Thursday, September 4, 2008

Thursday PM Open Source Interoperable VWs

Moderator:
Panel:

(Missed blogging first bit.)

Focus on OpenSim. Different variations on the same building blocks mandates the need for standards beyond OpenSim. Modular software plus a protocol to allow distribution across a series of clients and models.

WHAT PEOPLE ARE DOING TO PROTOTYPE APPS IN OPENSIM
Still in alpha code mode. At a cusp. Fashion Research Institute will make an enterprise solution using OpenSim to produce ... ???

OpenSim is open source, but also public domain under the BSD license. Historical context on why tcpip and BSD worked well and got adopted by M$, etc.

What is meant by interop? The long term goal is to go from small set of homogeneous worlds, be able to take your identity to worlds that are less and less alike.

Successful standards emerge from working code. Learn from existing code, learn from mistakes, and get more diverse.

Thursday PM VWs Meet the Web

Moderator: Skip (?) Constable (Electric Sheep)
Panel: (focused on consumer side)
Mitch Olson (Smorgles)
Daniel James (Three Rings)
Keith McKurdy (Vivity)
Sean Ryan (Mees)

ACCESSIBLITY, USABILITY, AND THE USER EXPERIENCE
2.5D vs 3D and the decisions you have to make. 2.5D has a range of different communications apps. Facility social interplay and immersive experiences. Dimensionality doesn't necessarily have a significant impact, acc to Olson. Ryan seemed to disagree. Keith said it doesn't have to do with dimensionality, but rather usability. What is easiest for the consumer? Assuming that is solved, then users gravitate towards the richer experience. Dan thinks immersion takes place in your head. A text game can be "immersive." If you require a plugin or download of any type, you are dead if you are going after a mainstream base. This brings you back to Flash or Java. If you have a download, 5% of people who check you ought will actually sign on. If there is no download, it will be more like 50%. Dan said with their app, a popup would come up and ask for permission to run it and 95% would not continue, agreeing with that previous statement.

SOCIAL NETWORKS
People want to be in VWs with people they know. Vivity is baked inside Facebook. Dan disagreed with this. 10% of Facebook users are logged in at any one time. So asynchronous use is important. The strength and weakness of VWs is that it is real time. Most people online are looking to meet people (consider that most users are teens). So they don't want to limit themselves to just the people they know in the RW.

HOW WEBSITES WILL CHANGE BASED ON VWS
The rise of community...forums to chats to ...there must be freedom to allow the users to disagree with you.

Quote of the day regarding EA: "If they could sell their mothers for money, they would do it."

The market and consumers dictate. They want 3D games/VWs.

Question: we have heard a lot about Facebook, but where is the next big thing coming into VWs coming from?
Answer: Sony's Little Big Planet (physics based, site scrolling game). Make your own physics-based games and share them with your friends. "Release some tools and see what happens." Someone will release a killer app for content creation. Only 1% of users are content creators, about 20% are modifiers, and everyone else just sponges. Give them more tools for creation and, even though more people will not create, it might generate the killer app. Very interested in how Spore will evolve.

Thursday PM VW Roadmap

Moderator: ???
Panel:
Sibley Verbeck
Victoria Coleman
2 others (the intro was too quick)
Mic Bowman (the tech guy)

An intro of the obvious of why VWs are cool.

Proposing to drive an open source-like activity to develop and frame mainstream uses of VW tech, study them, publish their results, enumerate a timeline by which VW industry can succeed beyond a single generation. A platform-neutral roadmap.

A forum where people are invited to participate.VWRSIG is a project of the Contact Consortium (http://www.ccon.org).

What is the VW Roadmap?
-Executive summary of the future of deployment
-Case studies of example usages and potential best practices
-Vision of future of each usage area and a table of key performance benchmarks and technical characteristics

VWRM vision documents include a core anticipated value, foreseeable barriers to mass adoption, guidance as to how barriers can be overcome, predictions.

Investment landscape: $184M in 23 VW in 1Q 2008. $425M in 15 VWs in 4Q 2007. $1B in 35 VWs Oct. 2006-Oct. 2007. Club Penguin was the largest investment by Disney of $70M. SL was not on the list shown.

http://www.virtualworldsroadmap.org and ./visiondocs/LiveRWEventsVisionDocument.pdf .

Workshop to be held 10/14 in Northern CA (tentatively).

Showed video of interacting with VWs using an iPhone competitor device using the accelerometer to navigate.

Question: how is this difference from the Metaverse roadmap?
Answer: this IS that roadmap as the industry has moved along. That was a single session with view towards the 10 year horizon. It was a single exercise to produce a single document. Want to take that exercise to a much more rigorous level. What docs will be helpful for moving the industry forum? Make it an open effort. This is the mature version of what got started there.

Question: what is the right balance between interop and business models?
Answer: there needs to be a balance. There is the business and tech sides. This provides the business use cases to provide context and priorities. What does it mean to have av portability? This group will think through the use cases and make recommendations back to the tech organizations. This group will not solve those problems, but will create the requirements. Our presence protocols are designed with very few state changes, but presence in a VW is much more complex. Provide the business requirements for why we need this and the business case for why it should be done.

Not trying to tackle interop. Trying to define the feature sets that are needed and the business arguments for why to do it.

MASS ADOPTION has not yet occurred because there is the lack of a compelling need for most people to go there. Also, you need certain tech enablers to get people into them (ex: a SL client that doesn't always crash). People like browsers, but don't know what they are missing by using it over VWs. VWs are boring because there is no activity going on and no people there. Need to get the tech in the hands of many people.

Challenges: interop, mobility, scalability, AI.

Thursday AM Augmented Reality

Moderator: Roo Reynolds (BBC)
Panel:
David Orban (Chief at WineTag Inc.)
Eric Rice (Producer)
Marc Goodman (Alcatel-Lucent focused on innovation, runs University's Innovation Program)
Blair MacIntyre (Georgia Tech professor, working on games on hand-helds)

Started with wikipedia on Augmented Reality. Panel defined it too...understanding the world better through different means. Instances where you are tightly merging media with the real world. More collectively, Blair called it "mixed reality." Not important to make strong distinctions insofar as it lets us talk. An overlay for all of the invisible data were are constantly surrounded by.

"Augmented Reality Character on Smartphone" YouTube video.

"Magic Lens" -- YouTube.

Bringing real stuff into the virtual world and virtual stuff into the real world.

Showed several of the Georgia Tech YouTube videos. GTU's goal was mixed media machinima. This was funded by Alcatel-Lucent. A-L showed this at CTIA. Showed a video of this using cell phone. By mixing realities, this is one area where businesses are going to want to go for the purposes of telepresence conferencing.

The input devices are now changing through touch screen, accelerometers, and cameras. Gaming VWs tend not to look at each other, but they have the most to learn from each other. You don't necessarily need a sophisticated headset.

It is useful to consider separately the data collection aspect versus the data visualization aspect. Having the data first is the most important thing. The semantic nature of AR...without the ability to tag knowledge prohibits any AR visualization. Much more data is needed where you don't need to rely on people to waste their lives collecting. We need sensor networks to really enable autonomous collection --> AR display. This is the biggest challenge for making the AR world useful.

The reason black markers are needed to triangulation is so you don't have to model the entire environment. This is one of the biggest advantages of AR. But without these coordinate queues, you can't really get a realistic path, for example, with AR yet.

The tech that we tend to use to interface with this stuff has a certain amount of "social isolationism." It makes it harder to interface with the real world since you have covered yourself with all of this hardware. Which tech lends itself best to this right now? iPhones, cellphones. We are already isolating ourselves thru interaction with these things (example: the driver trying to hide looking at the cell phone)...we are just in denial about it.

The HMD will never be popular or mainstream. When it is a nice pair of glasses with everything embedded, no chords hanging off, and a good software app is when it will take off.

Book: "Everything that is good for you." It is not necessarily true that we are isolating because of these devices. Like social networks, they are bringing us together in new ways. The intensity of sharing is so high that it feels like you are there. As a society, we learn what makes sense and what doesn't and we don't need laws for it. It will eventually be normal to hold your phone up in front of your face to get additional information about a thing.

From a business standpoint, it is advantageous because it gives the users the information they want when they want it. They are pulling the information rather than having it pushed at them.

INTEROPERABILITY
...between AR and other bits of the web. Is there a roadmap? Are there standards or is it adhoc?
No, it is still messy. There are already things (company: Media Powers) like QR codes in the works. Maybe there will be in ARML that combines geotagging, semantic web, etc. How do we make the sensor nets available to phones? Commercial services like Frommers? Eric was very negative about commercial entities moving towards this. Academics need to be careful about rolling out standards too soon without a lot of input from users.

As we build communities or societies that have to be accurately represented across AR, we must take cultural value into account. What happens when an avatar finds a place where slavery is legal? What about automated XML contracts?

Book: "Halting State"...by Charlie Strauss. About AR and the police use of...this generation's "Snow Crash."

How do we tag real life data? Or is it just a spillover on what we are already doing with photos online? If you are typing things (ex: Twitterific) and it is geotagged, is that really useful? What is the best way to display that? Plastered on the walls? And how do we tell the VW where the walls are? If we are going to get to real apps, we need to not only geotag, but also have a meaningful representation of the real world with an understanding on where on the world these data get displayed. Is the best tagging geotagging or will it be situational (ex: "I am in LA" versus "I am at dinner with ... ").

Book: "Rainbow's End" by Verner Venge.

Expressed concern over the social issues of using the data taken by the average citizen to inform governments (example: the UK wants to use the photos of its citizens taken with cellphones for monitoring for terrorism).

This session made the whole conference for me.

Wednesday, September 3, 2008

Wednesday PM Mobile VWs

Moderator: Wagner James Au (Author, The Making of Second Life)
Panel:
Martin Dunsby (CEO,Vollee...launched Second Life Mobile last week)
Victoria Coleman (VP, Samsung Electronics)
Tom Churchil (CEO, Earthscape...virtual globe app for the iPhone)

See www.vollee.com for free download. The frame rate is pretty good relative to SL on the computer.

Their privacy policy says that they don't watch what people are doing.

"Haptic" phones...touch screen-based technology like the iPhone. "The interaction with this phone is very different." Showed a means of shaking the phone to simulate rolling dice. You can't take the PC experience and shoe-horn it into a phone. Why do people in SL not do IM? We have to learn from both the VW side as well as the media group (phone) side. VW currency that goes across the VW to the mobile phone.

Earthscape: like Google Earth for the iPhone. Does geo-tagging for photos to make a mirror world. Lets you see other photos people have taken there as well. End goal is an "immersive Twitter" where people could see real time photos of events like parades. The business model for virtual globes is different than for VWs. This mirror world attempts to model the RW. The "locate me" button provides your location to any user. You are getting the ability to real-time wiki the world.

There is a move to ease the restrictions that customers have on the end devices.

Your world will be more extensive if more people spend time in it.

One audience member asked the panelists to explain their optimism. Handset OS's are all broken in different ways. Accelerometers are not being fully taken advantage of. We are still years away from seeing any of this as a true possibility.
RESPONSE: there are a few things that are ubiquitous: web browsing, text SMS, video. So the VW apps are being dumbed down to pass by the OS to act as a dumb terminal. There are ways around it, although it is the holy grail. It is difficult to get apps deployed on them in any meaningful way.

Technology limits by the companies to prevent unlocking phones. The operator is subsidizing the price of iPhones, so the investor wants a return on this. This is a chronic problem with cell phone companies. So the operator is a channel for the apps. These are strictly for business reasons and not technical ones.

Wednesday PM How to Convince Your Boss You're Not Off Your Rocker

Erica Driver (Principle at ThinkBalm)

Immersive Internet: a new dimension of engagement and context.

Why spend money:
-reduce business costs
-prepare for workforce of digital natives
-workforce proficiency at lower cost

Michelin IT pros on EA concepts:
98% felt immersive environment appropriate way to learn about EA concepts

Pre and Post Test Scores at Stanford Med: World Journal of Surgery, Special Issue, 32(2008),161-170. Shows that test scores with hands-on training vs. VW training are the same.

Virtual events cost about 1/3 that of traditional events.

Accenture Careers island paid for itself in 5-6 in-world recruiting events in SL.

Data visualization: very simplistic examples

Still waiting to get something useful out of this talk.

Early adopters: pick one problem and one target audience. Learning and training; meetings and conferences; recruiting and inerviewing; remote system and facility management; collaborative design and prototyping; data visualization.

Read metrics to show business value:
-Increase # of event attendees while reducing event costs
-Reduce travel costs
-increase workforce proficiency gain without increasing cost
-Hire the right employees more quickly
-Reduce the number of hours employees spend travelling

Leverage experience gained in enterprise collaboration initiatives

Question from the audience: Technology diffusion curve...where are VWs?

Not too much of use from this talk.

Wednesday AM Technical Visionaries

John Sorts (Director of Business, Electric Sheep), moderator
Panel:
Ven Gurzel (CEO of NoPomente...makes AI)
Christian Renauld
...
Can't hear in the back to hear who everyone else was

Group agreed it might be time to talk about VW 2.0. Last few years were 1.0. What characterizes the changes that are coming?

Are there better ways to interact as humans other than avatars and islands? Motion tracking? Does this constitute 2.0? And O'Reilly-esque conversation on what it might be?

There is a shift that will take place (~5 years) when VWs and tech behind them resemble more closely Web 2.0....more plug and play and easily combined. Do away with MMOG minus G and integrate with the rest of the web. Don't use general purpose social networking tools. More robust simulation modeling, education.

EXISTENT: that which has its own individual being. Generalize the notion of an avatar. Have its own individual controlability through scripts, AI. The world is no longer at the center of the VW, but the avatar is. It has its own state with its own knowledge. They are portable across different VWs. Interop goes beyond avs to machines. Things centered around identity. But this is a more general problem on how you tie identity across multiple domains. This will merge to a unified identity.

Ubiquity: not making email programs the place you go to do the work, but a generalized solution. Context is not important for the software agent.

In the consumer social space...for Facebook, do you want to see it in 2D or wander around it in 3D? Offer it as an option to the user. New things like SLim are not a step forward because it is just another thing that you have to deal with.

PERSISTENCE and PRESENCE and IDENTITY
VW interop like the broad internet. Transparent integration...no longer about polygon counts and avatars but software integration.

Redeeming values of VW 1.0 carried forward: it is about people. Being able to connect in an enterprise, education, or socially. It is about what they can do and create and express themselves. VWs give us a deeper degree of connectivity. Need to take this to the next step. What is next? Bring in the emotion (through video etc.?) and then bring it back out again. What is driving that? As interactions get more rich...this will not push it over the edge into the next phase of development. There are still apps that are missing that have to do with bringing people together to actually have the interactions. This is solved on the web through search engines to do discovery. This has not been done in VWs. In SL, you can see cool content if you know where it is. But it is hard to find if you are just dumped in.

SYNCHRONOUS, REAL TIME...WHAT HAPPENS WHEN IT ISN'T SYNCHRONOUS ACROSS THE RL?
Asynch tools will resemble those of the web. Ways to discover new info and form new connections with offline users. #1 use of email statistically to arrange face to face meetings.

doppler (can't find it online)...an engine to share your travel plans with your friends. Will the same serendipity happen in VWs?

Are we just teaching a virtual dog new tricks in AI? We need to go further than 3D worlds in terms of interactions with the real environment.

USING 3D WORLDS TO INTERACT WITH 2D CONTENT
Can you interact with software in 3D rather than a 2D document? For reasons of costs and time constraints, software is done in a restrictive way.

HOW WILL OPEN SOURCE PLAY A PART?
Answer was just a statement that it is important and why we all know it is. OS would have been looked on as a threat to the traditional software company, but are presently looking at it as an opportunity via OpenSim. OS is driven by people really waning something so bad that they are not paid for it. This will drive innovation. Several companies are waiting to adopt VW tech until there is a stable platform that will not be made obsolete during the development process. Open Standards exist but have not been given mass adoption because they did not consult the big users to see if the standards fit their needs.

IDENTITY AND AVATAR PORTABILITY: DOES IT MATTER?
Technically, it is not a big problem to do it. There are layers on top about data meaning and translation that need to be solved. The big thing is the social side and the acceptability of moving stuff. But when you pay money for your identity to belong to a particular VW (ex: WoW), they don't want you to do this. But who does your virtual identity really belong to? Identity port is a macro issue, but av port is a subset. What about asset port? If you paid for it in one world, you should not necessarily be able to take it into another world if that is the agreement. We should not dictate to companies with business models how this should go without understanding the difference of these things. But if something belongs in an environment, we should not think about asset stripping. There are way to stream things, existing in their original place and use it in some way in a different VW. This allows each environment to be appropriately hit (and billing). There is no reason that business models can't adapt to support interop.

THE FUTURE OF THE HARDWARE
Intel moving away from the GPU to a standard x86 core with like 32 cores. But there are some applications that more graphics power is not necessarily better (ex: a cartoon character smiling does not need to be high rez). Better physics engines, better AI. Use the physics engine to mediate interactions between avs and objects. Main problems are bandwidth and processing power.

GAME-CHANGING TECHNOLOGIES
-3D printers...see shapeways.com for a paid service of this.
-AI that goes out and gather information and report back.
-Virtual goods.