Wednesday, November 17, 2010
Web 2.0 Summit 2010: Eric Schmidt, "A Conversation with Eric Schmidt"
Eric Schmidt (Google),
John Battelle (Federated Media Publishing),
Tim O'Reilly (O'Reilly Media, Inc.),
"A Conversation with Eric Schmidt"
Web 2.0 Summit 2010: Mark Zuckerberg, "A Conversation with Mark Zuckerberg"
Mark Zuckerberg (Facebook),
Tim O'Reilly (O'Reilly Media, Inc.),
John Battelle (Federated Media Publishing),
"A Conversation with Mark Zuckerberg"
Katherine Savitt Interviewed at Web 2.0 Summit 2010
Katherine Savitt is the Chief Executive Officer and a director of Lockerz, LLC, a website offering social network, entertainment and commerce for 13-30 year olds. Formed in 2009 with 50 members, the invitation-only site has rapidly grown to 14 million members in 195 countries. Before founding Lockerz, Savitt was Executive Vice President and Chief Marketing Officer at American Eagle Outfitters, Inc. from March 2006 to January 2009, where she led both the global marketing efforts of the company's four brands as well as the digital and ecommerce businesses for the Corporation. Prior to joining American Eagle Outfitters, Kathy served as Vice President of Strategic Communications, Content and Entertainment Initiatives of Amazon.com from 2002 to February 2006, and also served on Jeff Bezos' S-team (senior management committee). From 1993 to 2002, Kathy served as President and co-Founder of MWW/Savitt, an integrated marketing communications and public relations firm. Kathy has extensive senior leadership experience with retail, e-commerce, content and social media businesses and, in addition to being a Lockerz director, currently is a member of the Board of Directors of Vitamin Shoppe, Inc. and Build-A-Bear Workshop.
Adam Lashinsky Interviewed at Web 2.0 Summit 2010
Adam Lashinsky covers Silicon Valley and Wall Street for FORTUNE. He has been on the magazine's staff since 2001, and for two years before that was a contributing columnist. In addition, he is a weekly panelist on the Fox News Channel's "Cavuto on Business" program on Saturday mornings, and he appears frequently on other Fox News and Fox Business Network programs. He also co-chairs FORTUNE's annual technology conference, Fortune Brainstorm Tech, and is a seasoned speaker and panel moderator.
Lashinsky's articles focus on finance and technology. Recent cover-story subjects have included Apple, Hewlett-Packard and Google. He also has written in-depth articles on Wells Fargo, Intel, Oracle, eBay, Twitter, the venture-capital industry, and the post-Katrina economic recovery of New Orleans.
Prior to joining FORTUNE, Lashinsky was a columnist for The San Jose Mercury News and TheStreet.com. Before moving to California, he was a reporter and editor for Crain's Chicago Business. As a Henry Luce Scholar, he also worked for a year in Tokyo as a reporter for the Nikkei Weekly, the English-language version of Japan's main economic daily. He began his career in the Washington, D.C., bureau of Crain Communications.
A native of Chicago, Lashinsky earned a degree in history and political science from the University of Illinois at Urbana-Champaign in 1989. He lives in San Francisco with his wife and daughter.
Cyriac Roeding Interviewed at Web 2.0 Summit 2010
Cyriac Roeding is the Co-Founder and CEO of shopkick Inc., focused on bridging the mobile and the physical retail worlds. shopkick's first mobile application CauseWorld turned into the fastest-growing and one of the largest mobile retail applications on the iPhone App Store and on Google Android within weeks after launch. shopkick was funded in 2009 by Kleiner Perkins Caufield & Byers (KPCB), as part of its "iFund," in cooperation with Apple, and by Reid Hoffman, Founder of LinkedIn and investor in Facebook. Previously, Roeding spent a year at Kleiner Perkins Caufield & Byers as an Entrepreneur-in-Residence to identify next-generation cross-platform mobile and online venture concepts. Prior, Cyriac was Founder and EVP of CBS Mobile. He launched CBS Corporation's (NYSE: CBS) mobile businesses across CBS Entertainment, CBS Sports, CBS News and the CW Network. In partnerships with technology start-ups, he created U.S. industry firsts such as location-based mobile advertising, mobile video breaking news and sports alerts, and avatar-based mobile games tied to major TV shows. He grew CBS Sports Mobile into one of the Top Ten highest traffic ad-supported mobile websites in the U.S.; created the first virtual reality video gaming experience across TV, online and mobile with "CSI: NY"; and launched the first Alternate Reality Game with full primetime TV drama episode (Numb3rs) and outdoor billboard integration.
Prior to CBS, Cyriac was the Co-Founder of 12snap, a European mobile marketing and entertainment company with clients such as Coca-Cola, McDonald's, L'Oreal, adidas and MTV Europe. Before that, he developed growth strategies at McKinsey & Company for global media, software and high-tech players in Europe and in Silicon Valley.
Cyriac has received several industry honors, including the first Lion Awards for mobile concepts at the Cannes Lion Festival in France (in 2003 and 2004). In 2007 and 2008, Cyriac and his team received the only Emmy Award nominations for Mobile. For 2007, he was elected as Global and North American Chairman of the MMA, with 600 member companies worldwide. Cyriac received a master's degree summa cum laude in engineering and business administration from Germany's Technical University of Karlsruhe, and studied Japanese management at Sophia University in Tokyo.
A Conversation with Robin Li, Chairman & CEO, Baidu
A Conversation with Robin Li, Chairman & CEO, Baidu
Average rating
(4.40, 5 ratings)
Add your rating
Moderated by:
John Battelle (Federated Media Publishing)
Panelists:
Robin Li (Baidu, Inc.)
5:00pm Monday, 11/15/2010
Location: Grand Ballroom
John Battelle
Federated Media Publishing
John Battelle is an entrepreneur, journalist, professor, and author. Currently founder and chairman of Federated Media Publishing, he is also a founder and executive producer of conferences in the media, technology, communications, and entertainment industries and “band manager” with BoingBoing.net. Previously, Battelle was founder, chairman, and CEO of Standard Media International (SMI), publisher of The Industry Standard and TheStandard.com. Prior to founding The Standard, Battelle was a co-founding editor of Wired magazine and Wired Ventures. He is the author of The Search: How Google and Its Rivals Rewrote the Rules of Business and Transformed Our Culture” (Portfolio, 2005).
Web site
Robin Li
Baidu, Inc.
Robin Li is the co-founder, chairman and chief executive officer of Baidu, Inc., and oversees the company’s overall strategy and business operations.
In the nine years since founding Baidu in January 2000, Robin has turned the company into the largest Chinese search engine, with over 70% market share, and the third largest independent search engine in the world. In 2005, Baidu completed its successful IPO on NASDAQ, and in 2007 became the first Chinese company to be included in the NASDAQ-100 Index.
Prior to Baidu, Robin was already regarded as one of the world’s top search engine experts. His hyperlink analysis, patented in 1996, is among the inventions that shaped today’s search engine technology. Robin worked as a staff engineer for Infoseek, a pioneer internet search engine company, from July 1997 to December 1999, and as a senior consultant for IDD Information Services from May 1994 to June 1997.
Robin received a Bachelor of Science Degree in Information Management from Peking University in 1991, and a Master of Science Degree in Computer Science from the State University of New York at Buffalo in 1994.
Gravity introduces system that studies your social interests
We built Gravity to help the right information find you.
Today, we live in a world in which we’re constantly bombarded by information: 90 million tweets per day, 35 hours of video uploaded per minute, 1.6 million blog posts per day and an average of 130 friends demanding our attention. With so much information being created on a daily basis, it’s hard to find what you’re looking for and impossible to know what you’ve missed.
Our answer is the Interest Graph: an online representation of your real world interests and a new lens through which to view the internet. Your Interest Graph is your own personal electromagnet. It pulls the best stuff to you based on your interests and leaves all the noise at a safe distance where it can’t distract you. We build your Interest Graph by analyzing social data (like tweets, retweets, status updates, likes and shares) to create a holistic picture of who you are and what you’re interested in. When you click personalize on a site that uses Gravity, your electromagnet gets activated and the right information starts flowing your way. We connect you with content, people and advertising based on the probability that you’ll love it. It’s about helping the right information find you.
With our platform, any website will be able to tap into Gravity's Interest Graph and enable personalized experiences for their users.
Technology
Overview
If you want to geek out on the inner workings of Gravity, read on. If you’re looking for a more simple description of what Gravity is, head to About Gravity.
Your experience on the web should reflect who you are. We call this the personalized web and we’re developing technology to make it happen.
In theory, it’s a simple idea. In practice it’s challenging to implement, a pain in the neck to collect the right data, and near-impossible to do something useful and fun with the data… but that’s our plan. Why embark on such a challenging endeavor? We think personalizing the web will make user experiences more interesting and more fun.
Two big things are needed to deliver a personalized web experience. First, you need to index the web and tag everything. You need to know what websites, content, media, products, ads, etc. are out there and what topics and interests they cover. Second, you need to create an interest graph for each person. You need to understand what people are interested in and how interested they are. Combine a rich web index and an interest graph and you can do some pretty interesting things, like filter the web based on a person’s interest.
At the heart of Gravity is a semantic engine that extracts interests from any source of information. The engine is used to understand the web and the people interacting with it. It enables us to classify information with our web ontology, to build interest graphs and to deliver a variety of personalized web experiences. We’ve filed several patents related to our technology and its application to building interest graphs and enabling the personalized web.
Semantic Engine
The basic function of the semantic engine is to take a blob of text and figure out what it’s about — like what you’re brain does when you read a newspaper article.
The first step to is to analyze the blob. We use natural language processing (NLP) extensively. Instead of simply identifying keywords, our semantic engine analyzes a variety of linguistic and statistical factors. The output of our analysis is highly structured data that contain the key characteristics of the blob — sort of like a sample of the blob’s DNA.
The next step is to compare the blob’s DNA to a DNA database; our DNA database is our dynamic web ontology. The goal is to match the DNA of the blob to DNA signatures of topics in our ontology. Your standard blob of text (a sentence) will typically generate several matches. A paragraph or page of content may generate hundreds of topic matches.
Finally, we use a modified convergence algorithm to convert a list of topic matches to a handful of interests. Our algorithm prioritizes interests based on the number of topics related to an interest and the average distance and the strength of the relationship between each topic and the interests.
That’s it. Like we said, the engine takes a blob of text and figures our what it’s about.
Web Ontology
The web ontology is our DNA database. It’s actually a data graph (not a database), which does an excellent job at representing relationships between data and lends itself to graph traversal algorithms. Our ontology has 7+ million topical nodes, further augmented by metadata we’ve created or sourced from other web services. Gravity’s semantic engine uses the dynamic web ontology to match the signature of unidentified text blobs to the signatures of known topics.
Our web ontology is built on top of Wikipedia, DBpedia, Yago and OpenCyc. These resources give our ontology its basic structure, including relationships between topics. Next, we mine the open web for topical content and related social data and map this content to the ontology — thus creating a web ontology. The web ontology is different from a normal ontology in that it’s augmented with linguistic, statistical and other metadata extracted from the web. This metadata provides the basis for the DNA signatures of topics in our database. We’ve crawled terabytes of data from major publishers, blogs, forums and other public sources of information to create the most accurate DNA signatures possible.
We built our web ontology to be adaptive over time. One of the most powerful aspects of open sources resources like Wikipedia is crowd-sourced human curation. We used a combination of human and algorithmic curation to enable our ontology to grow and adapt as the world evolves.
The Theory Behind the Interest Graph
The interest graph is an online representation of your real-world interests and a new lens through which you can play with the web. We want your interest graph to become your personal filter for the web and to help you to discover content, ideas, people, events, products and services that you’ll like.
Building an interest graph is an iterative process. Instead of just asking you what your interested in, we start by pull in information about you from the web (this might include your tweets, status updates, things you’ve liked or shared, blog articles you’ve written, or information form your profile on social networks). We use our semantic engine and dynamic web ontology to comb through your information and to identify potential interests. Our convergence algorithms reduce all your potential interests to the most meaningful ones and enable us to establish preliminary interest intensity.
Once we’ve created your initial interest graph, we give it a life of its own. Your interest graph changes dynamically over time as you interact with Gravity powered services, change your presence on the web, or makes changes to your interest graph directly. For example, we might add legos to your interest graph because of several recent tweets related, but maybe you don’t really like legos. You’ll always have an opportunity to remove interests or reduce your level of attachment.
Tuesday, November 16, 2010
Robin Li : Speech
"China 2.0' is a series of 2 two-day events at Stanford University and Beijing looking at the latest development in Chinas digital media and e-commerce sectors. Hosted by SPRIE, the Stanford Program on Regions of Innovation and Entrepreneurship.
Watch The Livestream Of The Web 2.0 Summit Here (Day 2)
Watch live streaming video from web20tv at livestream.com
Today is Day 2 of the 2010 Web 2.0 Summit. We will be livestreaming the event all day right here for those who couldn’t make it out to San Francisco (thank you, John Battelle and Tim O’Reilly). The event should start in about ten minutes at 8:30 AM PT. Speakers today include Mary Meeker, Vinod Khosla, John Doerr, Fred Wilson, Yuri Milner, Ron Conway, and Mark Zuckerberg.
Optimizely A/B Tests Its Way To $1.2 Million In Funding From A Roster Of Top Angels
Optimizely, a startup that makes it easy to run A/B tests on your website, has closed a $1.2 million funding round, with an impressive (and remarkably long) list of angel investors. The full list: Ron Conway, Chris Sacca, Steve Chen, Paul Buchheit, Ashton Kutcher, Mitch Kapor, Chris Dixon, Joshua Schachter, Naval Ravikant, Ram Shriram, Ariel Poler, Aydin Senkut, Brian Sugar, Deep Nishar, Sam Altman, Steve Huffman, Nils Johnson, Jonathan Heiliger, Keval Desai, Elad Gil, Avichal Garg.
Whew.
Optimizely, for those who haven’t tried it, lets you run A/B tests on your site by simply adding a JavaScript snippet (you configure the A/B tests themselves and can analyze the results from the service’s dashboard). The service launched in July and has since added a handful of key new features, including integration with Google Analytics (you can track your metrics from within the Analytics dashboard instead of Optimizely’s, if you prefer).
The company says it’s currently tracking around 250 million ‘events’ per month, but it isn’t sharing how many unique users it has signed up so far. It does, however, have some big customers, including the Democratic National Committee, which used it to help optimize its ‘Commit to Vote’ Facebook application.
Optimizely is also using today’s funding announcement to share some older (but cool) news: following the Haiti earthquake earlier this year, Optimizely helped the Clinton Bush Haiti Fund optimize its site — and it wound up driving an additional $1 million in contributions.
Google Goggles Tests Ads Triggered By Your Mobile Camera
Google Goggles, the search giant’s mobile visual search technology, is getting a new test case: advertisements. Google is working with a number of high-profile brands, including Buick, Disney, Diageo, T-Mobile and Delta Airlines to offer “Goggles-enabled” print ads.
When users take pictures of these individual ads with Google Goggles on their Android or iPhone, they will be able to click to a mobile website from the brand.
The advertisements are similar to scanning a QR Code and receiving further information about a product. I wonder how many people will actually be interested in unlocking the information via Goggles without an incentive like a deal or coupon associated with the ad.
Of course, there are a number or interesting use cases Google could turn on for Goggles that involve advertisements or discounts. For example, Google could allow users to take pictures of products in stores to access a coupon.
Patrick Chappatte: The power of cartoons
So yeah, I'm a newspaper cartoonist -- political cartoonist. I don't know if you've heard about it -- newspapers? It's a sort of paper-based reader. (Laughter) It's lighter than an iPod. It's a bit cheaper. You know what they say? They say the print media is dying. Who says that? Well, the media. But this is no news, right? You've read about it already.
(Laughter)
Ladies and gentlemen, the world has gotten smaller. I know it's a cliche, but look, look how small, how tiny it has gotten. And you know the reason why, of course. This is because of technology. Yeah. (Laughter) Any computer designers in the room? Yeah well, you guys are making my life miserable, because track pads used to be round, a nice round shape. That makes a good cartoon. But what are you going to do with a flat track pad, those square things? There's nothing I can do as a cartoonist. Well, I know the world is flat now. That's true. And the Internet has reached every corner of the world, the poorest, the remotest places. Every village in Africa now has a cyber cafe. (Laughter) Don't go asking for a Frappuccino there. So we are bridging the digital divide. The Third World is connected. We are connected. And what happens next? Well, you've got mail. Yeah. Well, the Internet has empowered us. It has empowered you, it has empowered me, and it has empowered some other guys as well.
(Laughter)
You know, these last two cartoons, I did them live during a conference in Hanoi. And they were not used to that in communist 2.0 Vietnam. (Laughter) So I was cartooning live on a wide screen -- it was quite a sensation -- and then this guy came to me. He was taking pictures of me and of my sketches, and I thought, "This is great, a Vietnamese fan." And as he came the second day, I thought, "Wow, that's really a cartoon lover." And on the third day, I finally understood, the guy was actually on duty. So by now, there must be a hundred pictures of me smiling with my sketches in the files of the Vietnamese police.
(Laughter)
No, but it's true: the Internet has changed the world. It has rocked the music industry. It has changed the way we consume music. For those of you old enough to remember, we used to have to go to the store to steal it. (Laughter) And it has changed the way your future employer will look at your application. So be careful with that Facebook account. Your momma told you, be careful. And technology has set us free. This is free WiFi. but yeah, it has. It has liberated us from the office desk. This is your life. Enjoy it. (Laughter) In short, technology, the internet, they have changed our lifestyle. Tech guru, like this man -- that a German magazine called the philosopher of the 21st century -- they are shaping the way we do things. They are shaping the way we consume. They are shaping our very desires. (Laughter) (Applause) You will not like it. And technology has even changed our relationship to God.
(Laughter)
Now I shouldn't get into this. Religion and political cartoons, as you may have heard, make a difficult couple, ever since that day in 2005, when a bunch of cartoonists in Denmark drew cartoons that had repercussions all over the world, demonstrations, fatwa. They provoked violence. People died in the violence. This was so sickening. People died because of cartoons. I mean -- I had the feeling at the time that cartoons had been used by both sides, actually. They were used first by a Danish newspaper, which wanted to make a point on Islam. A Danish cartoonist told me he was one of the 24 who received the assignment to draw the prophet. 12 of them refused. Did you know that? He told me, "Nobody has to tell me what I should draw. This is not how it works." And then, of course, they were used by extremists and politicians on the other side. They wanted to stir up controversy. You know the story. We know that cartoons can be used as weapons. History tells us, they've been used by the Nazis to attack the Jews. And here we are now. In the United Nations, half of the world is pushing to penalize the offense to religion -- they call it the defamation of religion -- while the other half of the world is fighting back in defense of freedom of speech. So the clash of civilizations is here, and cartoons are at the middle of it? This got me thinking. Now you see me thinking at my kitchen table. And since you're in my kitchen, please meet my wife.
(Laughter)
In 2006, a few months after, I went Ivory Coast -- Western Africa. Now, talk of a divided place. The country was cut in two. You had a rebellion in the north, the government in the south -- the capital, Abidjan -- and in the middle, the French army. This looks like a giant hamburger. You don't want to be the ham in the middle. I was there to report on that story in cartoons. I've been doing this for the last 15 years. It's my side job, if you want. So you see the style is different. This is more serious than maybe editorial cartooning. I went to places like Gaza during the war in 2009. So this is really journalism in cartoons. You'll hear more and more about it. This is the future of journalism, I think.
And of course, I went to see the rebels in the north. Those were poor guys fighting for their rights. There was an ethnic side to this conflict as very often in Africa. And I went to see the Dozo. The Dozo, they are the traditional hunters of West Africa. People fear them. They help the rebellion a lot. They are believed to have magical powers. They can disappear and escape bullets. I went to see a Dozo chief. He told me about his magical powers. He said, "I can chop your head off right away and bring you back to life." I said, "Well, maybe we don't have time for this right now." (Laughter) "Another time."
So back in Abidjan, I was given a chance to lead a workshop with local cartoonists there, and I thought, yes, in a context like this, cartoons can really be used as weapons against the other side. I mean, the press in Ivory Coast was bitterly divided. It was compared to the media in Rwanda before the genocide. So imagine. And what can a cartoonist do? Sometimes editors would tell their cartoonists to draw what they wanted to see, and the guy has to feed his family, right. So the idea was pretty simple. We brought together cartoonists from all sides in Ivory Coast. We took them away from their newspaper for three days. And I asked them to do a project together, tackling the issues affecting their country in cartoons, yes, in cartoons. Show the positive power of cartoons. It's a great tool of communication for bad or for good. And cartoons can cross boundaries, as you have seen. And humor is a good way, I think, to address serious issues. And I'm very proud of what they did. I mean, they didn't agree with each other -- that was not the point. And I didn't ask them to do nice cartoons. The first day, they were even shouting at each other. But they came up with a book, looking back at 13 years of political crisis in Ivory Coast.
So the idea was there. And I've been doing projects like this, in 2009 in Lebanon, this year, in Kenya, back in January. In Lebanon, it was not a book. The idea was to have -- the same principal, a divided country -- take cartoonists from all sides and let them do something together. So in Lebanon, we enrolled the newspaper editors, and we got them to publish eight cartoonists from all sides all together on the same page, addressing the issue affecting Lebanon, like religion in politics and everyday life. And it worked. For three days, almost all the newspapers of Beirut published all those cartoonists together -- anti-government, pro-government, Christian, Muslim, of course, English-speaking, well, you name it. So this was a great project. And then in Kenya, what we did was addressing the issue of ethnicity, which is a poison in a lot of places in Africa. And we did video clips. You can see them if you go to YouTube/KenyaTunes.
So, preaching for freedom of speech is easy here, but as you have seen in contexts of repression or division, again, what can a cartoonist do? He has to keep his job. Well I believe that in any context anywhere, he always has the choice at least not to do a cartoon that will feed hatred. And that's the message I try to convey to them. I think we all always have the choice in the end not to do the bad thing. But we need to support these [unclear], critical, responsible voices in Africa, in Lebanon, in your local newspaper, in the Apple store. Today, tech companies are the world's largest editors. They decide what is too offensive to too provocative for you to see. So really, it's not about the freedom of cartoonists; it's about your freedoms. And for dictators all over the world, the good news is when cartoonists, journalists and activists shut up.
Thank you.
(Applause)
Vidyo shows off multi-party corporate videoconferencing on iPads and mobile
If you are considering a room-based videoconferencing system, you should check out Vidyo, which lets multiple parties participate. Really cool, too. But now they are also showing off that they can run on iPads and other mobile devices and slates.
Torsten Reil builds better animations
托司登里爾(Torsten Reil)介紹了一項在牛津研究發展的科技，這項科技能夠模擬人類，模擬身體及其神經控制系統。這項科技可以用在特技演員，因為特技很危險且很貴，還有更多的特技是鏡頭做不到的，為了不讓特技演員受傷，可以大大改善電影的製作方式。除了可以改變好萊塢電影和電玩的製作方式，還將被應用在幫助腦性麻痺兒童開刀的外科醫生來預測孩子手術的結果。
I'm going to talk about a technology that we're developing at Oxford now that we think is going to change the way that computer games and Hollywood movies are being made. That technology is simulating humans. It's simulated humans with a simulated body and a simulated nervous system to control that body. Now, before I talk more about that technology let's have a quick look at what human characters look like at the moment in computer games. This is a clip from a game called Grand Theft Auto 3. We already saw that briefly yesterday. And what you can see is it is actually a very good game. It's one of the most successful games of all time. But what you'll see is that all the animations in this game are very repetitive. They pretty much look the same. I've made him run into a wall here, over and over again. And you can see he looks always the same. The reason for that is that these characters are actually not real characters. They are a graphical visualization of a character.
To produce these animations an animator at a studio has to anticipate what's going to happen in the actual game and then has to animate that particular sequence. So he or she sits down, animates it, and tries to anticipate what's going to happen, and then these particular animations are just played back at appropriate times in the computer game. Now, the result of that is that you can't have real interactivity. All you have is animations that are played back at more or less the appropriate times. It also means that games aren't really going to be as surprising as they could be because you only get out of it, at least in terms of the character, what you actually put into it. There's no real emergence there.
And thirdly, as I said, most of the animations are very repetitive because of that. Now, the only way to get around that is to actually simulate the human body and to simulate that bit of the nervous system of the brain that controls that body. And maybe if I could have you for a quick demonstration to show what the difference is -- because, I mean, it's very, very trivial. If I push Chris a bit, like this, for example he'll react to it. If I push him from a different angle he'll react to it differently, and that's because he has a physical body, and because he has the motor skills to control that body. It's a very trivial thing. It's not something you get in computer games at the moment at all. Thank you very much. Chris Anderson: That's it?
Torsten Reil: That's it, yes. So that's what we're trying to simulate -- not Chris specifically, I should say, but humans in general. Now, we started working on this a while ago at Oxford University, and we tried to start very simply. What we tried to do was teach a stick figure how to walk. That stick figure is physically stimulated. You can see it here on the screen. So it's subject to gravity, has joints, et cetera. If you run the simulation it will just collapse, like this. The tricky bit is now to put an AI controller in it that actually makes it work. And for that, we use the neural network which we based on that part of the nervous system that we have in our spine that controls walking in humans. It's called the central pattern generator. So we simulated that as well, and then the really tricky bit is to teach that network how to walk. For that we used artificial evolution -- genetic algorithms.
We heard about those already yesterday and I suppose that most of you are familiar with that already. But, just briefly, the concept is that you create a large number of different individuals, neural networks in this case, all of which are random at the beginning. You hook these up -- in this case to the virtual muscles of that two-legged creature here -- and hope that it does something interesting. At the beginning they're all going to be very boring. Most of them won't move at all, but some of them might make a tiny step. Those are then selected by the algorithm, reproduced with mutation and re-combinations to introduce sex as well. And you repeat that process over and over again until you have something that walks -- in this case, in a straight line like this. So that was the idea behind this. When we started this I set up the simulation one evening. It took about three to four hours to run the simulation. I got up the next morning, went to the computer and looked at the results, and was hoping for something that walked in a straight line, like I've just demonstrated, and this is what I got instead.
(Laughter)
So it was back to the drawing board for us. We did get it to work eventually, after tweaking a bit here and there. And this is an example of a successful evolutionary run. So what you'll see in a moment is a very simple biped that's learning how to walk using artificial evolution. At the beginning it can't walk at all, but it will get better and better over time. So this is the one that can't walk at all.
(Laughter)
Now, after five generations of applying evolutionary process, the genetic algorithm is getting a tiny bit better.
(Laughter)
Generation ten and it'll take a few steps more. Still not quite there. But now after generation 20 it actually walks in a straight line without falling over. That was the real breakthrough for us. It was academically quite a challenging project, and once we had reached that stage we were quite confident that we could try and do other things as well with this approach -- actually simulating the body and simulating that part of the nervous system that controls it. Now, at this stage it also became clear that this could be very exciting for things like computer games or online worlds. What you see here is the character standing there, and there's an obstacle that we put in its way. And what you see is, it's going to fall over the obstacle. Now, the interesting bit is, if I move the obstacle a tiny bit to the right, which is what I'm doing now, here, it will fall over it in a completely different way. And again, if you move the obstacle a tiny bit, it'll again fall differently.
(Laughter)
Now, what you see, by the way, at the top there, are some of the neural activations being fed into the virtual muscles. Okay. That's the video. Thanks. Now this might look kind of trivial but it's actually very important because this is not something you get at the moment in any interactive or any virtual worlds. Now, at this stage, we decided to start a company and move this further because obviously this was just a very simple, blocky biped. What we really wanted was a full human body, so we started the company. We hired a team of physicists, software engineers and biologists to work on this, and the first thing we had to work on was to create the human body, basically. It's got to be relatively fast so you can run it on a normal machine, but it's got to be accurate enough so it looks good enough, basically.
So we put quite a bit of biomechanical knowledge into this thing, and tried to make it as realistic as possible. What you see here on the screen right now is a very simple visualization of that body. I should add that it's very simple to add things like hair, clothes, et cetera, but what we've done here is use a very simple visualization so you can concentrate on the movement. Now, what I'm going to do right now, in a moment, is just push this character a tiny bit and we'll see what happens. Nothing really interesting, basically. It falls over but it falls over like a rag doll, basically. The reason for that is that there's no intelligence in it. It becomes interesting when you put artificial intelligence into it. So this character now has motor skills in the upper body. Nothing in the legs yet, in this particular one. But what it will do -- I'm going to push it again. It will realize autonomously that it's being pushed. It's going to stick out its hands. It's going to turn around into the fall and try and catch the fall. So that's what you see here.
Now, it gets really interesting if you then add the AI for the lower part of the body as well. So here we've got the same character. I'm going to push it a bit harder now, harder than I just pushed Chris. But what you'll see is it's going to receive a push now from the left. What you see is it takes steps backwards -- it tries to counter-balance, it tries to look at the place where it thinks it's going to land. I'll show you this again. And then finally hits the floor. Now, this becomes really exciting when you push that character in different directions, again, just as I've done. That's something that you cannot do right now. At the moment you only have empty computer graphics in games. What this is now is a real simulation. That's what I want to show you now.
So here's the same character with the same behavior I've just shown you, but now I'm just going to push it from different directions. First starting with a push from the right. This is all slow motion, by the way, so we can see what's going on. Now, the angle will have changed a tiny bit so you can see that the reaction is different. Again, a push, now this time from the front. And you see it falls differently. And now from the left. And it falls differently. It was really exciting for us to see that. That was the first time we've seen that. This is the first time the public sees this as well because we have been in stealth mode. I haven't shown this to anybody yet. Now, just a fun thing. What happens if you put that character -- this is now a wooden version of it, but it's got the same AI in it -- but if you put that character on a slippery surface, like ice. We just did that for a laugh, just to see what happens.
(Laughter)
And this is what happens.
(Laughter)
(Applause)
It's nothing we had to do about this. We just took this character that I just talked about, put it on a slippery surface, and this is what you get out of it. And that's a really fascinating thing about this approach. Now, when we went to film studios and games developers and showed them that technology, we got a very good response. And what they said was, the first thing they need immediately is virtual stuntmen. Because stunts are obviously very dangerous, they're very expensive, and there are a lot of stunt scenes that you cannot do obviously because you can't really allow the stuntman to be seriously hurt. So they wanted to have a digital version of a stuntman and that's what we've been working on for the past few months. And that's our first product that we're going to release in a couple of weeks. So here are just a few very simple scenes of the guy just being kicked. That's what people want. That's what we're giving them.
(Laughter)
You can see, it's always reacting. This is not a dead body. This is a body who basically, in this particular case, feels the force and tries to protect its head. Only, I think it's quite a big blow again. You feel kind of sorry for that thing, and we've seen it so many times now that we don't really care any more.
(Laughter)
There are much worse videos than this, by the way, which I have taken out, but ... Now, here's another one. What people wanted as a behavior was to have an explosion, a strong force applied to the character, and have the character react to it in mid-air. So that you don't have a character that looks limp, but actually a character that you can use in an action film straight away that looks kind of alive in mid-air as well. So this character is going to be hit by a force, it's going to realize it's in the air and it's going to try and, well, stick out its arm in the direction where it's landing. That's one angle, here's another angle. We now think that the realism we're achieving with this is good enough to be used in films.
And let's just have a look at a slightly different visualization. This is something I just got last night from an animation studio in London, who are using our software and experimenting with it right now. So this is exactly the same behavior that you saw, but in a slightly better rendered version. So if you look at the character carefully you see there are lots of body movements going on, none of which you have to animate like in the old days. Animators had to actually animate them. This is all happening automatically in the simulation. This is a slightly different angle, and again a slow motion version of this. This is incredibly quick. This is happening in real time. You can run this simulation in real time, in front of your eyes, change it if you want to, and you get the animation straight out of it. At the moment, doing something like this by hand would take you probably a couple of days.
This is another behavior they requested. I'm not quite sure why, but we've done it anyway. It's a very simple behavior that shows you the power of this approach. In this case the character's hands are fixed to a particular point in space, and all we've told the character to do is to struggle. And it looks organic. It looks realistic. You feel kind of sorry for the guy. It's even worse -- and that is another video I just got last night -- if you render that a bit more realistically.
Now, I'm showing this to you just to show you how organic it actually can feel, how realistic it can look. And this is all a physical simulation of the body, using AI to drive virtual muscles in that body. Now, one thing which we did for a laugh was to create a slightly more complex stunt scene, and one of the most famous stunts is the one where James Bond jumps off a dam in Switzerland and then is caught by a bungee. Got a very short clip here.
Yes, you can just about see it here. In this case they were using a real stunt man. It was a very dangerous stunt. It was just voted, I think in the Sunday Times, as one of the most impressive stunts. Now, we've just tried and looked at our character and asked ourselves, "Can we do that ourselves as well?" Can we use the physical simulation of the character, use artificial intelligence, put that artificial intelligence into the character, drive virtual muscles, simulate the way he jumps off the dam, and then skydive afterwards, and have him caught by a bungee afterwards? We did that. It took about two hours, pretty much, to create the simulation. And that's what it looks like, here. Now, this could do with a bit more work. It's still very early stages, and we pretty much just did this for a laugh just to see what we'd get out of it. But what we found over the past few months is that this approach that we're pretty much standard upon is incredibly powerful. We are ourselves surprised what you actually get out of the simulations. There's very often very surprising behavior that you didn't predict before.
There's so many things we can do with this right now. The first thing, as I said, is going to be virtual stunt men. Several studios are using this software now to produce virtual stunt men, and they're going to hit the screen quite soon, actually, for some major productions. The second thing is video games. With this technology video games will look different and they will feel very different. For the first time you'll have actors that really feel very interactive, that have real bodies that really react. I think that's going to be incredibly exciting. Probably starting with sports games, which are going to become much more interactive. But I particularly am really excited about using this technology in online worlds, like there for example, that Tom Melcher has shown us. The degree of interactivity you're going to get is totally different, I think, from what you're getting right now.
A third thing we are looking at and very interested in is simulation. We've been approached by several simulation companies, but one project we're particularly excited about, which we're starting next month, is to use our technology, and in particular, the walking technology, to help aid surgeons who work on children with cerebral palsy, to predict the outcome of operations on these children. As you probably know, it's very difficult to predict what the outcome of an operation is if you try and correct the gait.
The classic quote is, I think, it's unpredictable at best, is what people think right now, is the outcome. Now, what we want to do with our software is allow our surgeons to have a tool. We're going to simulate the gait of a particular child and the surgeon can then work on that simulation and try out different ways to improve that gait before he actually commits to an actual surgery. That's one project we're particularly excited about, and that's going to start next month. Just finally, this is only just the beginning. We can only do several behaviors right now. The AI isn't good enough to simulate a full human body. The body yes, but not all the motor skills that we have. And, I think, we're only there if we can have something like ballet dancing. Right now we don't have that but I'm very sure that we will be able to do that at some stage.
We do have one unintentional dancer actually, the last thing I wanted to show you. This was an AI contour that was produced and evolved -- half-evolved, I should say -- to produce balance, basically. So you kick the guy and the guy's supposed to counter-balance. That's what we thought was going to come out of this. But this is what emerged out of it in the end.
(Music)
Bizarrely, this thing doesn't have a head. I'm not quite sure why. So this was not something we actually put in there. He just started to create that dance himself. He's actually a better dancer than I am, I have to say. And what you see after a while -- I think he even goes into a climax right at the end. And I think, there you go.
(Laughter)
So that all happened automatically. We didn't put that in there. That's just the simulation creating this itself, basically. So it's just --
(Applause)
Thanks. Not quite John Travolta yet, but we're working on that as well, so thanks very much for your time. Thanks.
(Applause)
Chris Anderson: Incredible. That was really incredible.
Torsten Reil: Thanks.
Subscribe to:
Posts (Atom)
