 My name is Sudhir. I've been with Google for about three years. And I've been in the learning field probably for the last 15 with NASA, Accenture, Deloitte, and now Google. Google does a lot in the education space, but in probably very distinct areas. So one is the Google Apps for Education. I don't know if any universities here are using the Google Apps product. But that is definitely one area where we're trying to make a headway in the education space. That's one area. The second area is there's a lot of work that Google is doing for external communities in terms of, particularly in the developing world, about how to actually raise educational standards in developing world. So there's lots of special projects going on within Google in that space, more so recently. The third space is what I'm involved in. And that's probably what you see up there, which is Google EDU learning apps. Now Google EDU is just an internal name at Google for our internal corporate learning. We have about 22,000 Googlers. That's what we call people that work at Google. And we really look at ways to actually make sure that this 22,000 workforce is always learning, is always learning. And it's not just a question of the fact that they're learning, but do we know what they need to learn? Like what's the next big thing that they actually do need to learn? So the learning labs team, which I manage, looks at this area. And we have a whole host of projects that are running to actually look at enabling our workforce to learn, and coming up in some cases tools and products to actually enable our own workforce. Some of these products we'd love to see taken externally, but right now the predominant focus is internal learning at Google. So this is not a comprehensive list of projects, because that would take a lot longer than I have. But I at least wanted to give you a taste of some of the things that we're working on. So before I kind of go any further, actually, I wanted to talk a little bit about what inspires our team. Because we don't feel like at Google that we have all the answers, but what we're inspired by is actually a lot of what I've heard today in lots of different sessions about what the world is doing. So the first thing I wanted to show you actually is this, the world. And this is actually a plasma screen that if you walk into the main building at Google's headquarters in California, there's this big plasma screen, and this is the image on it. It's a big earth, and it rotates very slowly, and it has all these spikes of light on it. And what these spikes of light represent is actually the volume of search queries that are going on at any particular time. So it's pretty awe-inspiring just to watch this thing and to see all these active search queries. And the different colors represent the different languages those queries are made in. And it's awe-inspiring. It shows that, OK, Google's working. People are querying. And for me, being in learning, what this actually symbolizes to me is that we're actually quite an inquisitive planet and constantly searching, constantly inquiring. And for me, it's one of the things that inspires us about learning is just people asking questions and wanting to know something. So let's look at what the world is doing. So what are these people doing when they're online? So one thing I mentioned is search. And so the world is searching for all sorts of information. And Google has a way of actually we capture search logs so we can kind of actually, when Seb walked into the London office, there's actually a live search query stream that's filtered somewhat that tell you what people are searching for. So I just remember when I was trapped in the volcanic ash cloud in Ireland in Stranton, maybe some of you were also affected. This was kind of my little learning journey to learn about what was going on is how can I learn more about these volcanic ash clouds? What are the newspapers around the world saying? I wanted to learn a little bit about volcanoes I didn't understand. So I could do a book search. I could look at Google Earth to actually figure out where that isatic volcano actually existed. And YouTube actually had a lot of videos on the ash cloud itself, too. So we see a search stream that people are doing across the different vertical products at Google that actually help them learn about a particular subject. So that's one thing that we're seeing is lots of searches and over a billion searches a day at Google. What else are we doing? Well, there's approximately 1.4 billion people online. And a good percentage of that, I think, according to the last Pew study, I think around 40% of those people are creating some sort of content. And I don't know if you guys remembered, but Time Magazine does this person of the year, man of the year, every year. And a few years back, they didn't actually award it to a person, but they awarded it to you, to the people of the planet, because they realized that there was a huge phenomenon taking place that was actually the individual contributions of millions and millions of people that were really changing what we can do today. So that's one thing we're doing, is creating lots of content. And where do we see this content being created? Well, Wikipedia still seizes to amaze me what a wonderful resource it is. And if someone actually pitched the idea of Wikipedia to me, I wouldn't actually expect it to be as successful as it's been. Over 3.4 million articles in English. And I think there's only 20 million plus overall that have been written on Wikipedia. What else are we doing? We're also answering each other's questions online. Is any developers in here, are you guys familiar with Stack Overflow? Or do you use it? Do you like it as a resource? Yeah, it's a pretty amazing resource. For those that are not familiar with Stack Overflow, it's a place where you just simply can post a question, usually development-related, and it gets answered by the community. And what's amazing is how fast things get answered. And things get answered literally in seconds, usually. So the developer community is very strong and uses a site like this to actually enable and help each other learn. What else are we doing? We're teaching. There's a whole host of products, but obviously Google sees a lot of activity in YouTube. And this is an area that we've looked at. I remember spending one night actually trying to stump YouTube, seeing how to almost anything just to see if I could stump it. And actually, I really couldn't. I couldn't, you know, as how to program in Python. And there was a whole host of tutorials out there. There are actually some of them which were actually quite good. And you can learn almost anything. I remember typing in how to moonwalk. And people are just teaching anything, and lots are watching. Just to reiterate what we said this morning, we have over 7 million views on this one video. If you just think about the audience that this guy in his bedroom teaching moonwalking is able to reach, it's really an incredible phenomenon. And it's not just moonwalking, and it's not just programming tutorials, but some people are really using YouTube in the video format to really form learning academies. This is a picture or an article about Salman Khan in his Mountain View studio where he's created over 1,100 courses. He started in math, but now it covers a number of different areas that are actually been highly rated and really good. And he's formed this Khan Academy. I don't know if anyone has seen this before, but there's an article written about the New York Times. And it's gotten Google's attention as well in terms of the quality of content that's being put on YouTube. And we've been talking, I've heard folks here talk about YouTube EDU. It was mentioned this morning that this university is also on YouTube EDU, but that's another area where we see all the lectures, be it a bad word or a good word, I'm not sure, being put online. So that's another area that we see a lot of activity. So the thing is, what does actually this mean for Gula? We see this kind of incredible phenomenon. We're actually humbled by the world, what they've been able to do, answering each other's questions, creating all sorts of content, collaborating at such a large scale. How can we actually learn from what the planet is doing and actually apply it to an enterprise like Google? So the question that we ask our team, and this is one of our projects, is how can we take what's going on in the world and actually apply it to the people at Google? So I'm not suggesting that what I'm about to talk about is necessarily based on academic findings or necessarily best practices. It's just Google's a very experimental company, so we're trying things. We're trying things to see if we can sort of create this enabling learning environment that we're seeing in the world, this ecosystem, if you will, and actually try and create that same sort of ecosystem within Google itself with our 20,000 plus employees. So Google's kind of in a good place to do that. One, we have a lot of products, as you guys probably know, that actually already lend itself to this space, and the ones I've already mentioned. So they happen to be in the search, content creation, and collaboration space. So there's a lot of products as our disposal, as a learning team, to actually figure out how we can actually use in the learning space. So that's one thing we have in our advantage. The second thing we have in our advantage is that Google has, luckily, it has a culture that really encourages this type of behavior. They want people to share knowledge. They really want to enable people to do it. I mean, you could argue that Google's trying to do it for the planet, so obviously they're going to do it for their employees as well. This is actually an excerpt out of the IPO letter that Larry and Sergey wrote saying that Google's not a conventional company. I just wanted to highlight a few words that create an atmosphere of creativity and challenge. And I see that that is true. They do try to infuse that sort of thinking, to try and do things differently, to really try and stretch you, think creatively. The engineers plus more in the company have what's called 20% time, where you can dedicate 20% of your time to something that you're passionate about. Of course, that is also in the business interest of Google, but not as directly as you would think. When you talk about the atmosphere of creativity and challenge, for people that, when Seb came and visited the office, some people kind of see it in a physical manifestation, the fact that there's kind of the lava lamps and the hammocks and the food and bringing your dogs to work. To me, that's nice to be able to say, and you invest, create an atmosphere, but I think it's more than that, actually, too. It has to be actually ingrained in people, and it's kind of the attitude that they bring. For me, this was kind of apparent, actually, when I joined Google, because this was all new to me as well, and when I went to the Lou, actually. And this was when I was really thinking about, OK, Google's an unconventional company. How are we going to actually do learning and development differently at this company? And that's when I went to the Lou, and I saw this poster here, which said Learning on the Lou, Managing Meetings and Stretching on the Toilet. That's kind of a joke, but this is actually pretty serious. Learning on the Lou is just a series of training nuggets that was just created by an individual employee. It's not really sponsored or anything by the Learning and Development Group, but it's actually quite a popular series. I mean, it actually sort of works. And captive audience, it's nice and quiet. So there's a whole series, and now there's a second series for the test engineers called Testing on the Toilet that's also quite popular. So this is the kind of environment that we had to work with. So it seemed to be right to try and create this ecosystem at Google. To be honest, a lot of this would happen without us, we felt. So let me talk to you about just a few different areas that we really wanted to enable. And I'm just going to cover three projects today, kind of in line with what I just presented earlier, one being around search, one being around content creation, and the last one I just wanted to talk about answering each other's question, this whole idea of expertise location. So the first project I wanted to cover is search. And what we really attempted to do here is can we create an educational search engine? A lot of people actually come to a search engine and they want to learn something. Can we actually create something that's more specific for a learner? Actually deliver results that result in them learning about the subject as opposed to just being informed. This is just a very early thing, and we have a lot of work to do. In fact, there's another iteration of this that I don't have available yet to show. But this is one of the early things that we did. And just to let you know, the state of learning and development when I joined was pretty haphazard. Lots of different training groups. Things just started, very startupy sort of environment. I think we're at a stage now in the company where we really want to think about it more strategically and do things more thoughtful, I guess is the word. So one was thing that we wanted to do in the company was actually really hard to find learning resources. So we attempted to create a search engine for learning. So here's what you get when you type in leadership in our learning at Google portal. You get lots of different programs that have been created by various learning and development teams. You get what instructor-led courses they are. You can check the info and the availability. Some of this actually for most companies, they may not be earth shattering. But for Google, to be honest with you, we're actually starting from scratch. And so this is one of the things we developed. And lastly, we also developed Google recommended resources. So we have a mechanism where Googlers can recommend what they think is good for their role. And then that would show up in your search results as well, things that have been recommended by other employees depending on that search term. So that's one of the projects that we've done and completed and that we're iterating on. One reason I put this up here, because not that this is necessarily earth shattering, but it's kind of one of the side benefits of the projects, that we have a tangential project coming off of this. Because what we do is actually capture that search log on this site. And so when people type in leadership or people type in things they want to learn, we capture that stream and we collect that data. And we found it to be really interesting, actually. And I don't know if you're familiar with some of the traditional companies I've been in. We've done an annual training needs analysis where we would survey the entire company. We would talk with leadership and management and actually figure out what the company needs to learn. What are the top five topics that folks in different groups need to learn? It was kind of an arduous process and relatively painful, actually, to actually conduct that sort of training needs analysis. We ran a training needs analysis at Google in the traditional way the first year. But what I also did was we had the site running for some time. And what I did was compare the top search log entries with the results of the training needs analysis tool. And we're able to look at the search logs by role, function, et cetera. And what we found is actually the top five search queries and the top five things in this painful TNA process actually correlated exactly, which was really kind of interesting to us. So we're thinking that maybe there is more interesting, implicit ways to do training needs analysis in a company by looking at these sorts of data. Could be at least one data point. So that's one thing. Second project I wanted to talk about is our goal of getting Googlers creating content and teaching each other. So we started this program that we called G2G which stands for Googler to Googler. And it's been fairly successful. We have a whole program where we encourage Googlers to actually teach a class and we enable them to teach them or we have a whole design support team. We train them how to teach. And because there's a lot of people in the company that do wanna share their knowledge but they just feel like they just need that kind of catalyst from the L and D team to do it. So we have some fun classes. We have some serious classes. We just kinda let anything go. And then we've also started sort of a video sharing site within Google, kind of our internal YouTube, if you will. And we just call it G2GTV where we encourage lots of folks to create short snippets of video on a wide variety of topics. And we have a search engine and everything that looks at those videos. In fact, that search engine I just showed you also scans all the videos as well and makes them appear. So here's a little two minute video that Emma, who that is, did on time management. So that's another project. What we found with the video creation is that we're trying to, our strategy in the company is to kinda lower the barrier to create content. Because the biggest barriers at Google that we're finding is just simply time. Like people say, this is not my job. I'm just too busy doing my normal work. Or am I gonna find time to record a video? So we're putting things in place to actually make it easier and easier. So one of the things that we're doing that we started this year is that we started this concept of a learning media studio that we rolled out to 11 offices at Google where this is in its very kind of early days. But we wanna kinda create an environment where people feel inspired to actually create what we're just calling learning media, be it an online course, be it a quick video. And right now the rooms are about kind of having all the right equipment already configured, soundproof, microphone, the room is set up. But we wanna take it beyond that actually. We actually wanna make the whole environment a little bit more conducive to the whole design process. Cause it's not just about the tools and the technology, it's about the whole design process. So we wanna make the rooms actually facilitate the design. So that's another thing we're working on with these studios. Which brings me to the last project, and this is the one I believe I so just said, and which is one of the problems that we identified, and that's completely how we work, we actually identify what problems we're trying to solve and figure out ways to solve them, is with 20,000 people at Google, it's actually a very hard problem for us to figure out who is good at what. And I think it happens in most companies and how do you know who is good at what? Because it's great to say we wanna tap the knowledge of everyone, but actually we don't know who in the company is good at stuff. And you can't make the assumption these days that just because someone is in the role of finance that they have so much other knowledge that has nothing to do with finance because they may have been in product in their previous job and they had actually quite good at product as well. So every individual is unique, they come up with a unique CV and skill set. So we wanted to make sure that we understand who these people are. And in previous companies and even in my first year at Google, you might have these skill surveys that are set that where people actually self-mark or rate their own skills and you collect all the data and that data is outdated after a year and no one wants to update it. And you kind of have CVs that are out of date which seems to be, at least in my previous company was always a problem, but of actually just maintaining that sort of data on folks. So what we decided to do is try something a little bit different. And what we did was we came up with this product that we're calling G-Wiz and what it's based is, if I can summarize it in maybe a tagline it's basically crowdsourcing your CV, actually getting others to tell you what you're good at. You can also, whoops, sorry you shouldn't do that. Lost my train of thought. So telling, getting others to tell you what you're good at including yourself. So it's simply based on skill tagging. So if I think SEV is good at conference management I would just, I would just, in this tool I would be able to tag SEV with conference organization and management and it's simple. We don't restrict what the tags, they can be anything you want. This is the homepage of the site. Those are the tags that I've been tagged with by my peers and myself. And here's some of the stuff that in a word cloud that people have tagged with other stuff. And you can see some serious stuff like Python and Java and JavaScript. You also see stuff like making cheese and making stuff up and I've seen weaponizing office supplies. But we let kind of anything go and it just makes for interesting data and people actually have more fun with the tool and if people have fun with the tool that's actually great for us because we just get more and more data. So what happens is that people use this tool to tag other folks but first and foremost it's a utility. So it's sold as a search engine within the company. Like if you want to find somebody that's good at a particular skill this is where you would go. And so for example if I were to type in CSS into the search box and hit search Gwiz I'd get a list of everyone in the company that's been tagged with CSS. And it was actually the first time in the company that wow this list has been creating. It wasn't created by a group it was actually created by Googlers. So and the way we rank folks is like right now the ranking algorithm is quite simple. It's just the more tags you've received like if somebody received five tags for CSS from five different people you'd be ranked higher. So that's just at least a rudimentary way of ranking people. We have some ideas for more advanced ranking algorithms but that's what we've done for now. So there you go and then there's a little plus box next to it so if you click the little plus you can see everything else that person has been tagged with so Damian in this case has been tagged with CSS, foosball, JavaScript, Prometheus which is Google technology and Python. And if I agree that Damian is good at these things I can just click these green buttons saying add your vote and effectively what it does is just invoke another tag from me to this person and then it would adjust the ranking if necessary. So it was actually quite a it's a simple tool and concept but we have been running it for a few months and we have over 4,000 Googlers tagged over 10,000 skills in there and we're looking for ways to market and grow it and we wanna just let it grow. So we're just letting the people tag each other and we're just kind of hopefully we'll just continue to see an uptick but it's a way of us actually identifying folks that are good at particular skills. I personally would love to see this grow some more and oops let me talk about Gwizzle just a little bit more. What am I doing on time? Okay now we're good, we're good. And then so with this data I mean it's actually a very simple data set it's just if you look at the data set it's just the person that's doing the tagging the person they've tagged and what skill they tagged them with and a date timestamp. So it's actually a very simple data set but what we're trying to do is visualize this data even more because by merging this with organizational data we're creating things like if I wanna see across the world where the Python skills lie within Google I should be able to do that with this. And in all these skills eventually we wanna make it more of a it's kind of a rudimentary CV if you will. If you start to look at all the tags and there's other data that we can put next to it in this next iteration it starts as kind of your crowdsourced CV. Another idea that we have with this is skills correlation. You know the fact that we have actually these kind of wacky skills in there or hobbies are in there like if I type in skiing or chess or any hobby I'll get a list of folks at Google. And I'd be interested in doing some sort of skills correlation because I think what we found we think where it's again we're not we haven't done this very scientifically but I think people that are good at crossword puzzles also happen to be good at AdWords optimization which is a Google skill. So it's just it's really interesting to actually try and correlate even skills like that. We haven't that we need a lot more data to do better analysis. So hopefully one day we'll do that. And actually one thing is even just a community just creating a community of folks around a skill. So if I were to type in skiing or CSS at least I haven't kind of an instant list of folks which make up an instant community of folks that have something to do with CSS of the company and before that was a little bit more difficult thing to do. So those are the projects I wanted to highlight. So just talk a little bit about G to G what we're doing in search. Our team is exploring a number of different areas virtual instructor led platforms what we can do better with mobile. But these are some of our key projects that we're working on at the moment. So with that is back to the beginning as you know as a team we've been inspired by just what folks have been doing in the world. I actually think the situation's reversed now right. It's used to used to look at companies to see what new technology they come up with and then they'd be adopted at home. Now you know stuff that's happening at home is usually more advanced that's what's happening in the workplace. So we're inspired by that. And so with that can take questions. Thanks, Sylia. And we've got two microphone holders and we'll take questions. So please be ready to ask. Yeah, Alistair. Just suddenly coming down with a mic and the best way because it's a big rule is to grab some water. And if people can be indicating while Alistair's speaking I can get the mic organized. Thanks, Alistair Clark from NiAce. I'm just interested as to whether you see, for example, G-Wiz as a tool that might be available outside of Google. I personally would love to do that. We have a lot of work to do on it to be honest with you and my personal goal would be, I would be great for us to as we've done with some of our other products I'd love to open source it actually. But that's, we're probably not in a stage to be ready to do that at this point in time. There are some other learning products that we have recently open sourced. There was kind of a rudimentary LMS created called G-Learn. It was more to showcase what you can be doing in Google App Engine which is one of our development environments. And that was open sourced at Google IO which was our conference just past May. And you guys probably know that Google Wave was recently open sourced too. I was just curious whether you meet any resistance from employees. I mean I can understand that people who work at Google are a particular kind of person if you'd like to start with but you're really, you know, the bridge between home and work, you're really crossing that over there. Do you meet any resistance from people wanting to share with you what the kinds of skills they have at home? Oh and they, well this is completely generated by the individual and his peers, right? So like, are you talking about G-Wiz in particular and what skills they may put on their... Well certainly because you're talking right the way across the board of different kinds of skills, yes. True. So one thing I didn't mention is that the person has the final control over what actually shows up in his profile. So suppose somebody nominated you with something and you're thinking actually either I don't wanna expose that or I don't agree with it, you can take it off. So the employee always maintains his own profile. So it's never actually shown as a part of your profile. You get an email message first saying so and so has tagged you with X, you know, and then you can have the option to edit that profile. So that's how we manage it. Tom. Okay. Actually, you'd like to take the person there because they're nearer you and then in the front. I was interested in the educational search engine and how is that going to be integrated with services like Google Scholar and is that integrated with it? And I think somebody kindly answered my question on Twitter was about restricting searches to things like Creative Commons. Does the same advanced search mechanism exists in that as it does in Google generally? Right. Are you talking about the search engine that I showed you there? Yeah, yeah, the one that I saw earlier on, yeah. We're still thinking about the direction for that. I mean, the reason we built the search to begin with was to just solve a particular need. It was just actually very hard for people to find the right learning resource at Google, things that have been recommended by other employees, videos that now have been created. There's lots of different kind of repositories of information and we didn't have a good way to search that, surprisingly. But so that was the focus of that project. So I think that's the initial step but I think there's something bigger here that I think would be interested in exploring is just like how do you actually, if a user is a learner and they're going to a search engine and they really want to learn about the key where they've typed in, how can we actually create a results page that's more suited for that? I think the thing that the comment I'm making though is as soon as you put educational resources you've got a kind of worms there and teachers are gonna go in there and use those learning resources and it puts them in a position as regards things like copyright of knowing what's safe to use and what's not. And I think it can be quite a loaded term, I think in comparison to other initiatives that are going on at the moment. Just that's just my observation. Okay, yeah, this is not an external product yet and it just scours internal resources, internally created resources at the moment but I think that's definitely something to be cognizant of and bear in mind if we ever decide to actually try and create something external that looked at external resources. Got Steve at the front. I think the question's just been answered actually but it's been pre-empted but thank you anyway. Questions that answers that I'm going to anticipate the next question are very difficult to chair a session where that keeps happening. Anybody else? Yeah, Nick. And I think we'll close at that point because time has moved on. I'd just be quite interested, it wasn't what you were talking about but about the progress with Google Waves because I think quite a few people perhaps here including myself were quite enthusiastic takers up in the first place and it kind of seems to have died to death. I don't know if that's the experience in the States as well. In the States? I don't know what the, as is very recent news, the fact that we've, I mean basically with Google Wave there wasn't as much user adoption that Google was hoping for more, for more adoption. And as a result of that, they decided to basically discontinue Google Wave but make it available as an open source project. So what we're hoping is that the open source community picks up Google Wave and actually does more with it. I don't know if that answers your question. I will say, I mean I can just be honest internally, on my team we're trying to figure out ways to use Google Wave. I think it's one of those products that was actually, I'd be interested in hearing even how this audience may have used Google Wave or do they, do we have any success stories with Google Wave and education and. I wonder if it was almost a case of product overload, you know, sort of almost too many alternatives. It still seems to me like a very useful tool but if other people aren't using it, then you get fed up with it yourself. I think we've a response to your question studio for Fred over here. I know of one very successful project between University of Illinois and an institution in Southern Ireland, in the Irish Republic, which was making extensive use of Google Wave to link students and teachers internationally on course projects. Okay. Which I can send you details. Okay. Yeah, I was asked to comment by Paul Lowe, who's run Google Wave on, I guess, a not dissimilar course of the MA in photojournalism at the University of the Arts and they have students around the world on that and they use Google Wave as a coordinating tool. So he was asking you, can you continue it, because it works for us. So. Okay, I'll take that one back. Any other feedback on through this particular question? Okay. Thank you very much for, was there a, there was a hand up. Oh yeah. Yeah, I missed it. It's about the West thing, because for instance, I work in departments and I do Python, 20 people say I do Python, but everybody who does Python knows I can't really do Python. Do you also have that kind of connection between who is judging, like 10 people say you make cheese, but does it say you make horrible cheese? Do they know about cheese? Who are the judges of the skills? Who's judging? 20 people, 20 people can say, yeah, I tag him with HTML, but every web designer knows that your HTML is really not that good. So is there any? Well, we're hoping like with large numbers, you know, large number of tags that those kind of things will naturally, like if 20 people, if there was a bad Python developer and 20 of his friends tagged him, but there'll probably be other Python coders that are more respected that are just completely overshadowing the 20 tags. People who are tagging for Python work in the marketing departments, what do they know about Python, you know? Right, so who's doing the tagging? So yeah, that's an interesting point. So one thing that we thought about doing when I was talking about, we're thinking about new ideas for the ranking algorithm is to do more of a kind of the way Google does PageRank. In other words, like, if a lot of people tag this one person and this person then tags another person, in other words, if you're an expert in the field, your tag should have more weight than someone that's not good at it. That's the way it is. We're trying to, there'll be ways we can change that, where we can do that, and hopefully that'll take some of that noise out, but we don't know.