kind of take a somewhat user driven, but largely passive role in consuming content. I think that there is a misunderstanding there. I think it is a cause and effect thing.

I believe that human beings are intensely entertained by creativity, whenever they can find a venue in which they can actually do it. Unfortunately, but understandably, the way technology has structured the world, particularly over the last few decades, has been to make it very, very easy to broadcast content, meaning that we just have no choice but to largely be in a passive or consumptive mode.
When Second Life was really small, and there was only a thousand people using it, we would look at what percentage of people's time do they appear to be noticeable? It's hard because Second Life is so open ended. We would look at the data and try to measure, "what percentage of people's time are they spending creating versus consuming?" It was something like 30 percent creating and 70 percent consuming.

We said to ourselves, "well, surely as Second Life matures and a more normal mix of people, so to speak, are using it, that 30 percent is going to go down to one." Because, that's what everybody says. In some sense, we'll know that the usage of Second Life is mature when that number has dropped to one.

What we saw, then, over the last five years, was that the amount of time that people spent being creative really didn't drop very much at all. In fact, I don't even know what it is now. We stopped looking at that as a useful statistic. In fact, it seems that people want to spend a large percentage of their time being creative. It is almost like, if you see the number dropping, you can almost assert that the Second Life toolset is too hard to use, not that people don't want to be creative.

I think that there is the great thing about virtual worlds in general, which is that they finally give us the opportunity as people to discover what the balance is. How much of our environment do we want to create? If you look at the typical avatar of a typical house, for business, or anything else in Second Life, what you see is way more diversity of content than what you see in the real world.
What does that tell you? It tells you that, given the choice, people will continue to customize their environment to a point where it's almost one to one. It's almost like everything is a one off artwork. That would be the limit case. The real world today is a case that is driven by the economics, distribution and manufacturing conditions of the real world, which has it that most things we have in our house have been replicated thousands or millions of times.

I just think that the conventional thinking that brand drives people to really want the same bleached, stark chair that I see at all my friends' houses. There's not bleached, stark in all of them. But, the same thing from Wal Mart, or the same thing from that Swedish company ...

Industry Standard: IKEA.

Rosedale: IKEA. Again, the thinking that that's what people want, in their hearts, is somewhat wrong. If you look at Second Life, it really proves that. It's not actually what people want. People want to be quite distinctive and creative. The environment in Second Life demonstrates that that's true.

Increasingly as it grows, and as business use grows and everything else in Second Life, it's hard to argue that it can't be. It's got to be a reasonably normal cross section of people, and increasingly so. And so it's uplifting for me to see that little creativity space.

Industry Standard: Let's compare your virtual world vision with the way the Internet was, probably 15 years ago. A lot of people, and probably you did, too, started on something called Gopher. It was like a menu driven way of navigating the Internet, and that limited, how many people could use it or how many people might really realize the power of [the Internet]. Then the Web browser came along and that's changed everything.

Are we at a stage, right now, with virtual worlds, where there is some invention or application out there that we need in order to reach stage three or stage four that you were talking about?

Rosedale: Yes, there might be. It might be something about search and the interface.

Industry Standard: Could it be a hardware thing, like an input device?

Rosedale: It could be. I spent a lot of my life working on that stuff. Did you know that? We worked, initially when we started the company, on this thing called The Rig, which is a hardware interface device that is really, really cool. I'm super into interface devices.

Industry Standard: Are you still working on that now?

Rosedale: Yeah. Not right now, it's all packed up right now, but I could explain it to you and we do little demos of it sometimes. It's pretty cool. It's basically an immersive interface where you get that dreamy VR thing of having your hands in the virtual world. You can look around, but it doesn't use gloves and it doesn't use head (muddled) space. We came up with something else and it's not the Matrix. It's not invasive! [laughter] You don't have to part with your brain stem. I'll explain it you. It's a longer conversation, I'm sure.

It's based on force feedback. I'm sorry, not force feedback, but something (muddled) or something, but it's based on a very interesting idea that it was my idea from 1993. We actually have a patent on it.

If I take this Ottoman and I put these little things that are called strain gauges on it, I don't know if you know what these are?

Industry Standard: No.

Rosedale: They're very, very cheap. It's kind of like the Wii; it's just such a clever idea. They're these little pieces of almost tape that cost nothing. The circuitry associated with reading them is totally free and it's 25 cents for the circuit. I can put these little pieces of tape -- this is a bad example, we've purposely built them, I'm just giving you the example. If I put my hands on this and I can't move, like this thing is really heavy, but I can try to move kind of like when you use a track-point mouse. I do that with both of my hands, my whole hand.

It turns out that if I put the screen in front of me and I render my avatar's hands, it turns out that I cannot move my real hands at all, that all the latency and nausea on this that is associated with VR, this goes away. What happens is my brain sees the hands and then my hands are here and I go, "Turn." I try to go like this, but of course I can't move my hands because it's a glove or whatever. I'm just not going to hold on to it too long.

If I tell my brain, "Turn your hands up like this." You see the digital hands go (motions with hands) and your brain, because the strain gauges can tell how I'm pushing on the chair. So, even though I don't have to move the chair, the strain gauges can detect easily how I'm distorting the wood and they can read. You can do a bunch of math and you can go all the way back to what my muscles must be doing. There's a mapping between what my muscles are doing and how that wood is bending. You can just compute it. It isn't easy but it's not a big deal.

Industry Standard: Now, you mentioned the Wii before. You have also the iPhone which has an accelerometer in it.

Rosedale: I was dreaming about that stuff when I was in college. I was like, "When is somebody going to a make a massive accelerometer?" And I mean, everybody was, not to say that I was always inventing the idea like, "You've got to a chip that can do acceleration!"

Industry Standard: Is there an opportunity, these types of devices and touch interface?

Rosedale: Multi touching interfaces are certainly fascinating. I actually think 3D cameras is one of the things ... we actually have all of the camera prototypes here as well. Do you know what those are?

Industry Standard: Is it looking at you face and then ... ?

Rosedale: Well, it's a really cool thing: there's three companies Canesta, 3DVR, I may get them wrong, but there's like three or four companies that are doing this. They're all competing.
What they have is a camera. So, their goal is to replace the little camera that's in the edge of your laptop screen that's looking at you with a camera that's a little bit smarter. And it makes a big difference.

The "little bit smarter" is it has a radar. So, every pixel in the image that it generates has a fourth pixel which is the distance to whatever it's hitting. So, it's a depth map, a "z-map" of the thing.
It turns out that's the biggest deal ever. Do you know why?

Industry Standard: Why?

Rosedale: It can't find my hands. It takes [artificial intelligence], formal AI. Like, if you're a computer and I say, "I want to build this interface where I can go like this and grab a virtual object or do multi touch without even touching the screen, I can just go like Minority Report"? (waves hands around) It's impossible because the problem is, if my face is in the background, finding my hands ... It's not impossible, but it's basically I need two gigaflops of continuous computing power to find my frickin' hands in the image. It's a very hard problem and our computers are not fast enough yet to do it.

I'm saying it's almost at the edge of AI. It's like pattern recognition that almost requires a human brain to find a hand.

Industry Standard: Is that a Moore's Law issue that will go away in five or ten years?

Rosedale: Yes, it will go away in five or ten years, but it's going away right now because those cameras -- now imagine, all you have to do is adjust the depth filter.

Industry Standard: So, it only gets stuff that's one foot away.

Rosedale: Because the camera's here. Yes, it's just that everything that's more than a foot and half away is painted black. Now, only my hands are left.

Now, any idiot can write a program that tells, for example, which way a rectangle is turned. That's long axis short, axis angled between the top and you've got it.

And then it can also see distance. So, I can take a door in Second Life and just push through it, or I can move like that.

So, I think there's some very interesting opportunities. I think it's possible to -- what we see, multi touch, the iPhone -- in a few years time, [they] will become these hand interfaces. Maybe they'll use pressure, multi touch hand pressure or something. It would be very cool. Maybe we'll touch the screen, I don't know. If you touch the screen you get them dirty, the big ones, but you know, Microsoft's always experimenting with that.

Industry Standard: For 3D cameras ... you've read Snow Crash before?

Rosedale: Sure.

Industry Standard: There's a character in it who makes her mark by developing very good facial expressions.

Rosedale: Juanita.

Industry Standard: Right. And for these cameras, is that another potential killer app for you guys where you can really have real expressions?

Rosedale: Actually, the thing I'm debating working on right now, because I'm now one of our team leads -- part of my job, because I'm back doing development more, I'm working on a map right now. We're about to release some simple speed ups to map and stuff that are kind of cool.
It turns out that I think the highest impact part of that will be for meetings in Second Life. Making them just crush video conferencing once and for all. I mean, that's what we're experimenting on.

Industry Standard: Mark [Kingdon] was also talking about this, too.

Rosedale: Upper body matching between your avatar. So, if you've got a MacBook and you've got the camera, the thing I can do -- I can show you a demo of this, if my little team chooses to work on this -- I can show you a demo of this in probably less than a couple months. It is trivial now to read that camera and not to find hands, that's hard, but to find your upper body and your eyes is very, very simple.

So, what that means is that I can easily detect head nods, maybe shoulder shrugs, definitely postural changes in the upper body, definitely lateral head sway and stuff like that.

What that means is that if you have an avatar in Second Life and you got to a business meeting and it's six months from now (my engineers hope less than that) and you just fire up Second Life, that little light's going to come on your MacBook. That's not going to be video, because everybody hates that. It's going to mean that your avatar suddenly is doing this.

So, if you think about it, of course it's fun and you're a little biased. It's delightful to come and sit in our little cool meeting room here, but we would do these kinds of interviews in Second Life then, right? I mean, it would be so easy.

It could also see gesture, in which is a little bit harder. If I can emote as an avatar like that, like if I can nod, if you can nod, if you can follow each other's conversation like that? That's huge.

Industry Standard: What's the timeline for this?

Rosedale: We could do it in two months. It can be done. There're open-source programs, there's open source code that refines your eyes and is flawless, 30 frames per second. It's literally done, it's just finished. It's just a matter of assembling the parts.

Industry Standard: What about the mouth?

Rosedale: The mouth is a lot harder. The mouth and the other 60 facial muscles is a big problem, however, I've heard -- I'm not as familiar with the latest work on that -- but everybody's pushing on it and that one is going to go the way by Moore's Law.

Because once we have your eyes, which everybody can do now -- the eyes are just unusual looking so they're very findable in the image unless there's very weird background, a bunch of eyes in the background or something, people looking over your shoulder, obviously.

But yeah, because the eyes are easy to find, the facial markers and facial morphology cannot be far behind, because, again, it's kind of an easy problem, once you find the eyes, because if you've found the eyes, you can find the mouth.

Though, the problem is that the human brain -- you've heard this before. It's this "uncanny valley" thing in filming. There's this word called the "uncanny valley." It's kind of what Pixar has so brilliantly navigated. If you present an animated figure in a movie, you have this interesting problem, which is the more it looks like a real human, the better it has to be rendered, and the more emotionally correct it has to be.

If it looks like, say, a big, furry monster, like in "Monsters, Inc.," what happens is it can kind of "Hey, George," and it can move, and we just love it. It just looks so cool to us. It's like a talking dog. It's just so amusing. It looks correct to us. That's because we have nothing to compare it to. We've never seen a talking monster. Or a talking car.

So, Pixar has built an enormous amount of success around making these amazingly rendered images, but not going into the "uncanny valley." They've got graphs on this. If you look it up online, people talk about, the "uncanny valley..."

Industry Standard: "Polar Express" is an example of that. Some people said it looked kind of creepy.

Rosedale: Yes. I actually though "Polar Express" was kind of cool, myself. But, yes, "Polar Express" is a great example, where you saw those cherub kids' faces, and you'd see them speak and your brain is kind of stuttering at it, like, "Oh. Something's wrong."

That's the challenge with avatars is, if I start moving your lips, boy. I mean, the amount of neural circuitry that we have targeted on watching somebody's mouth is gigantic. Huge parts of the neocortex. So, we are able to see the masked stuff in the way mouths move. And that's why somebody that has a slight lisp, or even very small altercations to human morphology are arrestingly disturbing to us in the real world.

So now, we're trying to render that with an avatar. Very hard problem. The way you look inside your mouth, your brain's looking for all that, like at almost hundreds of frames per second.

Industry Standard: Now, the other side of this is, of course, the cost for consumers. I mean, even if Moore's Law makes it, and is able to do this, and you guys develop the applications that can handle this, too...

Rosedale: Yes. If it's a $3,000 terminal, it's going to slow the adoption. I think what's going to happen -- and we've open sourced the client, which I think is exactly the right move. Meaning that everybody's going to force us. We're all going to go in the same direction, because the water sort of seeks the level. And I think that people are going to make Second Life run faster and faster on cheaper and cheaper machines. And I think that's going to be more of the evolutionary direction that's helpful.

But, like you said, I do think it's possible that there will be some seminal breakthroughs that are related to hardware or interface that do potentially require more expense. But, I think the two will be hybridized. In other words, I think the basic, just sitting at your MacBook or your PC experience of being in the virtual world, interface has got to get better and experience has got to get a lot better.
I think, at the same time, in parallel to that, there will be, I don't know, a small number of people in the world -- and I think it's going to be very interesting, too -- that are empowered by more sophisticated interface technologies.

I would love to see a Second Life in which, you're cruising around. I love this idea myself. It's a very Snow Crash kind of idea. Like you're walking around in Second Life, and you have this better interface, so it's really quite easy to move around. I'm thinking about working on that, too, where you really can come into a group of people and really feel facile, the way you do in reality -- or at least maybe the way you hope you do in reality.

So you feel like that. But, then, you're in that public space, or that bar or whatever, and you look over and you see an avatar. And all of a sudden, you're watching them, and you kind of realize like, "Oh, my God. That person is using some sort of a full body interface." You know what I mean? You see like their hands moving, and you realize like, "Oh, wow! Whoa! That's one of them."

And if you think about it, I think that if one in 100, or even one in 1000 people had interfaces like that, it benefits everybody. You know what I mean? Because you'd be like, "Wow! I want to follow that person around now and see what they do, or talk to them about it." We used to talk about, when we did this interface work, we used to say that the coolest thing in Second Life will be like big brawls, like for fun. Where you have like Godzilla. You know what I mean? You're the avatar, you're using the full interface, and you're trained in it, or whatever. And other people are like shooting guns at you or something. [laughs] You're dodging the bullets, and people are like, "Whoa! How did you do that?"

I don't know exactly how it would work. But, I think it will be both things.

And I don't think that it is necessary that we evolve the hardware, any more than it was necessary that we evolved the hardware of the Internet. It was more the software and the idea that you presented a well formed page with a hyperlink and with the ability to have embedded images.

Which, really, when you got those three things, arbitrary hyperlinked URLs, pages themselves contained hyperlinks with paragraphs of text, and then the ability to put images in, which I think was probably, it can be argued, to be crucial for mainstream adoption. And that all came together in, what, '93 or '94, right?

Industry Standard: Yes.

Rosedale: And that was kind of what everybody was hoping for. It was a step for me. My experience of kind of discovering the net was around '94, '95. And my first experience was with a SLIP, a shell account, to like Berkeley or some service provider on the East Bay.

But, I only did that, I have to say, probably like all of us, probably like some people that use Second Life today, I wasn't hooked. You know what I mean? It was just text.

Industry Standard: All right. One last question, because you've been talking for 50 minutes. If I come back to you a year from now and you want to tell me about the problems you've solved or the accomplishments that Second Life has made, what would those areas be?

Rosedale: Well, I think that we ... there could be so many. I don't know if this is too broad as an answer, but I would like to see half the people that sign up for Second Life actually stick around.

Industry Standard: What is it right now?

Rosedale: Like 15 percent. So, I'd like to triple that number.

I also think, by the way, that that's the tipping point, where it kind of goes super global, and we see whether what I said about presence and language.

Oh, I didn't talk about crossing the language barrier, by the way. That's the other thing. Second Life crosses the language barrier. Not perfectly, but you have translators. And the other thing is that the navigational metaphor for Second Life is spatial, and that doesn't require language for a lot of things. You know what I mean?

If I rez you in a sim in Second Life, and it is the typical thing, right? It's the center of four or five radiating streets with a fountain and some signs. You already kind of know what to do.
If you go to some Beijing newspaper and you don't speak Chinese, the Website, you're not going anywhere. You're done. You're going to hit the back button. You're not going to click one time on that page. Why? Because the navigational metaphor of the Web, as powerful as it may be, demands knowledge of the language.

You have to be able to read the language. And so, what that means is that we have five Internets today which are virtually unconnected by hyperlinks: the Chinese Internet, the Japanese Internet, the Spanish Internet, the German Internet, and the English Internet. And who am I missing? ...

Industry Standard: Arabic.

Rosedale: Arabic. They're not connected at all, right? You know it. If you did one of those link diagrams, like the pictures of the Internet? They'd be totally discrete. There'd be like, "Whoop," one little link. And you'd be like, "Ooh, what's that?" Al Jazeera is connecting to Yahoo!
It's totally disconnected. But, Second Life, obviously, already is way heavily intermixed.

Industry Standard: But, how's the international nut going to be cracked? When people are talking with each other, they still...

Rosedale: Oh, no. Well, the other way it's going to be cracked is that -- so let's say you go [to] Tokyo. If I dropped you in Tokyo and I tell you to survive there, or even have fun, you'll be totally fine. Why? Because one in five people speak English. No problem. Same with thing with Chinese. So, it's not a problem.

In other words, if there are people wandering around in the virtual world, even if you can't read the signs, it's not going to be a problem, because you can ask somebody, just like in the real world.
But, on the Web, you cannot ask somebody, because there's no one wandering around when you're at that Website. So, I think that's another reason why Second Life, or virtual worlds in the most general form, will go across language boundaries. So, it's another reason they'll be huge.

But, I think that that will start to happen. We'll see. When we get to 50 percent of the people staying around, that's, I think, some sort of a sociological tipping point, because, if you go out with a bunch of friends today, will you tell them about Second Life?

Industry Standard: Some, maybe. But, most, probably not.

Rosedale: And the reason I think you won't is because I think -- I think we all do -- you have a natural psychological aversion to failure.

You know, intuitively, that, as exciting as Second Life may be, you know, intuitively, the fact that you just asked me to quote you a number on it, but you could have told me it was less than 50 percent. You know in your heart it's less than 50 percent.

Industry Standard: Right.

Rosedale: That's easy to ascertain. So, what that means is that you won't talk about it because it's a loser's bet. Less than half the people that are listening to you in your group of friends at dinner are going to come back a week later an say, "Man, that was so cool." The majority of them are going to come back and go, "I don't get it. Why do you think that's cool?"

Industry Standard: So, if it's closer to 50 percent, it's much easier to relate.

Rosedale: Then you can talk. And then, once you talk, then it undergoes hyper viral, just like anything else, like Facebook.

Industry Standard: Do you think we're going to reach that point in a year?

Rosedale: No. But, I think we'll think we'll reach it in two years. I just think that...

Industry Standard: Two years.

Rosedale: Yes.

Industry Standard: OK. So, you think the interface, or other things that are going on...

Rosedale: Yes. I think these are largely design problems. I mean, I could be wrong. It could turn out that we just don't crack that. But, I do feel like the Moore's Law curve of computing power, even on laptops, is giving us enough gain, giving enough advances in frame rate and stuff, we'll be able to get the launch time and the load time... I mean, that's another thing. A year from now, I'd like to see Second Life start up as fast as a browser. I'd like to have you in world that fast. I think when you double click on Firefox, or IE or whatever, you kind of go [pauses] and the page is up and stable? That should be the way Second Life works.

Industry Standard: Let's say even you double what you're doing now -- so it goes from 15 percent to 30 percent -- that has enormous implications on the scalability of the whole system.

Rosedale: Oh, yeah. We'll be making lots of money. I mean, yeah. We'll have more growth than we can handle at that point.

Industry Standard: Doesn't that set off alarm bells in your head?

Rosedale: But, I think we're doing everything we can to do the right things there. I mean, we're trying to architect the system in an open, standardized way, so that we can accelerate, for example, the rate at which servers are added to the system.

So we're constantly looking at how to open the grid up. Behind the firewall, we're working on Second Life systems that can run on different servers than our servers we own, to anticipate that kind of growth.

In fact, that's the primary problem. The primary problem is that the simulation space itself is CPU intensive. Not a problem, because it generates positive revenue. I mean, we're profitable, and the people using Second Life are profitable, so it's worth it to install those CPUs. But, nevertheless, this thing, much like the early Internet, much like the Web, it expands by the attachment of physical hardware to itself.

So, unlike like Facebook, there's a natural break on the rate of growth that's imposed by the fact that you literally have to build additional machines on the network. So, it's more like the growth of the Web, which was sort of steady but inexorable. It was just like, every month, you just got another few percent.

And I think that, right now, we're at like one percentage -- or not one percentage, but we're at whatever we're at. We're growing more like five to 10 percent a month. And I think that as we crack those usability issues, it'll come up some and it'll want to be even higher than that. And if you just scale those numbers out, they get really huge.

But I think the big problem will be putting servers online, and making the system scalable to that, and also making it have no central points of failure. And frankly, we've done a really good job on that.

I mean, I would love for us to have done better. And if you talk to Second Life users, they'll bang on it. They'll tell you that we're doing a terrible job on the staff. But, we're the only game in town. We're the only people doing this. Everybody's going to say that we're not doing it well enough. And there's going to be a subset of people who say we're doing great. I say that kind of naggingly, because it's very painful sometimes to be as emotionally connected to all this as we all are and hear people be so angry and say, "Why can't you do this better or faster or whatever?"

I don't know what to answer to that, other than to say I think we're a very good software development team. And this is a software development problem. It still is. It's not a marketing problem, and it's not a social problem. It's basically a design and engineering problem, at its heart.
And I think that the team of people that we have here, especially with Mark [Kingdon] and all the new people that are here, they're really good people: Tom Hale, Howard Look, and Brian Michon. These guys are very heavy hitters. We're going to do it as well as anybody else has done it. So yeah, I think we can get there.

Search needs to get a lot better. That's the other one I would call out. Like there's that new user experience, to the point where you're like, "OK. I'm comfortable with this." There's making the experience just more inherently delightful. I would love to see Second Life -- and I think this is totally doable. This is something I'm personally interested in working on.

When you move an object around in Second Life, if you ever try this, like if you pick up a physical object in world and you move it, it's like, "Yeah." It really is. If you have a high frame rate, it kind of follows you like a spring. It's so cool. You're like, "Yes. That is so neat. Whoo!" I can juggle things in Second Life. You get really good at it.

Industry Standard: You must be a pro. I can't do that.

Rosedale: But the rest of the interface, it's not fun. Walking around isn't fun. Flying is kind of fun, but the objects don't rez in in the right way. We can fix all that stuff.

So, I think, if we just make the actual experience of just being there more delightful, so that it's more like the iPhone -- you just log in and you're like, "Oh, man. That's just so cool. I just want to sit and look at flowers and goof around" -- that anyone who did it felt that way, I think we can achieve that. I think that's software and design, and I think we can do that.

And then I think we can make search really great. That's a concrete thing that I'd look out in a year. I mean, it's hard, and I hope we make lots of progress. But, if you look at how search works today in Second Life, well, it's just not very good yet.

Industry Standard: It's actually an improvement over two years ago.

Rosedale: Thank you. [laughter] It's true. But, come on. I mean, it could be fabulous, right? It could be really fabulous. We can look at where people are. We can look at people's pics. We can pull together all kinds of metadata and make the search work really well, where you just type in "live music" or whatever and you get something that really makes sense. And it's very Web like in its -- I don't mean in its presentation visually, necessarily, but in its scope, in the way we open it up to everything and we get all the data.

The Industry Standard - Interview with Second Life creator Philip Rosedale - Comments

Interview with Second Life creator Philip Rosedale