Earth 2 Avatars was the first annual conference of the Contact Consortium, an organization whose aim is to transform the Internet into a medium for human contact. Bruce Damer, president of the Consortium, described the conference as part of the launch of Avatar Cyberspace - a new place to work, learn, communicate and be entertained.

The opening speakers were Mark Pesce and Tony Parisi, reunited as a double act to take turns presenting their ideas on process and people. Tony looked at the technology while Mark looked at the literary inspirations behind Cyberspace. They both concurred on the fact that, "we will have Cyberspace, but it will not be built in a day".

There were numerous discussion panels: more than I could possibly attend, although they all sounded interesting. "The Galactic Edge" - a special program of Biota looked at the future of synthetic life forms. There were some bizarre ideas floating around in these virtual spaces, such as algorithmic plants that grow using a combination of L-Systems and neural nets. At the moment, the ‘digital jungle’ rapidly acquires too many polygons so the designers of the Nerve Garden are working on virtual animals to eat the plants.

Charles Ostman took a look at self-constructing entities and the concept of ‘virtual terraforming’ to spawn new systems in cyberspace. One of his projects is the ‘xenomorphic entity’, a synthetic sentient entity. These entities would come in a range, from personal agents to entities who manage your life. He believes that knowledge engineering will become the most important human concern and that the Internet will become a self-modifying system, an organism itself.

Marty Stoneman introduced the ideas behind anthrobotics. He constructs virtual reality inside his entities ‘heads’ and uses a story to link objects and ideas to help his entities achieve high level goals. He pays attention to things like metabolism, emotion and external circumstances.

"Kids in Digital Space" was a panel examining what can be done to involve children in virtual worlds and how to design interfaces to help them enjoy their experiences there. David Vroney from ImaginEngine, the makers of a number of CD-ROM games for children, spoke about some of the ideas they had developed, in particular with their MonsterMaker product. This product targets children as young as three years old so it has to work without words, dialog boxes, menus and the rules that stop adults having fun. David pointed out that messages that tell you ‘you can’t do that’ are not only infuriating but also a sign of bad design. So his kids products don’t prevent users from, for example, joining an avatars hands to a leg rather than an arm.

He works with a totally embedded interface; characters within a scene give feedback on user’s actions and may offer advice. "Leverage off the user’s expectations from an environment," David advises. "Don’t make a user leave the game or the world to change clothes." "Don’t make a user leave the game or the world to change clothes."

Choosing between creating or editing a scene is crucial for kids’ interfaces" he continues. "Editing is easier and faster". The four rules he cites can be applied to any interface design:

1. remember that Legos are easier to use than paint brushes
2. match granularity of controls to the task to be accomplished (and don’t merge tasks)
3. eliminate illegal operations (Don’t you wish MS and Netscape followed this one?)
4. keep operations single step.

1. make the expert match the task - an architect to build a house, a hardware store to supply parts,
2. try to make your help mechanism like the Holodeck interface on Star Trek (give me a table, no shorter, now darker wood, more fancy etc..),
3. remember the relationship between shopping and customizing - when buying a car you first choose a model, then its color and other parameters can be adjusted,
4. balance education in how to use the environment with ease of use for novices.

ImaginEngine must be doing something right: they’ve just signed a deal to integrate 2D avatars with The Palace, for adults this time round. Later they’ll start work on a semi-private Palace just for kids.

The "Worlds for Gaming and Fantasy Role Playing" panel gave us a chance to move through games with the power of the system administrator and gain an overview of the design. Rusel De Mario introduced Fabrice Florin of Zenda who showed Charebus (a Palace parlor game), Steve Nichols, Designer of Sierra Online’s The Realm (a third-person, 2D MUD), and Mike Sellers, now of 3DO but formerly of Archetype Interactive, who gave us a tour of the game he developed: Meridian 59 (a first-person, 3D MUD).

This game is phenomenally popular. Even the story of how 3DO discovered and helped complete it is a high speed trip. Mike had it up on the Web as a free alpha last December. By April Archetype had been acquired by 3DO and had its retail launch on Oct 1st after beta trials attracted over 27,000 participants in four months.

The game requires purchase of the basic CD environment for around $40 then monthly payments of $9.95 for unlimited play. The combination of action, adventure, puzzles and chat has led to a remarkable level of player involvement extending to weddings and wars. Player profiles indicate two thirds of the players are male but the age range runs from kids to people in their sixties, with an odd concentration of women lawyers in their forties. He worries that he may be creating an addiction problem for users: "digital crack"

To explain ‘Why 3D?’ Mike says mainly it provides immediate immersion and a sense of place for the players. 3D encourages a suspension of disbelief and eliminates screen edge problems as the player naturally has to turn to see someone beside him or out of shot. The avatars used in the game are bitmapped graphics "because it works" for producing reasonably detailed figures which can be easily customized from a parts palette then animated with gestures and expressions. The decision not to use VRML was partly a matter of what users’ machines could be expected to handle. At the moment, 3DO’s assumes that users have a 486 or better and only 8MB of RAM, although there’s a push to move to 16MB of RAM and Pentium machines. VRML is seen as being too slow and its representations of avatars as too limited, "all form and no function," as he put it. Mike, who comes from a background in Cognitive Psychology, has an analysis of the levels of social computing:

1. Sensory /Perception level e.g. DOOM
2. Short Term Cognitive e.g. MYST
3. Long Term Cognitive e.g. CHESS
4. Interpersonal e.g. MUDs

LivePicture demonstrated the relevance of their FlashPix file format to the Web, 2D printing and 3D VRML worlds. Stunning panoramic scenes had fully zoomable areas. These showed the difference between the pixelated bitmaps and the high-resolution FlashPix images. LivePicture is working with RealSpace, LiveWorld and Seismic Entertainment to produce a 3D virtual, multi-user, photorealistic, guided exploration and discussion of new places.

Oz Interactive revealed their new intelligent agents,a friendly help angel and a disputational girl form. To experience a conversation with them, download Oz Virtual and visit any of the Oz multi-user spaces.

Both 3RD Dimension and Sven turned up will full photographer's studios, ready to turn volunteers in avatars. While 3RD Dimension gives all its personal avatars hats or helmet's, Sven was turning our avatar heads complete with hair and generally improving on the model’s bonestructure at the same time. Most people approved their results and went on to get a whole body avatar from the people at Sven. By using ‘standard’ body or head models, Sven figures they can save on bandwidth, as they only have to transfer a texture map and the control points to map the texture onto the geometry. They can then animate the model and use the control points to keep the texture accurately mapped onto the underlying geometry.

But 3RD Dimension they had a new product to demonstrate: Puppet - a free 3D Chat Server. Avatar heads could text chat from one machine to another via the Internet. The interface allows users to either select a standard avatar head or enter a URL for a personal VRML avatar. The Puppet server uses Java to download to a user so no client software is necessary. Puppet runs inside Netscape Navigator 3.0 and is native on both Windows 95 and NT.

Closer Look Creative set up their Avatar Bazaar. Mike Messing showed the bizarre VRML 1.0 creatures - part TV, part spider - that he’s working on. He will be setting up an avatar server to act as a showroom for his avatars and their biographies. These avatars can be used in Black Sun worlds by copying them to your Black Sun preferences file. The next step for Closer Look Creative will be to let users combine parts to build their own avatars then add behaviors using drag and drop technology from VREAM.

The conference was a ‘coming out party’ for KATrix and their NeurRule Technology Biped API. They have been working on motion control for some time and have both expert systems and neural net learning systems which interact to produce motion triggered by a high level task language. They also have a limb coordination technology which they are incorporating into a skeletal control plug-in for browsers. VRML is just one language that can be coordinated to the behavioral engine. Their demos showed Metalhead coping with a range of reaching and stepping motions without losing its balance and always maintaining consistency of its body movements. The behavioral engine can also deal with multiple-jointed aliens.

Intel’s Distributed MOO version 2.0 was being demonstrated by a team that took turns exploring all the other demos. Intel has developed a framework which enables arbitrary object interactions in a virtual space. This means both world builders and users can originate interactive behaviors. The framework also enables users to add to or change the environment of the world itself, even to the extent of letting them create animated or interactive behaviors. Intel feels that because their approach is fundamentally extensible, it enables a much wider range of changes to take place without having to re-architect the basic communication infrastructure. The ID MOO 2.0 is a library of Java classes for building multi-user 2D or 3D spaces. Included with the code is a demonstration 2D underwater virtual environment built on top of the library.

Robert Saint John, IDS' Administrator of VRML Development, was at the conference to demonstrate their new (still alpha) VRML 2.0 browser. Unfortunately the computer with the software on it wasn’t in SF until the Monday after the conference. Although IDS was a sponsor of the conference their focus extends beyond what Robert calls "simple 3D chat worlds" to "online collaboration and building technologies".

Michael Hilgenberg, their IDS' VP of Sales and Marketing, spoke about "Worlds in Our Future", a peek at what virtual worlds will look like in 2 or in 5 years and how VRML 2.0 will impact the medium. IDS are developing a multi-user server technology but wonn't be available until their other VRML 2.0 tools are ready to work with it.

3Dlabs had SENSE8 at their table demonstrating the alpha version of their World Server, software which simplifies the process of sharing information between worlds and minimizing network traffic. As Tom Payne, product manager from SENSE8 says: " Their (3Dlabs) cards makes our stuff work really well." World Server will be a future stand-alone product available as an add-on with WorldToolKit and WorldUp. A specific use for the multi-user software was demonstrated by NTT’s Cybercampus, with 3D graphics created using WorldToolKit from SENSE8. Cybercampus comprises a series of interactive tours through 3D simulated environments but is not available on the Internet yet; it’s still on private ISDN lines linking advanced PCs.

Several companies were present to represent the interests of two dimensional interaction. Among them, and probably the most well known to regular Internet users, was The Palace. Operating on PC, Mac and UNIX platforms, it claims to be the leading virtual chat software on the Internet. Thousands of Palace servers have been set up by individuals and businesses around the world. Fabrice Florin from Zenda demonstrated his company’s adaptation of The Palace format: Charebus - a party game combining elements of charades and rebus. He explained that games enhanced social interaction in cyberspace because they gave people something to do rather than just chat. Zenda has developed a games engine to franchise on to other Palace users who want to run their own games environments. The games/party take place in the new Minds Palace, developed by Zenda Studios as a companion to Howard Rheingold's new Electric Minds web site. To join in you will need to download the Minds Palace software.

It features the day-glo visuals of comic artist Jim Woodring, the creative direction of Mark "Spoonman" Petrakis and all sorts of new inter-activities. For example, you can make your own character from a large collection of outrageous body parts and, perhaps, get your moment of fame in the monument to cool avatars.

AOLers may be familiar with Virtual Places software developed by Ubique Ltd. Virtual Places is a variety of chat environments for users with simple 2D GIF avatars. It uses an open client server architecture for live interaction on any Web page. Virtual Places users can use all sorts of Web page building tools to create and enhance their virtual communities. These tools include HTML, Shockwave, RealAudio, Java applets and whatever else can be used on a Web page. Release 1.0 lets developers work with either components running inside a Virtual Places window or a completely independent 3D application launched from Virtual Places. Client software is available for Win 3.1 and 95 and shortly for the Mac. Server side software runs on Sun or HP machines. If you’ve had enough of text chat, you can use Internet Phone in conjunction with Virtual Places.

CompuServe users have their own 2D chat spaces provided by Fujitsu’s WorldsAway. These 2D avatars are tremendously popular with around 40,000 registered users of the Dreamscape space. MSN has V-Chat a 3D space where 2D avatars can move around. The demonstrators at E2A were showing avatars with multiple expressive gestures using text chat in a variety of 2D and 3D spaces.

ART

The importance of artists was not forgotten at this conference. A special display room was devoted to the collaboration of artists and programmers who have produced avatars and worlds for them to inhabit. The most complex world was Epona, the result of over four years of collaborative work by scientists and designers, joined together as Worldbuilders, to create a complete solar system containing a planet with a fully elaborated eco-system. Epona’s inhabitants were represented by a real 3D creature.

Karen Marcelo and Frank Revi showed what will become a familiar concept : a ‘beta version of art’ produced by the collaboration of an artist and an engineer. As Karen says: "Now that artists need to deal with technology they can update their art or put out a new version." They can also behave more like programmers and stay up until 4a.m to finish coding a behavior. At the back end of the piece is an algorithm, iterations of which change the shape of a VRML creation. Marcelo and Revi have created an interface with the user, by using the Java API for VRML 2.0, and allow the user to control the shape of the object. Eventually they will set it up so the finished result can be published as a .WRL file and displayed by the user/creator on a personal Web page.

The Web Design Group was showing a project in preparation for a Japanese client. Lisa3 is one part of a display of futuristic fashions originally designed to "live" within a Japanese fashion company's online storefront. The designs are shown on a cartoon character, a mixture of Japanese Anime and TankGirl, who also exists as a VRML avatar with a fictitious biography. As C. Scott Young, the Avatar Design Team’s VRML designer puts it: "Cyberspace needs a myth, a story, characters, not just polygons." So the Web Design Group have developed a history for each avatar. They use a graphic novel type format in which each avatar has a resident alien card containing the restrictions on its movements, the worlds it can enter and its dimensions in virtual space. The ‘physical’ avatars exist and can be taken into virtual worlds. The Web Design Group say they are currently forming strategic alliances with major VRML companies to provide Lisa 3 and other avatars in development with artificial intelligence engines.

The Avatar Standards Discussions
The Living Worlds session was the official introduction to the ideas behind the Living Worlds proposal for how to deal with multi-user VRML 2.0 worlds. Representatives from all the supporting and collaborating companies were present to say a few words of encouragement and make it clear that this was a cooperative effort to define a conceptual framework and to specify a set of interfaces to support the creation and evolution of multi-user applications in VRML 2.0. Also introduced at the conference was Universal Worlds, which has since changed its name to Open Community and is yet another layer to enhance the interactive possibilities of multi-user worlds.

Living Worlds was co-authored by Mitra at Paragraph and Bob Rockwell at Black Sun. Universal Worlds was co-authored by Dave Anderson at Mitsubishi Electronics, Dan Greening, Maclen Marvit and Moses Ma.

Developing a framework conducive to free movement and retention of identity in a variety of proprietary worlds is a delicate and controversial undertaking. As the Living Worlds specification says: What is needed is a mix of standardization to ensure interoperability and openness to leave space for innovation. As with the development of VRML itself the development of standards is being handled as a community project. This conference provided an opportunity to review progress in many different areas.

THE GREAT DEBATE
Five industry experts consented to be coordinated by Bruce Damer into a source of sound bytes on the great issues of the day. The panelists were: Reed Hoffman - WorldsAway Product Manger for Fujitsu, Konstantin Guericke - VP of Sales and Marketing for BlackSun Interactive, Bonnie Nardi - Agent Researcher at Apple, Alexander Beshir - the author of Rim: a novel of Virtual Reality, and Maclen Marvit - Product Manager for Worlds Inc.

Why are we doing this?
Konstantin: We’re trying to make the computer more transparent (by using 3D) so we can interact with people.
Bonnie: We have nothing in common with our real neighbors - we have specific interests because we are so cultivated.
Beshir: We bring the best and worst of what we have now to build a higher consensus view. Cyberspace is a transition to another view of the world.
Maclen: We’re looking for people that we’re interested in interacting with. Time is hard to find. I can find Go players on the IGS Go server. It’s a community.
Bruce: We’re trying to get some contact back.

How is anyone going to make money in Cyberspace?
Maclen: Big movie/media companies will be dumping big money into 3D worlds. A real attraction is doing interactive things - it will pull people from movies to games.
Beshir: Creatives will be OK
Bonnie: End user tools.
Konstantin: Snowcrash gave a vision we all share. This draws money in, generates new jobs.
Reed: The focus is on consumers themselves adding value.
Bruce: Money will be made on intranets, collaborative workspaces are needed to save commuter journeys.
"Big movie/media companies will be dumping big money into 3D worlds"

Are there ever going to be virtual worlds worth going to?
Reed: WorldsAway is already being paid for - so yes. So many virtual mall projects are being set up some have got to end up selling stuff.
Konstantin: People go to Black Sun worlds and stay a long time. Accessibility is important so bundling VRML into the browser helps.
Dave Marvit (from the floor):. Luddites rose against an expression of technology, not the technology itself. We have no economic model yet for VR. We should speculate on the implications of the economic model associated with the technology. And we should look at the social context the technology generates.
Bruce: The virtual world can enhance the real world. If we want to go in that direction.

John Sculley gave the closing keynote speech. It was clear from his quick look at the history of personal computing that he has known all the main movers and shakers of the industry when they were just learning how to wobble things a little.

He sees a future where one of two revenue models apply:
1. Expensively produced, free content supplied by big companies. Sculley forecasts the huge multiple-media asset companies, such as Disney and Time-Warner moving onto the Internet and building interactive personalized Web-casting channels. This leads to the Net taking on the high quality associated with TV, CD-ROMs, and the movies. He sees VRML as just one of the tools to achieve this. "The existence of high quality free content will make it hard to have any content that is not free for users"

The existence of high quality free content will make it hard to have any content that is not free for users. Content is free because of advertising and the revenue from multiple outlet channels for any product. The big guys can have a free space on the web and it will advance the sales of the movie/video/TV-show/lunch-box/T-shirt. So, it is a revenue generator for them in a way it cannot be for the little guys. True advertising will appear not the junk mail banners we see today. We’ll start to see quality interactive commercials."

He sees telephone companies concentrating on being ISPs, investing in satellites, backbone upgrades, cable modems and then start charging for connect time to the improved infrastructure. "The Regional Bell Operating Companies(RBOCs) have learned their lesson and are likely to stay away from the content side and stick with plumbing. As for connect time, there is always the bottled water analogy. Buy the bottled stuff and get a better version (high bandwidth w/ low latency) of something that was available anyhow (connectivity that is slow, unreliable etc.…)."

2. The second model is that of free content provided by the users themselves in chat environments and games. This is similar to the idea behind CNN - use camcorder technology and unknown presenters in order to bring in news cheaply.

"This does not take into account the possibility of an Internet equivalent of Public Television paid for by charitable donations. But probably people who want alternative programming will just make it. Cable, with its diversity of alternatives, is already killing PBS. The alternatives on the web are so vast, it’s unlikely we’ll see a PBS style web presence."

At the end of it all there was the CCon-VeRGe-nce Post Conference Bash. This excellent party gave attendees the chance to finish off all those conversations that got cut short by pressure of events at the conference itself. Thanks to Curve, VeRGe, and CyberOrganic Corporation.