Friday, December 16, 2016

brown leather boots


>>watson: good morning everyone. welcome back,and welcome to this very special presentation. my name is mark watson, and i am the interimdean here at the woodrow wilson school. it is my pleasure to introduce the speakersfor this special presentation. during this session we will be discussingthe role of information technology in the next phase of globalization.arguably, there is no better symbol of globalization today than the complex connections of theworld-wide computer network. our experts will be discussing how it willtransform the global information landscape. i am pleased to tell you that we have withus today 2 of the best people anywhere in the world, to help us explore future trendsin the information revolution.

since our time is limited, i will be verybrief in my introduction and outline the format for the session today.first, we will hear from our experts. then we will have an extended period, i hope,for questions, answers, and for discussion. serving as our moderator and provocateur--ilove that expression--is ed felten, a professor of computer science--we all know him here of course as a professor of computer science and public affairs inthe computer science department in the woodrow wilson school,and the director of the center for information technology and policy, here at princeton.his research interests include computer security and privacy, especially relating to media,consumer products, and technology law and

policy.ed's research on topics such as web security, copyright protection, and electronic votinghave been covered extensively in the media. his web blog at is widelyread for its commentary on technology, law, and 2004, scientific american, named him to its list of 50 world-wide science and technologyleaders. along with this--this morning--is also ofcourse eric schmidt, chairman and ceo of google inc.earlier today, i will apologize but, i referred to you as the "google guy" in our breakfastmeeting. you were in the company of the "un guy" whowas here yesterday. [all laugh]

we have the "gates guy" here too.eric was recruited to google from novell, where he lead the company's strategic planning,management, and technology development as chairman and ceo.since joining google in 2001, eric has focused on building the corporate infrastructure neededto maintain google's rapid growth as a company. he earned his bachelor's degree in electricalengineering here at princeton, and then went on to earn his master's and phd in computerscience at uc berkeley. in 2006, eric was elected to the nationalacademy of engineering which recognized his work on development of strategies for theworld's most successful internet search engine company.clearly, obviously, why even say it? [all

laugh]eric was inducted as a fellow of the american academy of arts and sciences in 2007.i am going to quit speaking so we can hear some experts. [pause] [applause] >> felten: while eric was being introduced,i googled the "google guy," [all laugh] and in fact your name does come up.we know that you are, in fact, the "google guy." that is the source of authority thesedays.

i guess i would like to start by asking youto talk about how google made the transition from being a graduate student project at stanforduniversity, into being the global company that it is did that happen, and what did you learn along the way? >>schmidt: it is lot easier now because ofthe internet, because the internet is global. americans have this view that all innovationoccurs in the united states, and that it is lately exported after awhileto these undeserving and otherwise uninteresting people who don't live in america.of course, we all know that america is just a very small part of the global creative commons.the internet really is--it is remarkable as

an institution that we built with essentiallyamerican values-- at least according to the other countries--butallowing very rapid globalization. in our case, what we did is we built a setof websites that people just started using. the most common question i get about googleis how is it different everywhere else. i'm sorry to tell you it is not.people still care about britney spears in these other countries. [all laugh] it is reallyvery disturbing [all laugh]in fact the key insight of my service at google has been that people are the same everywhere.i think it would be the simplest way to run the world, to recognize that the other people,other races, other cultures,

people who don't speak the same language haveroughly the same things that they care about as you do.we know this because we can prove it. the only thing that is really different, ofcourse, is the language. the way we solved that was we got volunteertranslators. we made it very easy to take our website andliterally translate it into the language that you claim to speak.we have ended up with a lot of languages that don't actually favorite is "bork, bork, bork", which if you recall is the swedish chef language. [alllaugh] you can establish the specific language andall of our answers are in that subset of swedish,

as best i can tell.part of the secret here is that you get the globalization of the internet, and part ofthe secret is to get volunteers. i think one of the things that people don'tappreciate about the global stage is people have a lot of free time.they do, they are not like us. [all laugh] it is very important to recognize that theyhave a lot of free time that could be used for good or evil, but they will use might as well give them some task like translating your website. >>felten: people will translate websites foryouã¢â‚¬â¦ >>schmidt: for free.

>>felten: for themselves and their own friends? >>schmidt: the next thing you do, since thereare 5 such groups, you get them to compete to choose the one that they claim is the best.since you can't read the language anyway [all laugh]once you sort of map the art of manipulating this free energy, it works and it works reallywell. >>felten: one of the remarkable things aboutthe internet, of course, is that as soon as you put anything there it is set up a webpage yourself. you put maybe an embarrassing picture of yourfriend on it, and anyone anywhere in the world can see it right away.when that happens you are taking whatever

it is you are putting out there, and you areexporting it into many different countries, many cultures, many different legal regimes.if it just you and your friends, then maybe no one some point you get larger and somebody somewhere notices that you are doing somethingthat they don't like. that is when you get to see another part ofthe global nature of the web. google has certainly run across this, wheresomebody somewhere speaks up and says we don't like what is happening.this is contrary to our culture. this is contrary to our public policy.maybe they are confused, maybe not, but you do run into these problems all the time.

>>schmidt: they are usually not confused.they have an agenda. sometimes we agree with it, sometimes we don't.i would define google in the naãƒâ¯ve phase, and then the okay we are going to deal withit phase. the naãƒâ¯ve phase goes like this: we arehaving a good time; we have all these loud people and all of that.we put the stuff out, everybody translates everybody, and it is going to be this kindof kumbaya kind of moment around the world. the laws, the structure, and the culture areneither the same nor coherent; a simple rule. typical example: on the websitethere is information that is legal in the united states and illegal in another country.what do you do?

i now have hundreds of workers who can answerthis question precisely. it goes something like this: we assert thatif it is on the website then it is essentially controlled by google in theunited states and controlled by american law. we also, however, publish domains. they arelike .china, .brazil, .germany and so fourth. we do in fact follow those local laws.the internet community has managed to convince the governments, at least in most cases, toaccept that this is the answer. the first example we had was very early onthis. it turns out because of the despicable historyof the nazis in germany; it is illegal to have nazi websites.there are about 100 of these groups that are

truly horrific. they are legal in the unitedstates and illegal in germany. we have a german domain which omits that information.thatis widely accepted as legal, just, and correct. the next one we had was much tougher to dealwith, and it had to do with brazil. it turns out that we had a social networkingsite in brazil that had some child pornography on it.unfortunately child pornographers, again a truly despicable activity, upload this stuff,you catch them, and they upload again, so forth and so on.the legal question was whether that activity which was in a service that is hosted in theunited states was within the arms of the brazilian law.we were busy getting rid of the stuff, but

a prosecutor in brazil asserted supremacy.okay no problem; we have our government [inaudible]. then you would think it was resolved.then what happens is that the prosecutor decides to arrest the people that we had in our brazilianoperation, and put them in jail for a site that they have no control over,is not in their town, they have nothing to do with it.that was a wake-up call. our head of the country--and these peoplehave enormous personal liability. they have criminal liability, civil liability.we obviously indemnify them and all of this. they obviously had nothing to do with was something that was going on in the us.this goes on for awhile. that was one was,

by the way, successfully resolved after afew years. we learned a hard lesson, which is that inthe way governments work--including democratically elected governments--prosecutors have very strong and arbitrary powers, and many of them don't fully understandthis complex architecture of how the internet works. >>felten: this is one of the implicationsof doing business with people at a distance. even though you have an office in brazil,you don't have people in every area, every town. >>schmidt: in brazil we actually had peoplewho they could arrest. in thailand--we bought

youtube, and i thought youtube was prettyneat. then what happens is it turns out it is illegalin thailand to show a video--i have learned this very precisely--that involved putting shoes on the top of the head of the king in a character.[all laugh] this is not a video i would have picked astruly horrific, but you know. it is absolutely illegal. so the person wehad in thailand fled because of the secret police.they blocked us for a year. it has since been alleged that perhaps theblocking had more to do with the military junta's displeasure with all of those--it is legal to criticize the military junta

in thailand--but we don't know. >>felten: this is one of the interesting thingsabout youtube, of course, that anyone can upload any has been a tremendous engine of free speech, of speech of all kinds. >>schmidt: according to many people [inaudible]very bad. it is really bad according to these people, to let random people upload things[inaudible] [discussion about sound problems] we can hear each other. >>felten: that is right; we are having quitean interesting conversation here.

let's see if this works better. [pause for sound problems] >>schmidt: hand held devices. you'll be ourjudge. you have perfect hearing, and you have thecenter seat. >>felten: we were talking about youtube andthe way that it allows anyone to say almost anything, anyone who has a video camera. >>schmidt: they can say anything, not almostanything, and they do. [all laugh] >>felten: i stand corrected. yes they it meets our vital needs to see skateboarding dogs, as well as some maybe more importantneeds.

>>schmidt: the most alarming statistic aboutyoutube is that there 15 hours of youtube footage uploaded every minute to the site.i don't know the international percentage, but it is on the order of ã‚â¾ international.going back to the principal that americans are not the center of the universe, trustme people outside the us like youtube too. >>felten: to the extent that there are complicatedrules and laws about what people can and can't say in different is a real challenge for you to deal with. even looking at all of it before it gets posted,even having an employee look at it would really be completely unfeasible. >>schmidt: one of the central issues thatthe internet has is that there are a number

of organizations that would like prior censorshipto publication. i want to come back to that as a serious point.when we bought youtube, i did not realize that pictures were much more powerful thantext at the level that they are. we have tolerated on the web really nastystuff being published by truly despicable is tolerated because of the principles of free speech and so forth.a video image of the same thing is seen by humans all around world in a much worse light.all of a sudden youtube brought these issues much more forward.right after we bought youtube, we got sued by viacom for a billion dollars.without commenting on the merits of the suit,

except that they were wrong. [all laugh]one of their arguments, fundamentally, is that when people upload things there has tobe a check before they get published. we fundamentally disagree with this at a verybasic level. it is a principle of the web. an analogous point has to do with the questionof prohibited or inappropriate speech, libelous speech, or worse like pornography and thosesorts of thing. many governments would like prior censorshipbefore publication. a typical example here would be turkey.we all think of turkey as a pretty modern state, but i will tell you from a youtubeperspective they are a seriously backward state.for the last year they have blocked us.

the reason, allegedly, is because of a gentlemanwho published something which was critical of ataturk, who as you know is the founderof modern turkey. it is perfectly possible that the motivationfor this censorship has more to do with the political dynamic and criticism of the existinggovernment. we don't really know. >>felten: let's change the subject to anotherexample of information being posted online. that is google street view. >>schmidt: oh, my favorite. >>felten: street view is a facility that letsyou, if you are looking at a map, get a picture

of what it looks like to be standing on thestreet at that place. if you want to see the front of the buildingthat you are going to be going to later in the day, you can use google street view andjust call it up. the way that this was done was essentiallyto send cars driving around the city streets snapping pictures over and over.then processing and stitching all that together using computers.the result was essentially a very large set of candid shots of things happening on thestreets of america and elsewhere. when google street view first went up in theunited states, there was this sort of game that was launched where people would lookfor [inaudible] pictures.

look, here is a picture of somebody comingout of a pornographic movie theater. oh look, that looks like someone breakinginto that house. of course, if you take a zillion picturesyou are going to find all kinds of interesting things.this kicked off a kind of debate about the implications of street view,which on the one hand is a marvelously useful feature to be able to use if you want to navigatein a city or town that is covered, but on the other hand, involved the publicationof all kinds of pictures of all kinds of interesting things.what happened after street view was launched? >>schmidt: the most important thing that wedid is we brought in the face anonymizer.

it turns out that people are not excited ifthey are caught without their face being visible. remember these are public streets and so fourth.we won a lawsuit by a family who had a funny name, which i won't repeat.they actually sued us over driving through their public street and taking a picture oftheir house. we won that because it is a public street.we try very hard to do this. we have an internal video of a street viewcar--these are little prius by the way--you know it is google--they have little cameras on the top and this particular one goes into a--the guy takesa wrong turn into essentially a water treatment plant in chicago.people go berserk and he leaves the street

view camera running for the 8 hours that heis detained, inspected, hauled in by the police, and hauled in by the fbi, so forth and soon. they don't believe that all he is doing isdriving his little prius around with the little camera.last week, one of our cars was driving through a small village in england.there was a barrier set up by the local population, who blocked the car.the very friendly street view driver just sat there for awhile.they gave a speech about how they didn't want modernity to arrive in their town.we are on the edge of this question. the reason we take such a strong positionis that street view is our most popular, in

terms of growth right now, property.people love using street view to figure out where they are going and what they are doing.there is clearly value being provided for the mass. >>felten: on the whole, services like streetview and indeed the basic technology trends behind them make it much cheaper and easierto gather and store a lot more information that people are interested in.most of what people are interested in is information about what is happening out in the world andinformation that involves other people. there is an inherent pressure on privacy thathappens because of this technology. street view being an example, not that youset out to doã¢â‚¬â¦

>>schmidt: what is interesting about streetview is street view is precisely not about violating people's privacy.if anyone complains about anything in street view, we do our very best to take it downas long as it is not a picture of the eiffel tower. >>felten: but i guess my point was, that intrying to build this feature, you inevitably ran across privacy issues,and you had to think very carefully before the product went out and after about how tohandle this privacy issue. i think companies in the technology fieldare more and more spending time thinking through privacy implications of what they are doing,in a way that companies and a lot of other

businesses don't have to do as pervasively. >> schmidt: we do, and we are, and that'sexactly right. i think privacy is an evergreen issue for the are going to be having conferences on privacy for the next 50 years.people are obsessed about what other people are doing, and whether what they are doingis correct and appropriate. they are also obsessed about their own image,public and private. that's true for 7 billion people, becausepeople are the same everywhere. i think the best current example of this isa product called latitude, which we released a couple weeks ago.i'll tell you the story of how this came out.

i'm sitting in the usual product review room.we have all the executives, larry, sergey, myself, and the other senior executives doing--wedo product reviews for days on end--we're basically a product company.then this 22-year-old came in to demonstrate his new accomplishment.he shows a map, and he shows his current position on his mobile phone, and he shows the currentposition of his friends. and he has figured a way to calculate--notjust the real-time position of where he and his friends are,which of course he's logged on our servers, but also, he can predict how to meet up withthem. as he does his demo, i'm slowly sinking inthe chair. because, it's obvious that he has

no clue what he has just invented.and, of course, everyone else in the room says: oh that's so neat!and i'm saying: you have real-time tracking, which you're storing in our log files anywherein the world with an accurate predictor of where the person is going.everybody asks: what's your problem there? i said: can you start with the lawsuits? so think about if it were true, that we hadsuch real-time log information--ignoring the privacy issues.just imagine the number of police and legitimate business and government organizations--andperhaps some illegitimate ones as well-- wandering through, giving us subpoenas, orgoing into other countries and jailing our

people and so forth to get all this information.because, it is incredibly useful if you are trying to "catch terrorists," or "catch criminals,"or "persecute people you don't like." whatever it is your government does. there is a spectrum, as we know.eventually i won that argument. so it was agreed to do an internal test--whichi also then blocked. because i said: ok, so now we're going tohave lawsuits inside of our company over employee behavior, and so, eventually, the 22-year-old is wondering: what have i got myself into. you don't appreciatethe brilliance of my invention. we made a change, which is very easy to understand.with latitudes, you can turn it off.

it's very important to have an off button.also, you can tell it to tell people that you're somewhere else. [laughter] right? qednow from his perspective, that would never be we're sitting in the board meeting last week, and one of the board members says: where'sjonathan? and i said: he's at the wailea in hawaii.and we're obviously not in hawaii. and the board member said: how do you knowthis? and i said: real-time tracking--there he is!

the next day, we got a note from a nice lady,who had been mugged in san francisco. who reported on the great use of latitudewhen you leave it on in your purse, and tell the police where you, these things go both ways. and that's a good example of how we foughtthrough the invention by the 22-year-old. and then it's ultimate application into somethingwhich is also incredibly successful. >>felten: i think this is a good illustrationof what often happens in discussions about these products, and sometimes in the publicpolicy discussions as well. we now have access to this wealth of informationthat we didn't have before. it might not be so difficult technologically,to build a product like this once you get

the idea.the hard problem turns out to be figuring out what you actually want to do, how to housebreakit--if you will. >>schmidt: i should say that the ones thatare actually the most serious of all are called search logs.when google was founded, search google maintained search logs was obvious to me that there would have to be some restriction on that, because theselogs can be used against you. in the united states, we have something calledthe patriot act. there is also a successor called patriot 2,which can be used by a secret court. it is all legal in the united states.we do occasionally at least anticipate getting

such subpoenas through normal course of other countries, of course, it's far worse. the question was: how do you balance the interestsof legitimate governmental interests--there is obviously a legitimate interest--and themost extreme privacy concerns. we ultimately said: there must be a an operator, we are going to be regulated in this turns out the european data privacy commission and the european commission simply settledon a demand to us for 18 months. so we just said: yes. so that's an example. [both talking]

we actually don't delete it. we anonymizeit. and what i'm saying here is completely publicallyknown, we have described the details and so fourth. it has taken a lot of steam out ofthe system. again, i offer that as another example wherethe technology people think that because disks are plentiful and it's so easy to log everything,we might as well keep everything. you never know when that stuff might actuallybe useful. for example, we can use historical logs toimprove our algorithms. so there is a legitimate business reason toretain those logs, at least from a technological perspective.logs retained forever create all sorts of

long-term issues. >>felten: to what extent do these privacyissues limit your ability to get new business? if i were to put on my provocateur hat, imight say: i'm a little scared of all the information that google has about see all my email, things that i look up and read, you store my calendar, and all kindsof other things about what i have been doing. i may be personally not so worried, but ithink that if google fell into evil hands i would be pretty concerned about what happenswith all that data. >>schmidt: we have a rule of: don't be evil. >>felten: oh, ok. never mind then.

>>schmidt: we're computer scientists [inaudible]by rules. you're a computer scientist. you understand.the serious answer, i think, is that the entire relationship that google has is based on thetrust of our end users. if we were to do something that would violateyour trust--you as an individual--a reporter would find you,and they would make such an enormous storm that we would lose three-quarters of our userswithin a week or two to all of our competitors. because we've set such a high standard forourselves, we take it so seriously, that we've had a couple of cases where government hasgiven us subpoenas which we've actually fought. thank heavens for the judiciary in our country.we actually fought and won.

you really want to fight for the right ofprivacy within the context of what i'm describing. if you're not willing to fight for it, youshouldn't be in this business. to be just incredibly blunt with people whomay be confused about this, whether you're using google or not, your email system issubpoenable. in fact, the majority of the activity in policesections, now, appears to be going through people's email.because criminals use email, it's obviously terrible to be a criminal that uses should not assume that you have absolute privacy unless you're willing to use essentiallya chalkboard inside a room with no windows. it is just not fundamentally there. the systemsdo tend to retain information.

the solution is: just don't commit a crime. that's the easiest way to solve that problem. >>felten: in fact, even if you are in a roomwith a chalkboard--even if you are, say, sitting out here in the audience with the camera notpointing at you-- you are in a different world, from an informationprivacy standpoint, than you were a few years turns out that people like me have cameras in their pockets, and all kinds of thingsget recorded. recorded audio; recorded video; recorded andsnapped as pictures and then republished--which wouldn't have been in the we are living more in a fishbowl, even

if we are low-tech people now. >>schmidt: i think, technologically--i assumeeveryone here has a cell phone on them. everyone i know has a cell phone, or a mobilephone, an iphone, an android phone, a blackberry--what have you.if you think about that device, it has a 3 megapixels or greater has a cpu in it, which is more powerful than the cpu that i used when i was here asa student--which is sort of amazing. it has more memory as has the ability to do pictures, and in many cases, can now do a video.there are more photographs taken in mobile phones that there are in cameras.soon there will be more videos taken by your

handheld devices, and so has a gps. even if it doesn't have a gps, we have software which can simulate gps usingcell tower information--which is really neat. oh, and by the way--i forgot--it's also usefulfor phone use, this is its least interesting use.we can talk about browsers, and so forth. that change is a truly revolutionary architecturalchange. it means that people are carrying around thesephenomenal computing resources with them, which are hyper-connectedin my case, i like history. so when i walk down the street in new yorkor here in princeton, it should be possible for it to tell me the history of each of thebuildings as i walk by.

right? why not?google has the information. the phone has the's not doing anything, right? it is just sitting there.occasionally i'll say: oh what's going on at charter club?boom! there is the history of charter club, or whathave you. when was cannon club converted to a universityfacility, or something? my point is that people live in very localcontexts. now you have encyclopedia britannica streamingto you every day. i'm using the history example; because that'sthe mildest example i can think of.

think of the combination of the phone andface recognition. where i bring up--ed, you were very carefulto hold up your phone. using modern technology, you might have scannedthe audience, done immediate face recognition, and it would have told you every person inthe audience even though we don't have their name tags.that's when it gets scary. >>felten: very useful.the killer product is the one which will sit on your lapel in a cocktail party and tellyou who that person is. >>schmidt: you know, i hadn't thought aboutbuilding that one; but now that you suggest it.

>>felten: on the subject of mobile phones,another big implication of mobile phones is in the developing world,where you have a lot of people who don't have access to the kind of computers which we mighthave on our desks. you might not have the wired internet infrastructure.where a lot of people have these mobile devices, which are ever more capable, which will becometheir portal to the information highway. you see companies like google, who are movingaggressively to get their services onto mobile phones.maybe you could tell us something about what you see happening in the developing world,via the wireless infrastructure. generally, about technology taking off inthe developing world.

>>schmidt: one of the things which i'm mostproud of is something which we have almost no day-to-day concept ofwhich is the arrival of another billion people into the modern communications network overthe next few years, and the deployment of these very low cost wireless networks.for many reasons, the basic connectivity--the basic sms messaging, which we've taken forgranted here-- has in the last three or four years becomeavailable to people who don't have televisions, and have never made a phone call.through a combination of micro-lending and the other kinds of things people here knowabout. the fact of the matter is people like to's true even there.

this means, this platform, which today islargely sms--or short message based--is just ripe for the next generation of what are calledfeature phones. they are now becoming available at price pointsin the $30 to $50 to $70 range. this is low enough that they can be subsidizedinto those markets. all of the sudden, they can get basic connectivityand basic things. you sit there and you say: oh they don't care,they are just poor. they are just living in their little huts.this is a typical sort of an american reaction. if you are a farmer, and you're wealth isdetermined completely by agricultural prices, and the current temperature and weather.that sms message determines whether your family

lives or dies, or whether you have to sellthe cow in order to get through a bad winter. we think of this stuff as essential, for themit is fundamental to get that kind of information. there are many examples in markets where peoplewill be on a ship near a town, and they used to get ripped off because they would go tothis dock and get a non [inaudible] price for their goods.and now they can text to get their optimal price.they have all been able to globalize their information.there was a food shortage about a year and a half of the alleged reasons was that all of the sudden the arrival of this informationhad caused people to discover that they wanted

to dis-intermediate their suppliers,because their suppliers were ripping them off. >>felten: certainly everyone can find someway to benefit from increased availability of information.the people who have less information available to them already are the ones who, perhaps,gain the most from being connected. >>schmidt: we complain about the recessionhere. let me tell you that we have got it good comparedto people in the third world whose export markets just collapsed.where there the minerals boom, which so fueled their economic growth over the last ten years,has also just collapsed.

again, this is not a matter of humor, althoughit is fun to talk about. the fact of the matter is that these toolsare now essential for them to recover any form of economic growth. >>felten: so as people in the developing worldbecome increasingly connected to the global information networks,how do you think this will change the way we in the developed world look at the developingworld? >>schmidt: one of the other principles is that there is a lot of injustice in the world,and we are blind to it. you see this now with these horrific youtubevideos of women being beaten for violating

the alleged laws of the country.when will humanity say enough is enough? that kind of indignity is not ok.the fact of the matter is that the ubiquity of recording devices and the ability now topublish them, really does serve as a check and balance--in my view--on despots.the analogy is, in the 1970s, ted turner was the first person to cleverly use the adventof commercial satellites to create a global television network.this then allowed the coverage of horrific state'll remember. you know them all in the 1970s and 1980s.that served a check and balance against the grossest and most horrific state we have the ability, using private communications,

to serve as a check and balance against, atleast, the societal sanctioned horrific local actions.what i worry about is that we are still locked in this old zeitgeist--cuba being the currentexample. it is illegal for us to have any businessconnections, whatsoever, with cuba. and yet, if there is any group which couldbenefit from fax machines, personal computers, and bringing them into the modern world, itwould be the citizens of cuba. they would then immediately discover thatthe people running their country should be seems obvious. if you were a dictator--i hope none of usare, and i certainly am not going to be--

the first thing that you would do in a countryis you would take your tanks and you would encircle the boundaries of the would shut off all communication. you would make sure no one could see whatyou were doing. the internet is organized to prevent that.the internet is precisely organized to not allow those kinds of activities.people who say: we shouldn't do this, or we shouldn't do that--the way you invade thesecountries, is with information. that information, from their perspective isan invasion. from our perspective, it allows a levelingof expectations and of humanity. >>felten: this, i think, is a good point toopen it up for questions.

>>male audience member: my name is [inaudible].i am an alumni and an android developer. i was trying to take signs about what is goingto happen to android and my business throughout the stock.not having heard a word, a single time, i am getting more disappointed.the question i have is. >>schmidt: what are you disappointed about? >>male audience member: to begin with, i havebeen watching a work phone you have in your right pocket. it doesn't look like an androidphone. my question is: you have an absolutely wonderfultechnology. a great toy, a very flexible one, that can simply level and devastate the mobilemarket as we know it.

do you plan to dedicate more resources, morefunding, and better management to the program you have right now? >>schmidt: nice to meet you. [all laugh]. for the background, for the people in theaudience, android is an open-system operating system that we developed for mobile is by far the best linux based platform for mobile phones.we did a partnership with a company called t-mobile which has done very the next few weeks there will be a whole bunch more partnerships being part of that we have a developer program--it

sounds like you are part of that--where applicationscan be built on top of it. this will be the year when a very large categoryof phones--small phones, large phones, touch phones, non-touch phones will be availableon android. what is interesting to us about android isthat most of the phones that are delivered are ones we don't even know about.that is the promise of open system. so the answer to your question is, yes weare doing a significant investment increase in engineering and support.between now and september, which is this year, you should see a lot more partnerships. >>felten: there is an interesting issue here,i guess, which we might want to draw out for

those who are maybe less familiar with thetechnology development landscape. that is the role of open source technologiesand the way that when even a large technology company is rolling out a product,they rely on building a community of people to turn a product into a really useful ecosystemthat people can use. android being an example of didn't set out to build a packaged mobile phone with a fixed set of services.instead to enable people--like this gentleman--to help make a more interesting environment happen.when you do that, you are not 100% in control of what happens with your product. >>schmidt: for those of you who don't knowthe details, in the open systems world what

happens is you release the software with alicense that generally allows people to do whatever they want.sometimes they change it in ways you don't like, but in many cases they follow your lead.that is the balance that we are trying to has worked incredibly successfully for the web.those of you who are familiar with web architecture, the predominant web server architecture isbased on an open system platform. all the licensing came out of that, so thisis a very web centric approach. >>male audience member: you mentioned howyoutube has become very powerful. that you can sort of didn't really check how we wentback to the day of [inaudible] university

text. you didn't quite see how powerful itwas. you sort of suggested that it is so great for free speech and whatnot. while that iscertainly true, [inaudible]. they were required to say less interesting things on the campaigntrail, and fear that [inaudible] video format. [inaudible] it cuts both ways, the ubiquity of this technology.when it comes to car dealers, who may not want to public press or push the envelopehere or anywhere else, we might end up just [inaudible]. >>felten: let me repeat the question for thoseof you who couldn't hear. the question was about the impact of youtube, especially onpolitics.

the concern that the fact that things caneasily be videotaped and put up for the whole world to see, might make speakers--especiallypolitical speakers--very cautious. therefore, lowering the quality of politicaldiscourse. >>schmidt: if caution means they stop lyingwhen they say one thing in one place and one in the other thenã¢â‚¬â¦ [all laugh, applause] i want to distinguish between private behaviorand public behavior. i think everybody here understands the valueof privacy and private behavior. we are not talking about that.for years i had thought and had said publically

that there would be an election that wouldbe determined by the internet. it turns out that it was 2006. if you go backto then--we have such short memories around here.--the house, primarily due to the war issues, became democratic.the senate was not going to become democratic. there was a seat in virginia that was in play,and it was close. the incumbent was george allen, and a younggentleman caught him saying a racial slur word apparently by intent, not randomly.then that was used in virginia--which has a history of concerns over such things--andit on the margin switched to a conservative democrat.this then switched the senate. switching the

senate changed the political dynamic for presidentbush in a very fundamental way in our system. small things like that can have a huge politicalrepercussion. when nancy pelosi then took over the house,she tried to have her closest collaborator become, i think, the whip or something likethis. this is a person who had been in the abscam;it was a democrat who had been in the abscam scandal.a video surfaced of his testimony, on the internet.this became so concerning to everyone that he lost his appointment.if you are a public figure the reality is--and they have now all learned this as you said--you have to be careful not to play one side

or the need to be careful not to commit and crimes and stuff like that.i think it is reasonable for us as citizens to expect our politicians to say coherentthings in the same places, and also not to pander.there is a history in america of politicians pandering to one group, and then saying somethingdifferent. now we can check that. i am very optimisticon this from a governance perspective. >>felten: we see a similar issue, actually,for many of our students with regard to facebook. the students all have facebook pages, andthey all put things on those pages that they are willing to show to people now.twenty years from now thy may regret some

of the thingsã¢â‚¬â¦ >>schmidt: i have a solution. i think thereshould be a law that you can change your name at 21. it wasn't me, i don't look like that anymore,different person, different name. >>felten: i will be back to you next weekwith the algorithm for assigning, for connecting those names of the interesting things--we see this with facebook.when i talk to students about this, sometimes in a: are you sure you want that on the netconversation and sometimes just generally, what they say is, look you old guys don'tunderstand.

the norms will be different when we are 40and 50 and when i, or someone like me, is appointed to some government position.the norms will be different and people will be forgiven, not only the knowledge that therewere youthful indiscretions which we have dealt with for a long time in public life,but the vivid video evidence of them >>schmidt: if that is really true, then whyis the vetting process this year--which is presumably the most liberal of all the yearsin the last 50 years--the roughest? >>felten: well, that politics. >>schmidt: so it is also possible that theinverse will be true. that 20 or 30 years from now it will be evenmore vicious because people will have an even

worse summary of what people did because noone is perfect. >>felten: it could be. these days memoriesof what happened back in the 1960s and 1970s are maybe a little vague. >>schmidt: mine certainly are, and i was here. >>female audience member: at this point googledoes not charge for all of these wonderful services. you are adding a lot more, the picasaetc. do you foresee in the near future or in thelong-term future that this will change, or will ad revenue continue toã¢â‚¬â¦ >>schmidt: [inaudible]

i will give you the shorter answer, whichis no. >>felten: up near the back, the woman in thered shirt. >>female audience member: i am interestedin this tool, the use of the phone thingamabob as a further tool for research in medicine.i told pat about 10 years ago it would be great to have a hand-held device with evena sensor on it. you would be able to hold it and click atsomething, sort of like your history walk down the street in know what this organ is that you are looking at, this plant, or some sort of would be kind of hyper searchã¢â‚¬â¦ >>schmidt: there is actually a group doingthat. it is called the barcode of life.

what they are trying to do is to come up withessentially a upc code for the genetic structure of every living organism.they have developed a sensor which can actually do precisely what you described.if you use google, you will find all the details. >>felten: i should look to the left, and callon someone over here. here in the blue. >>male questioner: google's world-wide marketshare and search has grown monthly, based on the stats i have seen, about 1% a monthover recent years. my question is do you see a natural limit to that. >>schmidt: yes, percentages have limits. let him ask this question. go ahead.

>>male audience member: [inaudible] it hasto do with: do you see any concerns or issues about monopoly in the future? >>felten: the question is--asking eric: doeshe see any concerns or issues about monopoly in the future? >>schmidt: we don't necessarily confirm thatnumber, so that is somebody else's opinion on our market share.we have high market share in some countries. we have low market share in others.we argue very strongly that we are one click away from losing, it is possible to have high market shares but still be highly competitive.that is both my personal opinion, and our

legal position.the simplest answer is: we try very hard to not lock people in.we looked at microsoft and other historical examples, and wrote an internal policy abouthow to be good while being big--how to avoid being evil, essentially.the simplest rule is: we won't trap your data. so if you become dissatisfied with us, youalways have the opportunity of leaving and going to a competitor. >>felten: having been involved in some previoustechnology antitrust issues, let me comment on this as well.i think you have a problem when you have not only a high market share, but also you havesome kind of barrier to entry--or lock in--

which makes it hard for customers to go somewhereelse, [inaudible] problem. then also, when you have some bad can imagine situations developing in the technology field going forward, where thathappens. but i think you have to make a detailed argumentfor these other elements as well before google were to get into antitrust trouble. >>female audience member: hi, i'm an undergradin the economics department. i was just wondering what you thought abouthow people claim that as more and more people spend so much time on the internet, and lessface time, they become even more socially disconnectedthan they have in the past.

what does google thinks about it? >>felten: the question is about whether peoplespending more time online and have less face time become disconnected from each other. >>schmidt: you probably don't remember thebook "bowling alone." the book "bowling alone" is clearly false.we are more communicative. we are more together. we are more social.we are just social in these odd ways. from my perspective, the current text-basedcommunication mechanisms--which is what you were referring to--are just a transitional communication mechanism to much more interpersonal communication withpictures, movies, real-time video conferencing,

that kind of thing.the 160 character limit of sms is just a technological artifact--something which will be gone verysoon. from my perspective, the simplest answer is:people are spending more time communicating in different ways.i think we tend to talk in the negative context. think of it as: there is an explosion of communication.there is not just an increase. you do that when you use facebook informationor you use twitter. i especially think about it for somebody whohas not grown up in our rich part of the world. >>felten: one of the changes we've seen inthe way people communicate--these technologies--is a movement from synchronous to asynchronouscommunications.

synchronous means we are here together. weare talking at the same time. i speak, and you hear me at the same time.we go back and forth in real-time. asynchronous means i write something, i sendsomething, i put something on my facebook page, and you come and look at it we do see--as eric said--a tremendous amount of communication going on.but we are seeing a change in the mode of communication.that does change social interactions in some sense.but it is still true that the things that people want to do most of all is talk to otherpeople and communicate with other people. that is what they are primarily using thesetechnologies for.

>>schmidt: look at the rise of twitter, forexample. which is: what am i doing right now? >>male audience member: there is concern inthe newspaper industry is that organizations, such as google, repackage news and provideit free. thus the newspaper doesn't have a source ofincome. it goes out of business. and the very collectors of news--which youare then repackaging and sending out--will can google respond to the concern that newspapers such as the boston globe may justdisappear? >>felten: the question is about what is happeningto news and newspapers especially, concerns

about the vanishing newspaper, and what googlesays and thinks about that. >>schmidt: there has been a lot of discussionof this recently. i think the recession has really brought this issue to the fore.newspapers are suffering from three different is the rising cost of newsprint. another one has been the loss of classifieds.the third has been the loss of advertising revenue primary, by the way, from autos.especially in local markets, it is very severe. we use newspaper content with their permission.the first part of our answer is always if the newspapers don't want us to--as you sayrepublish--or make available copyrighted information that they own and control,it is trivial for them to put information

that would block us from doing every case they have chosen to give us that information in return for us sendingweb traffic to their sites. so many people come to google looking forinformation; we send it to their site, which they can then monetize.the rough analysis of the problem is that the money that they are getting on their websitesdoes not make up for the loss of revenue from print circulation.that is not a problem that we know how to solve.we have ideas about how to make the online news more effective, and we are working onthose. we fundamentally don't have any good insightsabout the other problems.

>>felten: we had a conference last year aboutthe future of news in the age of technology. my take away from that was that the transformationthat is happening, the challenges to newsgathering and newspapers especially, basically boileddown to two problems. one is that newspapers now face a lot morecompetition. you don't have to read the one paper or the two papers that are in your can read whatever you want. that means that there is a lot more competition.therefore, the price is likely to go down. this means less revenue for individual papers.the other problem is that the traditional newspaper was a bundle of different things,different components. some of these were cheap and easy to makeand also profitable, say the sports section.

others were expensive and not as revenue enhancing,say investigative reporting. the sports piece of the newspaper was subsidizingthe investigative reporting part, but they were distributed a days people go to different places to get those different components.they get their sports from a lot of different places.the result is that the parts that were profitable and subsidizing the sort of hard news canno longer do that. i think there are some basic economic changesthat are happening and would happen regardless of whether google or any other particularcompany was around. down here in the front, you have been patient.

>>male audience member: first off, thank youfor the transparency that google has provided recently with the [inaudible]. the fact thatthis video will probably go on youtube, some of us do watch them.and of course your push for social changes google [inaudible].my question is text search as an aim for all extends to the tune of roughly 20 milliondollars. in the last week or so there was a reportthat came out that said that youtube is basically costing you guys 5%, 500 million dollars.i wouldn't ask you to substantiate that. my question is: when are we going to see thenext type of ad, the geo ad that people have been talking about?

>>felten: the question was about revenue andalso about when are going to see geo ads--meaning ads that are targeted based on where you are. >>schmidt: you think that those geo ads wouldbe useful? today google's--the vast majority of our revenue,96-97%, is related to the text ads that you see.if you type digital camera in, you will see ads in various combinations on the top andon the right hand side. they pay for google. when you asked earlier will it remain free?where is the money coming from? it is coming from the way it is a good business. the other businesses that we are in are afraction of that.

in some cases they are growing more quickly,but they are not as profitable, or they take longer, or they take more investment.if you are going to have a core business, this is the one to have. that gives flexibilityto have a business and yet also have a lot of these other things.we can afford to build businesses and lose money at them while we are building the case of youtube, we are in the process of building video advertising systems whichwe believe will ultimately provide enough revenue.not only to cover the costs of youtube but more importantly to pay for the professionalcontent. which is the issue--it is fundamentally theissue with the newspapers.

we want to pay for that professional content,we are codependent with them. for an ads format perspective the most importantnew set of ads formats that we are doing are called display ads.there are essentially picture ads, where the picture ads are better targeted.our observation is that an ad online that is highly personal is useful, and an ad thatis not very targeted to you is pretty much useless.the example i would offer is my television. the television is on, and it shows the sameshows that i saw yesterday. it is a dumb tv. why does it not know that i already saw thatshow? it is just a computer, right? it should beable to know that.

it also repeats ads over and over again. whydoesn't it show me ads that are relevant to me?ads for baby diapers--there is no baby in the house, there is no need for diapers, andthat kind of stuff. changing that model to a more targeted advertisingis what google is about as a business. >>felten: we have seen this in a big way onthe web. one of the really interesting, often untold, stories about google is about thead model. this advertising model that google has developedand its role in really changing the way that print advertising works, and display advertisingworks online. we are all waiting for that to come to television,to come to video.

do you have a prediction about when and howthat might happen? >>schmidt: we have been working on it. thesethings are hard. you have to invent new technologies to do television we have a partnership with a satellite company where we have set top boxand we can actually do targeted ads to your set top box.we get a signal back as to whether the ad worked or not. that is pretty promising.we had an ad business for print ads for newspapers because we tried to help them out.we couldn't get enough information back about which ones were working, to make the networkand the auction that we run work very well. it is highly dependent--the same thing withradio.

we had a radio ads business where we couldn'tfigure out well enough which radio ad to have you hear on your radio to make it economicallyworthwhile. >>felten: way in the back >>male audience member: a question about internetpower laws. you mentioned in the speech [inaudible] in washington.that you divide the total number of blogs by the total number of blog readers, and yougot a number that pretty close to one. >>schmidt: basically 100,000 blogs createdtoday. the average number of readers for that is one. it actually 1.01 because the .01 isyour mother >>male audience member: so it hasn't gonedown at all since then?

>>male audience member: i think the gentlemanearlier mentioned and the effort by google to have a moto do no evil and accomplishsocial good. i am wondering what your conclusions are [inaudible]that you been at investing through and [inaudible] fund now.the focus on renewables, i know, is a personal interest of yours. can you talk about allthat? >>felten: the question is about,which is the philanthropic arm of google. >>schmidt: we put together rightafter the [inaudible]. we funded it with about 1% of has a serious amount of money. we started with a number of initiatives.the one that has had the biggest impact has

been in the area of green and renewable power.for example, we have been funding and working very hard for plug-in hybrids.the neat thing about plug-in hybrids is basically if they have a larger battery they use almostno gas most of the time. they can use off-peak power.another thing that we did is called the google 2030 plan, it was described.we can up with an architecture, which basically if we just to decide to rebuild america'senergy infrastructure by 2030 we would essentially not be reliant on foreign oil.we would also have created a huge number of american high-paying jobs and created a wholebunch of export industries. the net present value of that was a positive200 billion dollars. we actually make money

to do this.this, of course, means it won't happen. the economics are order to do that, for example, you need to have a smart grid, intelligent garage solutions,and so forth which we are working on with a number of partners.that is an example of the kind of impact that we have been able to have.another one that we are particularly proud of is called flu trends.what we do is we statistically look at the log behavior, anonymize the data, and we canpredict it looks like a flu outbreak three or four months before the authorities willpredict it. when people get the symptoms of flu, theygo to google and they type them in.

we can detect that. we believe that this willsave some number of tens of thousands of lives and many more if there is the repetition ofone of these horrific out breaks like 1918 flu virus.those are two examples with many more coming. >>male audience member: i wonder if you wouldgive us any thought to how the internet will develop and the way it can assist us in organizingourselves politically as local citizens. >>felten: the question is about how the internetcan help people organize themselves politically as a global citizenry. >>schmidt: you have actually done some researchon this. you are probably more the expert than is the rough problem. you watch television

as somebody says oh; well we have a millionvisitors that came to our site. everybody goes oh, well that is exciting.the problem is that we can't tell whether that million visitors is 900,000 computersand 100,000 people, or whether it is a million people who were paid by some lobbyist to gotalk to the site, or whether it is something that they madeup because their it person can't add all the websites correctly. we just don't know.when you start to think about trying to govern using the internet, you start to get reallyworried about misinformation, lobbying activities, spin, and that kind of stuff.indeed, even lack of a paper trail for voting machines. this has been sort of somethingthat people are very concerned about.

if you think of the notion of one person,one vote, and the traditional principles of democracy,you want to be sure that you are preserving some aspect of that or just view it as input.if you are an expert, you can view the sensing and so forth that goes on as input to you.i would be very concerned if people started to say--here is the example that i was talkingto ed about earlier-- let's assume for purposes of argument thatthere was a discussion about whether smoking caused far as i can tell in science everyone agrees, smoking is a primary cause of cancer. theonly people that do not agree with that are the tobacco companies.we do a survey of the average person on the

internet. the average person doesn't reallyunderstand all the details. they read the scientific literature, and thenthey read the opposing view. the tobacco companies have a huge financialinterest in making sure that the opposing view appears legitimate.they will spend an asymmetric amount of information to pollute, if you will, the conversationfrom generally accepted--if you will, communal wisdom of this outcome.i am using that because that is a very simple example that we can probably all agree with.of course, the real examples are much more pernicious. >>male audience member: [inaudible]

>>schmidt: my point is that you have to comeup with systems which allow you to understand whether you are being gamed, in order to effectivelydo these things. >>felten: this is one of, i think, the mostimportant questions about how the internet will transform politics.certainly technology increases the quantity of discourse. the question is: can you useit to increase the quality, or at least to make it easier for people to understandwhat is happening, to get a better picture and to make better decisions?that is a very hard problem. we have maybe a few inklings that it may bepossible by thinking carefully about the way you design systems, about the way you designsocial software.

this sort of nudges people in the directionof having conversations and forms of organization that are more likely to be fruitful.there is a tremendous amount of research to be done in social science and technology designto understand how to make any progress on this the other hand, if we make even a small amount of progress at raising the level ofquality of a political and public discourse, it can have tremendous benefit. >>male audience member: in another talk today,the person mentioned that we should get away from the old method of governmental securityintelligence, in the spooks and the super secrecy amongthe agencies and so fourth.

going more to public information such as canbe found on the internet. certainly you have described some of google'sproducts internationally. what is your perspective on this? >>felten: the question is about whether technologywill change the way that governments gather information in the sense of intelligence.will it be more like traditional intelligence gathering, or will it rely more on the useof public sources and use of the internet? >>schmidt: so transparency is really fundamental,i think, in modern governmental systems. you can see it now--at least i can see itin hindsight understanding the motivations of people who have information.the information can be used to embarrass them.

so there is a natural tendency of governments,regardless of political party, and regardless of which country, to want to not make informationtransparent. it is only negative from their can only ultimately be used to judge them and to embarrass is seldom used to make them heroes, even when they retire from government.i think modern functioning democracies have to have a bias for transparency, and theyhave to have laws that promote it. we have seen in the united states some ofthe costs of that. we discussed that at some length.from our perspective transparency means publishing the information. the bad information as wellas the good information can all be worked

through.ed and his team wrote one of the most fundamental papers on transparency out can talk a little bit about why transparency was so important.a simple example is do you know what the government did this week?you think you do because the media reported it, the people who live in the white houseand so fourth. why do we, for example, have all of the conferencesthat are public--that are published for attendance or they are public--why are they not streamed onto the internet? so that everybody can see what is going on,so that people who are directly affected can have an opinion?they can lobby or however our system works.

transparency is very fundamental.i will give you another example which is, i think, more controversial.president bush decided to invade iraq based on feedback that he given based on secretinformation. go through the thought experiment where thatinformation was all public and it had all been debated.he still may have made that decision, but he would have done it in the background ofa lot more conversation. for things which are of major import, secrecyseems to almost always be the wrong answer. public discussion--the difficulty of the framingof all of our debates. we seem to get to a better answer when the crowd is is all about the wisdom of the crowds.

there is a lot of evidence that the wisdomof crowds produces better outcomes than experts. the sum of people, in almost all cases, isbetter than a single decision maker. >>male audience member: could you talk tothem about how google health is doing, what the plans are for it and maybe more broadlywhy [inaudible]. >>felten: the question is about google health,and also about why the healthcare field might be resistant to adopting information technologyin the same way as others. >>schmidt: so google has a product calledgoogle health, which we will talk about in a in america is 16% of gdp. it is roughly a third more expensive in terms ofgdp percentage than other industrialized countries.

i am not an expert on costs, but i know thosenumbers to be true. i know that they are increasing. the current forecast indicates that they willget to 20, 21, 22% which becomes very difficult to sustain for all sorts of reasons.i went to announce google health at a conference. i gave a speech to 10,000 people.i thought the healthcare community is really large. then i discovered that it was the itcommunity. this was the little 1% of the community. itgives you a sense of the scale of the healthcare many cases the it approaches in healthcare have been hampered by secrecy, lack of commonstandards, proprietary standards, and those sorts of things.under the bush administration, they worked

very hard and, i think, effectively to openup many of those databases and to standardize. the obama administration has also indicatedtheir inclination to do that. you need, again, some level of transparencywith respect to, for example, healthcare codes, healthcare outcomes and so fourth to reallyaddress this. one other observation about health, why isthere not a wikipedia for doctors. it goes something like this: wikipedia isthe collective world's intelligence and it is a remarkable achievement by jimmy wellsand that whole group. we use wikipedia all the time. all of us doin one form or another. in medicine the collective knowledge of outcomes,diagnoses, and so fourth in a similar format

would have a huge impactin terms of raising the base case for doctors up to a certain level. that could be donerelatively inexpensively. google health is an attempt to have personalhealth records that are controlled by the end user.its growth rate is completely determined by whether we can plug into the existing healthcareit systems, which we are working on. >>male audience member: [inaudible]is thereany possibility, is there any way to know at this point--you would certainly assumethat you could see the trends in families. does bad information drive out good information?because of human consumption problems or because of simply [inaudible].does it go the other way, does good information

drive out bad information. [inaudible]human capacity having been defeated by the fact that human consumption of informationat the universal level describes everything [inaudible].>>felten: the question is about whether, in the information age, bad information drivesout good information or visa versa. >>schmidt: i think it depends on your viewof the public broadcasting system. if you fundamentally believe that pbs--whichwas as you know chartered to bring in this very high quality content.was it necessary? how do you think it played out, and so fourth?this is not a new debate, the interests of the common man against those of the intellectuals,or the interests of the governing versus the

interests of the governed.we have deliberately not taken a position on this question.if you want to consume the stuff, we will happily serve it to you. we try not to judgeyour tastes as being truly as terrible as we think they are. [all laugh]we just try not to. we just don't think it is appropriate forus to judge that. it is a question for the philosophers.i am an optimist about information. i am an optimist about the goodness of the human condition.the people really do want a better world. they do want social justice. they really dowant healthcare for their children. they don't want wars. they don't want conflict.they want safety and all those kinds of things.

this explosion of information allows us tosee it more clearly. ultimately it is a philosophical, political,cultural decision as to how we use it. we are going to continue to build these tools,the technology enables it. it is up to humanity as a whole as to howwe take advantage of it. i think the benefits of information so overwhelmthe negative concerns. the bad speech and those sorts of things thatare possible are so overwhelmed by the access to information, the good speech,and the fundamental benefits of using this to make the world a better me the answer is clear that it will be a better place because of it.

>>felten: i think that is a beautiful placeto stop. i would like to thank eric and all of you for this conversation. >>schmidt: thank you all.

No comments:

Post a Comment