welcome back today we're going to talk about Larry pagee Sergey Bren and the rise of the modern search engine by which of course we mean Google as we said in the last lecture by 1998 you can see it down here there 2.4 million websites and it's going up exponentially you can no longer hand compile lists that will be in indexes or directories you need a way to search it you need to search the internet to find what you want otherwise there's no good human Network computer interface and the whole thing is like a jungle and so a few search engines that come along there but they were pretty bad they didn't really get to what you wanted to get to but as we'll see in that year when we hit 2.4 million websites that's the year that Google is born uh one of the co-founders is of course Larry Paige born and raised in Michigan his parents were academics uh his dad was a math teacher the University of Michigan and like a lot of the people that we talk about from Steve Jobs to Allan K to at a lovely Sten he liked both the Arts and Technology he was deeply into computers but he went to music camp he played saxophone he studied composition mus uh he went to this place called interlock and in Michigan which was one of the great music camps and one of the things they did at interlock and was they ranked each kid it's almost like ranking a web page right but they would rank each kid and how did they do it they crowdsourced it which meant that two people would play the same piece and there'd be a few kids sitting in front of them but they'd be facing in the other direction they didn't know which of the two kids was playing and then they would rap the kids and it would be a process of elimination until finally you moved all the way up to the top in the rankings uh when he was at the University of Michigan himself where his parents went Larry pagee loved a class he took on human computer interaction he studied how you can design interfaces like you know we saw at zerox Park and apple that are intuitive in particular he looked at a email client something called udora back then and he figured out how much longer did it take if you used a command and then something to try to do an act action or if you could just point and click he said it took 0.9 more seconds to do it with a command than a point and click and so he had it perform various tasks and he said quote I developed an intuition for how people will interact with the screen he's rejected by MIT but he gets accepted by Stanford and in some ways that's very lucky for him MIT as we've seen earlier in the course was incredibly good at teaching and at academics and research but they frowned upon people starting their own businesses and uh so it wasn't quite like Stanford which had a Dean for many years named Frederick Turman who very much encouraged people who went to Stanford people taught at Stanford to get involved in the businesses of what became Silicon Valley uh when he goes to his orientation in 1995 at Stanford Larry pagee meets Sergey Bren that's you know a bit like you know Steve Jobs meeting Steve wnc or Gordon Moore meeting Bob noise or huet meeting Packard or James Watson meeting Francis cck they clicked they became a teen uh Sergey had actually been born in Moscow his father taught math at Moscow State his mother worked for the big state oil and gas company as a researcher but they were Jewish and it sort of hampered uh their ability to get promoted much higher lots of discrimin ation and when they applied for exit Visa both of them got fired from their job it took another couple of years before they got the exit Visa that was in 1979 and young Sergey is five years old then they end up in Maryland uh where his father teaches math at the University of Maryland his mother works at the Godard space flight center and Sergey is able to Breeze through High School in three years and Breeze through the University of Maryland in three years and Just Like Larry pagee he has a good luck to be rejected by MIT so he goes to Stanford and they meet during this orientation and they totally Bond he says you know they joke about it now like both of them say I remember that the other one was really obnoxious but they said but all we did was we talked a mile a minute and we ignored everybody else we found each other very obnoxious but they have remained and are sort of tot Al bonded in the way their minds work and in the way they interact uh they studied at a new building at Stanford that was called the William Gates computer science building because Bill Gates by this point Microsoft is totally taken off and he's endowed computer buildings at Harvard Stanford University of Washington and many other places and a whole new generation becomes like Paul Allen and Bill Gates they become bonded in computer labs and that was certainly true of what were known as Larry and Sergey as if it were all one word sort of one word and you can see it here with what's called Camelback where you uh capitalize the different words but smash them together you've seen that like in alter Vista and many different uh companies you see how they do that everybody called them Goofy smart both of them because they were indeed absolute Geniuses in math and Computer Sciences but totally Goofy and Paige in particular loves this notion of human computer interaction and they have a group at Stanford that's dedicated to human computer interaction run by a wonderful Einstein haired Professor named Terry winegrad who becomes the adviser for the two of them so when Larry's looking around for a dissertation he's thinking of doing self-driving car but he decides that he instead wants to try to rank the importance of pages on the worldwide web partly because his parents were academics and if you're academics and you publish a paper the importance of that paper is often determined by how many people footnote it or cite it in their bibliography and there's all sorts of programs that look at how many times did this research paper get cited or footnoted by somebody else and that determines the importance of the paper but the problem on the worldwide web is the links as we talked about when Tim burners Lee invented the worldwide web they don't go both ways so if I'm on a web page I can look at all those hyperlinks and I know what they link out to but I do not know what links are coming into the pages so it's very hard to rank a page by seeing how many people cited it but you could you because you cannot look at the page and figure out how many people have linked into it as Larry said hypertext had a flaw it didn't have bidirectional links didn't have the two-way links that Ted Nelson who had created a system called xanadoo was a purist about hyperx that's what he had wanted but Tim berners Lee said if we make people get permission and Link both in for incoming links as well as doing outgoing links then the web won't grow exponentially as it was indeed growing so basically Larry and Sergey signs a board for the ride have to figure out a way to figure out how many things are linking to a particular page and as I said by the time they're really going at it in early 95 there's a 100,000 websites with a close to a billion possible links all growing exponentially and what you do is you use a web crawler it's like a spider that crawls around the web Larry Page had his start at Larry Page's own homepage and what it would do is would click on every one of the links on the homepage go find out where that link went and then put it into a huge database here of all the links that the WebCrawler had found and then once you have that database you can look at it all and you can look at it backwards to see not only where the links going to but how many links do you have incoming to each page and they called it backrub using the camel case I talked about because it went back and F figured out uh what links back to each page uh backrub was there not to do a search engine it was simply there to say how important how much of an impact does each web page have and they started using up all the server space in Stanford in order to do it but they had no thought of building a search engine and then it occurs to them that once you have this all this data in a database and once you're figuring out the importance of each page you can use that information to drive a search engine so that you just search for something and it looks at that whole database of links and it looks at which are the most important of them which is why the other search engines really sucked because they would tell you oh yeah this word was sound on that web page but it didn't it didn't rank them by the importance of each page and they called it page rank because of course it ranked each page but as Larry said it's also a bit of vanity he named it page page rank a little bit after himself and it becomes an iterative process because you're figuring out how many links go to the page let's say that you have a page that's a two-lane University homepage you'll look at how many links go into it and you say okay two Lane's more important than uh Georgia Tech or LSU and the number of links coming in but then you can do something that's comp Lex but iterative which is not only see how many links that are coming in but how many links are coming in from important pages and you can say a link from Justin Hall's you know dorm room web page is not nearly as an important is a link that comes from the New York Times or the Washington Post and you rank each page and then it becomes the sort of mathematics you need to do to do it it that if you figure out the import importance of each page it tells you okay links from that page become more important and you get to continually upgrade the rankings of each of the pages and as they wrote in their in their sort of paper they did for their professors and actually published it with one of their professors says the page has a high rank if the sum of the ranks of its backlinks is high this covers both the case when a page has many backlines and when a page has fely as a few in other words not very many but they have highly ranked backlinks there were many other search engines I said and they didn't do stuff like that they just if you had University on it it would just go and see just randomly what web pages use the word university and it didn't really calculate which is more important how many got the most links to it all that sort of thing and there was Altavista and mellin and as GES and hotbot and you know they just totally suck so people didn't use them much but when Larry and Sergey create page rank they decide to test it out against the other search engine and they use the word university they type it in alter Vista mellin they just randomly they give you some universities they give you some web pages that has a name University in it or the word University in it but when Larry and Sergey did it it ranked it the top hits were Harvard University Yale University Princeton and Stanford and the University of Michigan which made Larry pige happy and he realized that they were creating a real search engine not just some random list of things that contained a particular word that you put in the search barx they decide because it's just doing millions and billions and then trillions of things to name their company that they were going to found and their search engine instead of page rank which can't really keep calling it that they decided to name it Google after the number that's a huge number some of you know exactly what it is maybe one with to the 100th or something and it's spelled g o o g e l but when they searched to see if the name was available they misspelled it they spelled it g o GLE e and that name was available the other one the correct spelling wasn't so he ended up with the name Google spelled a bit in incorrectly they kept improving it and what they would do is they would see if somebody uh searched for something and then they didn't find it get in the top five or six things on the page they had to scroll down and go to the next page they'd realize well we got to fix it that's uh that search wasn't very useful they learned how to use synonyms for in other words that there's a SE synonym if you're looking for dogs for example well puppies are kind of the same thing if you're looking for something that's really hot the word boiling might be the same thing but then you had to fix it so somebody might search for something that's boiling uh and really hot and they might also uh you know search for puppies but if they type in the word hot dogs they're not looking for boiling puppies they're looking for hot dog something you can eat all these things are pretty complicated but if you have an iterative process that almost learns from itself then you're going to keep improving that search engine and eventually uh their do it is their dissertation but people in Silicon Valley have also invented as I've said this notion of venture capital so a guy named Andy beelin one of the early Venture capitalists he had also been a great entrepreneur and engineer he drives to see them in his car and he says to him uh I don't want your pitch just let me borrow your computer I want to type in and see how well it works he types in a bunch of things Bill Clinton and in other places it might get you to the Bill Clinton joke page but in this new search engine Google it got you right to the white house.gov was the first thing uh Andy betal shim says this is great here's a check he goes to his car a convertible takes it out of the glove compartment takes a checkbook and says here's a check for $100,000 I want to be your first angel investor figure out what percentage of the company will have and they said you know they evaluated the company said no no it's too cheap uh the way you're valuing the company and he got them to Value it right uh and they said thanks for the check but we don't even have a bank account yet he said go get a bank account go cash the check and go incorporate your company you now have investors they went to Burger King Larry and Sergey soon as Andy drove off to celebrate as Larry said Larry paage said we thought we should get something that tasted really good although it was really unhealthy and it was cheap and it just seemed the right combination of the way to celebrate our new funding Silicon Valley had a lot of things as I said it had Venture Capital it had Engineers it had universities and had professors who not only taught but encouraged you to form companies it also had you know a lot of hippies and people who defied Authority and Free Speech movement could all work at places like Atari and then be engineers at Google but it also had something that you didn't find much on the East Coast which was garages two-car garages just like Steve Jobs's family had a garage even though it was a small house and that's where Apple was started uh likewise Sergey and Larry have a friend named Susan Miki and in September 1998 a month after they met with Andy Bine they Incorporated the company they cashed his check and they went to Susan's garage and rented from her a garage where they could uh start doing what they ended up dubbing on a big whiteboard in the garage Google worldwide headquarters there they are in the garage there's Susan by the way Susan she now runs YouTube no good deed goes well actually it's a good deed and it paid off so is this really a Triumph of what we called human computer symbiosis you remember the original directories we talked about were a mixture of human hands and computers and so it seems like that that's a real combination of machines and humans and perhaps Google looked like if you do it search ending that it just automating the whole thing with an algorithm in spiders were they really removing human hands from the product in fact No in fact what made Google so great was that it was the ultimate melding of machine intelligence and human intelligence every day people got up and they created web pages and if they wanted to write about Bill Clinton they put a link and they'd link to the White House website if they wanted to write about hot dogs they'd do a link and it would go to some wonderful Lucky Dog uh vendor in the French Quarter or whatever it may be every day there were hundreds of thousands millions of human judgments being made each hour by what links they were putting on their website it and what Google did was Harvest human intuition Harvest this work of the human mind and connect that work of the human mind with an algorithm and a directory and servers and an ability to melt human Minds human creativity with machine processing power in other words it was what aah love lace looked for What doug inkle looked for what jcr lick lier looked for what Steve Jobs looked for which is that highest form of human computer symbiosis that reaches its peak with the worldwide web the creation of the internet and then the creation of Google so that anyone anywhere can link to anything everywhere and find it you can connect to Minds around the world by the processing power of humans and you have gotten a great interface for human computer interaction thanks