Sims 141 - Search, google, and life - Sergey brin - Google

MARTI HEARST: We have a fantastic guest speaker today, Sergey Brin, who is a co-founder of Google. You might have heard of this company. [LAUGHTER] I actually got the Wikipedia article on you, Sergey, in order to give your history. So I could sit here and read things for a couple of minutes but I don't think I'm going to do that. SERGEY BRIN: I should sit down and do that. MARTI HEARST: Sorry? SERGEY BRIN: I should sit down that. MARTI HEARST: Yeah, why don't you do that? That'd be great. I'm not going to sit here and take up the time. We don't have that much time with you so I'm just going to let you take over. SERGEY BRIN: OK. MARTI HEARST: So, Sergey Brin. [APPLAUSE] SERGEY BRIN: So I mostly want to do some Q&A here today, but I wanted to start with a few opening thoughts. And actually you reminded me of one of them which is the Wikipedia -- Wikipedia in general. There are things out there that are very simple and you never think would work. And that's why you just don't do things that you assume they basically won't work. Wikipedia is one of those that it would never occurred to me that something like that would work. And I assume many of you-- has everyone here seen Wikipedia articles? All right. Yeah and it's amazing to think that you can build an encyclopedia and anyone can edit anytime. I've gone to Wikipedia pages at first when I said, look I don't believe they're getting this content this way. Here, I'll hit the edit button and see what happens. I go on a random web page. I don't know, it was some artist, 18th century. And I made some stuff up. He really liked the colors brown and orange, something like that. And I punched it in there. And I said, come on, there's no way this is going to work. And of course, I click submit and then I view and there's the change there. And then I quickly undid it. I didn't want to pollute it. But it does work. And it works for several reasons, many that I don't understand for sure. But one of them is scale. And by virtue of the fact that there are so many people out there that are reading these Wikipedia entries, that are editing-- well, there are a smaller number editing them. And then a still smaller number that really actively monitors all of them. But still, it's a small fraction of a huge number of people. They're able to keep it to be a pretty-- a very comprehensive, reasonably high quality site. Occasionally like some of the stuff I think above me is a little bit wrong. But I don't know how it would compare to like normal encyclopedia entries. I know it not. So I think that they do very, very well and I'm very impressed. With internet search as a whole, forget about Google for a second. That too, which today we take for granted in a sense. But it was a fairly simple idea that you take all the information out there which let's say 12 years ago when the first search engine start being developed wasn't that much. But the computers were a lot less hefty then too. And you just create an index. Even a fairly basic inverted index. In fact, in the earliest days, people didn't really worry about ranking even. It wasn't that big a deal. There weren't that many matches for most searches. And AltaVista probably made the biggest leap in terms of comprehensiveness and speed and what not. And you just index it and you let everybody query it. And today it's just it's very-- we all take it for granted. But this was just a short time ago. And it wasn't at all obvious that it would work, that it would be useful or anything like that. And I would extend the same idea to the web as a whole. They were a number of hypertext experiments and systems that people put up. What was the one with a funny guy, Xanadu? Did you cover that? Yes, Ted Nelson. Who's a very interesting guy. But anyway, so he had created this thing and it wasn't the quite the same as the web but it was-- people have tried that. And yet, with a few simple ideas-- and I won't pretend to know to identify the key features that really allowed the web to grow-- but it's really became a repository of the world's knowledge. So anyway, I guess I want to finish that intro just with the point that people who have taken fairly simple ideas, ones which you might not think would work at all really, at a certain scale and after they gain a certain amount of momentum, they can really take off and work. And that's really an amazing thing. Let's see, maybe I should try to relate that to Google a little bit. I want to leave time to-- At Google we had one simple idea which now seems obvious. But the idea that the ranking does matter. And in fact that was not a high priority in a lot of information retrieval web search research at the time. That the ranking is-- I mean some people worked on it, but it wasn't that important of a thing. And we decided that for queries that really return a lot of results that we could do something more reasonable. And we sort of stumbled upon a way to do that by studying links. And I don't know if any of you have-- what have we presented here in this class to date? MARTI HEARST: [INAUDIBLE] SERGEY BRIN: OK. So you've covered a lot of stuff. And page rank? Have you? OK. I'll go through in a high level. We originally developed page rank-- well, I was kind of playing around with studying all the links on the web. And that too was a pretty revolutionary idea though it seems very simple that you could even just collect them and then do anything meaningful. Because as a graph in the computer science sense it was a very large graph compared to computers of the time. Or at least compared to our budget of computers at the time. And anyhow, I really credit Larry pursuing that idea that it's even worth collecting the graph. And then that you could run any kind of processing on it. But soon after we had it, and we had a crawler that went out, and we have to kind of develop our own RAID to be able to write to the data to the disk fast enough. And it's kind of things that are trivial today, even probably on your laptops, but were hard back then. And then they started to play with it and came up to the notion that not all web pages are created equal. People are but not web pages and some web pages are inherently not worse than others but at least less important than others. And we developed this analysis of a graph of link structure of the web that imputed an importance for every web page. And we use a similar algorithm today. There are many other algorithms that we have to run. And it's evolved a bit over the years. But it is one of the things that we continue to use. And the general concept that not all web pages are created equal is very important. The other thing I want to highlight is that when we were studying this, and we actually weren't sure that we wanted to have search as the big application, at some point we realized that this actually worked really well for a search. That if you type Berkeley-- there are a lot of pages that mention Berkeley-- but some like the Berkeley homepage are probably somewhat more important than others. And I guess there's a UC Berkeley homepage and a Berkeley city homepage. Anyway, all the Berkeley pages. And we decided that was actually very useful to search. When you had a lot of results and that if you wanted ranking to matter, that was a good way to do it. But the other thing we were kind of thinking about at the time is how would you-- we weren't kind of thinking of this as how would we let millions or hundreds of millions of people use this. But how would you even make something anyone, a single person, could use or how could you make a search that would work well. We had a phrase for it: search for kings. No, you're not searching for kings but a search that a king would use or queen. But the point was, is given the resources that we had, how would we create really good search engine, not worry about how many searches it could handle or how large a user base it could support, but to make something really, really good for a small number of people. It wasn't that we wanted to make something good for a small number of people particularly. But we wanted to get rid of that constraint that you had to scale it up to a large number of searches. But ultimately what we developed we were able to scale. And in fact in subsequent years as a company at Google when we've had sort of projects which say, well throw as much compute power at it as you want. Let's say we just want this to work well for a small number of people. We've ultimately always found ways to scale it up and deliver it to everyone. Which is kind of interesting. It's kind of like technology as an inherent democratizer. Because based on the evolution of hardware, probably more importantly the evolution of algorithms and the system software that supports these, you're able to scale sort of almost anything you can think of up. Now it takes-- it's not trivial. It takes some hard work and effort. But I think that's an interesting observation that it's-- we'll have to see if in our lifetime if that means everybody has more or less tools that are equal power. And there's not much way that you can really spend a lot more for the search and get much better results because in a short period of time technologists are able to make it work better for everyone. So anyhow, that said, I just wanted to quickly go over a little background and open it up to some questions. MARTI HEARST: [INAUDIBLE] SERGEY BRIN: Yeah. Oh, thank you. AUDIENCE: Hi. I'm curious what keeps you up most at night from a competitive standpoint? SERGEY BRIN: Well I found over the years-- if I may say that though we've only had a company for seven years or so now-- it's important that those sorts of things not keep you up at night. I mean, we obviously have big competitors. And you may have heard of some them: Micro... soft? [LAUGHTER] SERGEY BRIN: And anyways, some others nearby. But you really want to make sure you don't-- that's not what you should be spending most of your time thinking about. Though inevitably in a company that happens and lot of people worry about this company doing this thing or the other. What I think is more valuable to spend time thinking of is that we have this really incredible opportunity and we have a lot of computational resources, a lot of great people that worked there. We have the ability to analyze lots of data. And there are a terrific products and services that come out of it. I'll just throw out one example: we developed this advertising network. And a real issue for the web is you can't find something if it's not there, even though it's very comprehensive. And the high quality sites, they do have economic issues. Like they have to stay in business. And a lot of sites during the boom, the bust of the boom, went out of business. And we did have-- we wanted to make sure people could get their content out there and could afford to pay for it. So we develop this program called Adsense which allows us to put ads. And you probably see them kind of on various parts of the web now, not just on search sites but on other sites. Sometimes you might even annoyed by them. But we like to think they're better than the kind of flashy, crazy ads. Anyhow, they also create an income stream for a lot of people who are developing great content. And that's something we can do by virtue of the size of our advertising network, the information retrieval technology we have to target the ads, the scale of our brand, various other things. And that's just one example. But there many things that we can apply, the different things we have in our company, both technologies as well as reach of various kinds, and create new things. And that's exciting thing to lose sleep over. AUDIENCE: My question is, at the time that you developed and released Google there were other search engines that had been accepted as the standard and were considered to work fine and so on. What is it that inspired you to go back to the drawing board, create a new search engine, and release it? SERGEY BRIN: Great question. So at the time there were probably five major search engines or so. And you might not even remember some of them. I don't know. People here are pretty young. But what we found was we-- kind of by accident almost-- we found that this processing of the link structure of the web, we could create a search that was better in important ways. In ways that these search engines had ignored. And also they had decided that the search wasn't that important at the time. All the search engines wanted to be Yahoo! which at the time, and actually until relatively recently, didn't have a search engine at all except one that it licensed. So we found that it was actually-- we thought we'd give it a try. That's another I think important thing. There was relatively little downside to trying. I mean, we thought we have something that was pretty good and we were able to get a little bit of funding for it. We didn't need huge amounts. And my adviser assured me that I could give it a try and if it didn't pan out, could come back to the PhD program, which I still need to finish. I'm technically on leave of absence. So we gave it a go. AUDIENCE: I was wondering, at the end of John Battelle's book on The Search, he starts discussing the project at IBM which is systematically creating a tagged or semantic index of the web and kind of approaches a very slow but nearly perfect search experience. And I was wondering how do you feel about the future of tagging and semantics in the role of search in the future. SERGEY BRIN: Great question. Actually I have not finished-- I may have seen the book but I haven't-- you guys had it as reading, I guess? Well you can tell me about it. But I spoke to him at length about it so hopefully some of that was in there. I think that the tagging and semantics are great as long as the computers are doing the tagging and semantics. Because if people are doing the tagging and semantics for the computers then there's something a little bit inverted about the relationship with man and machine there. So I'm a big believer of creating lots of innovative algorithms that can extract this kind of structure knowledge from lots of the text that's out there and created by people all the time. But I'm not a big believer that you're just going to have lots of people that enter the data very carefully so machines can then process it. AUDIENCE: I was wondering if you can give a brief overview of your China strategy. Also from the standpoint of technical barriers as well as business and governmental barriers that you foresee? SERGEY BRIN: That's a good question. Yeah in China there are some technical challenges. In particular the CJK languages as we call them: Chinese, Japanese, Korean. They're double byte. They don't have sort of the same idea of words and sentences and what not. And anyway, there's some special technology that goes into making the Chinese search work well. But we've been developing that pretty well and I'm pretty comfortable with that. There are a number of complicated government relations issues. And anyway there are a lot of trade issues. There are a lot of policy questions that haven't really played out I think as of yet. And I think that all I can know is that I don't think we can predict today exactly what will happen with China in the future. The approach that we have taken is to make sure that we stay in touch with both of the representatives of various people knowledgeable about China both in China and out, as well as human rights organizations which care about some of the censorship questions and things like that. And what we've deduced out of those conversations with both is that our participation in the Chinese market is probably positive, not just for us as a business to participate in China but also for the people having access to more information, more kinds of communication services, and things like that. And in some cases, just like we do in Germany and the U.S., we comply with laws which we may not necessarily support. But we do need to be law abiding. So I hope that answers the questions. AUDIENCE: Both the background from academia as well as your current opportunity having access to resources, infrastructure, and data. I'm wondering your views in terms of doing research, coming up with great ideas in the corporate environment versus say in the university. And what opportunities exist today you think in university research within the context of search or marginally in terms of distributive systems design? And what might some things Google could do to perhaps enhance-- make innovation happen in both big corporate environments as well as academic research? SERGEY BRIN: Great question. I've had some experience now with both academic research as well as within our company. I think that you can-- we have a lot of useful assets and I guess technologies, things like that, a Google. And I gave the-- I illustrated with one example. You can look at other things. Google maps takes a lot of computers and some data that we license. Things like that. But that said, there are plenty of interesting things were you don't need any of those things that we have. Like you might-- especially with today's computers, you could easily get by with just your laptop perhaps. I've certainly done experiments even while with Google where I didn't use any more computer power than that. Or maybe we'll use a kind of small cluster of machines. You probably have things like that around here. I think there are many interesting things that can be at the very least tested out without the particular kinds of resources that we have. There's some things that can't. Which is why it's a good idea for us to focus on those problems. And in some cases we've actually collaborated with researchers in different places. For example, we were just talking to the Venter Institute-- I don't know if you're familiar, they do DNA sequencing-- about using some of our compute farms to run their processing. And we've run experiments with that. And other kinds of things. There is always a cost, both to us as well as to the researcher of that infraction. There's communications, there's overhead. It does take some effort. So we try to apply it in the areas which appear the most promising at the time. At the same time, we also want to make public a lot of-- we actually publish a fair amount about our systems designs, the algorithms that we use. We also have published some software tools and some of our libraries and things like that. It's still a fairly small fraction. But that's the direction we found to be helpful. AUDIENCE: [INAUDIBLE] SERGEY BRIN: We can definitely hear you. AUDIENCE: [INAUDIBLE] SERGEY BRIN: I don't know. I guess I feel that human language is pretty well evolved. And it's certainly not perfect but attempts create better ones like Esperanto haven't been that successful. So I think that professional researchers and authors-- I'd rather they spend more of their time researching and writing. And I think it's our task as well as all of your tasks to turn those words into semantics which could be used for a variety of purposes. I just don't think it's good use of human effort if you kind of compound all the hours out there. I think the AI type of problems are actually very solvable. So that's my view. AUDIENCE: I have a question on the-- to follow-up the China question. So Google's mission is to organized the world's information and make it universally available. So the fact that Google is censoring, does it violate this mission in not making it universally available? SERGEY BRIN: So, excellent question. First I'll disclude the premise. We're actually are not censoring. In fact, our search is the only search engine I know of-- the only major one operating in China where we do no filtering. There are some properties such as news.google-- I forget the code--. There are some other sites-- I guess-- I don't remember the exact URL where we've had to have tailored experience. But the web search is actually not censored by us. It is, however, censored by the Chinese firewall itself where they're able to take out certain searches at the border. And there's nothing we can do. So I think we've taken a pretty good stand there. But in general with respect your question in terms of making information universally accessible and useful, it's not very useful if it's not there at all. And in fact we have had our side shut down multiple times, altogether blocked out of China. In fact had the news site shut down in China and Blogger. And things kind of go up and down. And we have to make a judgment call combined with people who really care about the political situation there and human rights there. And I've spent a lot of time talking to. We have to use our judgment to sort what the best outcome will be. We've been doing that. AUDIENCE: Baidu? Baidu, the Chinese search engine. Do you guys think it's a big competitor? SERGEY BRIN: Well, it's been a good investment so far. Yeah we actually invested in Baidu, I forgot, a couple of years ago or so. And that investment has done very well from a financial point of view. But in general, in this space we have-- there a lot of companies that compete with us in different local geographies. In China there are some. In russia actually there-- Russia and Korea have ones with much more market share than us. That's compared to China where we have pretty good market share too. I don't think we fear them. They were actually-- I mean, I know the founders of Baidu and I know the founders of most of these companies. It's a pretty good relationship but obviously we do compete. AUDIENCE: My question is also about China-- Google China strategy. So I know most of the company, they face two big challenges when entering China market. One is to find the right person. So it sounds like Google did a great job in this side. The second biggest challenge is to figure out what to change to fit the local market and what to maintain to keep consistency as a whole company. So what should Google do from the product perspective to adjust the second challenge. SERGEY BRIN: That's a great question. I think there-- well, I think many of you are well-versed and I should get your input on this question. But yes we have been hiring in China. We're very excited to have Kai-Fu Lee join us. Who, I don't know if any of you saw in the press, we had a little bit of a legal skirmish about that. But he's on board now and busily hiring all the other people who are going to help him. And most markets I find that it's actually sort of 90% is the same as the U.S. and other markets were serve in 10% has to be unique. China's probably a little more towards the 70-30 mark or something like that. And I think we're going to have to deal with a number of issues. Just for example, the connectivity issues there-- the network. We found that most university students have to pay or couldn't access Google at all because you have these-- you pay these fees for going outside the country like 10 cents a megabyte or something. I don't remember what it was. But that's a lot of money. It's even a lot of money for here. And to a Chinese university student, it was just unaffordable. And so we've had to work on our networking situation. And on top of that, there's also latency issues, things like that. And these things aren't true in-- well certainly the charging it's true no place that I'm aware of. A lot of the unique kind of licenses you need to get, things like that just aren't true in other parts of the world. So yeah we would have to address that. There are other markets that have their uniquenesses. Korea is very unusual in a lot of ways. Far more broadband adoption. And a lot of things about the end-user behavior we don't understand. So I think all these things are going to take work. I would say on average though 90-10 might be typical. Maybe based on these other countries that gets pulled down to 80-20. 80% of the work we do is the same across all countries. And I think that's been one of the reasons why we've been actually very successful internationally. Because there many markets which our competitors never paid attention to at all. And we've had a pretty scalable approach. We have interfaces in well over a hundred languages right now. And you can search in probably hundreds more. It's kind of hard to test. I don't speak most of those. But I think it makes sense to make sure that you continue to put the core efforts into technologies which will work across the board. And then don't forget that you're going to have to put some specialized technologies in each market., that you're going to have to budget somehow. SERGEY BRIN: Hi Sergey. Today is actually my birthday, but I think I speak for all of us-- SERGEY BRIN: Happy Birthday. AUDIENCE: -thanks. I think I speak for all of us in saying I think this is a great surprise present to hear you talk. SERGEY BRIN: Well, it's my pleasure to visit. AUDIENCE: My question is about the recent announcement of partnering with NASA and the implications of that in terms of branching out into different technology areas. SERGEY BRIN: Yeah, well the NASA announcement's probably a little-- actually even a small portion of kinds of things we do to collaborate with NASA and other research institutions. NASA for a long time at Ames ran one of the largest internet connection points. NASA Ames-- I'm not sure it's still such a big hub. But they pumped a lot more bandwidth than anyone else for a long time. And they also have some of the largest computers. Just sort of the kinds we use, clusters of PCs or similar kinds of things. And they've been experimenting with those for many years. Let's see, what else do they do? They've had their share of research in AI and things like that. And we've had collaboration on the Google Earth-- if any of you've tried Google Earth?-- with some of the NASA folks. In fact, Keyhole, which is the company that became Google Earth worked with NASA before too. So there's a lot that we work together in. Maybe someday we'll get exotic, we'll build space tethers and things like that. But that not in the-- the short term plan is just for some office buildings to start. AUDIENCE: Hi. My question is regarding the future of Google. I know Google started off as a search, but in the recent years they have been rolling out more desktop software like Google desktop search, like Google Earth or Picasa. Is that where the company is headed towards a more desktop environment rather than focusing on the search? SERGEY BRIN: That's a good question. Yeah, initially we didn't have desktop software because we didn't really have the expertise for writing it. And there's some cross-platform issues and we're not-- our desktop software is not on as many platforms as I'd like. It's also a little bit more work for users to get to because you actually have to download and install an application. But over time we've had things that we felt required desktop software to have a really great user experience. And you couldn't really have Google Earth in today's browsers. Maybe in tomorrow's browsers you can build Google Earth. But fortunately more and more, the browsers are becoming more capable. So we can do things like Google Maps, which is actually pretty close to Google Earth, and Gmail which because a pretty well-m featured email client just in the browser. I think we're going to continue to use both. I'm actually hoping we won't have to have as many clients because of those install issues and things like that. But we'll do whatever it takes in order to deploy the services we want. MARTI HEARST: I'm think I'm going to ask the final question. So we're very, very grateful that you went out of your way to come here. And I think one reason might be to make up for the fact that I gave lectures to you at Stanford and you're paying me back now. But probably more likely it's that we want to encourage people to go into computer science and related fields. And in fact, one reason that we're having this class-- one of my ulterior motives is to encourage some of you undeclared people to go into computer science and information science who weren't thinking about it yet. But I was wondering, what you might say as a few words to encourage young people who are thinking about this area, about why they should go into computer science and related fields? SERGEY BRIN: OK. I'm in a tight spot here. MARTI HEARST: Or why they shouldn't? SERGEY BRIN: I actually was going to discourage people because-- but I won't because I'm a visitor. I think you should-- I think it's important to do what you love. And I really loved computer science and sort of science and math in general. And I've loved the tools I've gotten the chance to use. And being part of the technology curve which has enabled some of these great applications. And then ultimately bringing these to end users and people in general, creating something that people really appreciate. So I'm very-- I feel very happy and very satisfied. But I did pursue what I loved. Now there are people out there who don't get over the hump. Like they don't really realize what they're going to enjoy because they don't know enough about it. I mean, I played with computers when I was very young. I got them as a present. My dad's a mathematician and whatnot. And I think there are actually a lot of people out there that I know that have discovered later in life, after they've actually had to go through a little bit of time learning or just as a hobby to realize that they really do enjoy working with computers. And I think you really want to make sure that you-- when you pick your career that you understand what it's about how you feel and that you're not kind of dropping thing because the 100 class wasn't exactly-- however you number them here-- that that class wasn't really representative. And I think that will happen. So the other thing they should consider as practical matter, many or most of the things they will want to do, you will be a lot better off if you understand computers and you understand programming. In fact, increasing increasingly at Stanford-- probably the same is true here-- but they built this whole Clark Center, which sit halfway between the computer science department and the med school. And has sort of these lab spaces-- you know if you guys have ever been in labs, there are lots of pipettes and tubes and things sticking all over the place. But increasingly when I visit there are just people who plop down their computers, kind of push away the pipettes and they're spending their time on their computer. They're not putting things in test tubes and whatnot. And depending on what you enjoy doing, you may or may not be happy about that. But as a practical matter you should understand it. So I do think they should get the education. And based on what you think of it, I think that you should choose your ultimate major based on that. MARTI HEARST: OK. That's great advice. I want to give you a thank you gift. SERGEY BRIN: Oh, how exciting. All right. Oh, it's a Whoopie cushion. MARTI HEARST: It's a bean bag. SERGEY BRIN: Oh, it's a bean bag. MARTI HEARST: Turn it around so they can see. SERGEY BRIN: OK. [APPLAUSE]