>> tell us what you're reading this summer. tell us a tweet@booktv. ♪ ..

you. number let me start by asking why think the most basic question about this your book is titled filterable that technology can't do anything on its own without choosing to make it do something. who's doing the fighting and what is it that we are missing when things are being hidden from us? >> guest: this is a short cut, but certainly increasingly in number of the top with sides and services even other applications are editing the experience and a personalized way. that means is essentially they are collecting data from us and using it to carry it to prioritizes the pieces of data and show things they think we

are going to want to see. afton google this happens in the search results. different people get a different search results based on with the click on in the past, what kind of person they are, on dahuk news it happens actually the stories themselves, and this is spreading across come it's been based in two or services online for the simple reason that if you can provide this sort of code word as relevance then you keep people coming back and presumably provide a better experience for them, but the danger is because it's very hard to see how your intranet differs from anyone else or what basis it's being edited, it's hard, you don't know what's being left out, you know, or not seeing what other people might d.c. and

most people don't have a sense that it's happening at all. >> host: the internet is turning the accessible into the unknowns. you're not even aware that it's out there. so, you talk a little bit about knowing what kind of people we are, using that as a way to filter search results in yahoo! giving the same thing with yahoo! news and so forth and get this isn't just a matter of being locked in. you are saying google is doing this even if i don't have an account if i'm not logged in as a user of edge e-mail or the calendar just as an ordinary citizen google is still doing this? >> guest: this is the big shift. one engineer told me they collect 57 different variables about people who even if you're not locked in if you take a new laptop and put it on the desk right here you can tell what kind of laptop is it, what kind of software, the size of the font on the laptop where is it

located and how long are you lingering before you click on the link. in all these things can be used to make guesses of what kind of person you are, there's big font and small font people, and all of this then allows you to make these guesses. this may not be a very good portrait of you don't need that much in order to be able to do this with an increase in optimization and talk to the folks at hunch, they got in personalization site and they say actually you need very little data in order to start to have a lot of predictive power five data points, five particular data points you can then get any other data point within acres it accuracy if. >> host: what would they need to know about me to make these

predictions? >> guest: . >> i can remember about four of them, so i will do the four. there's gender, introvert and extrovert turns out to be an important signal. there's political orientation, which is a proxy for other lifestyle things if you say that turns out to be -- [inaudible] some kind of proxy for the generational taste essentially. do you like the hilton or some sort of trend the hotel, this like -- >> host: generational case rather than age, so hilton verses the w. as a hotel choice,

is that a better predictor than knowing if someone is 25 or -- >> guest: it's a good proxy. he tries to ask questions that don't directly ask you the question they are trying to get that, and so that's -- you could say a liberal or conservative it is more fun. as a mix of people don't get weird about the answer. >> guest: it helps get the information out that you need, so with five data points, you can then predict out the eda% accuracy almost any of your question when you're saying here is five things frank in the order of this person. not only that if you have to people answering those questions and you know their friends with a third person to you can

predict with 80% accuracy having gathered these five data points. >> host: this is the social networks that we are all part of our exporting information -- >> guest: it's one of these things i don't -- i think people who say privacy is over often have an interest in saying that and i don't totally buy into that, but there is this way in which what is known about us can increasingly be inferred by what is known about our friends. and it's in their hands in some sense. we've crawled sourced it. >> host: there was a paper written out of mit singing we can infer your sexual preference and figure out whether you are gay or not. based solely on your social net for confirmation. but that is a general principle, that kind of inference, the general principal of the social network.

>> guest: you wouldn't want to make a lot of particular -- you wouldn't want to use that to make important decisions for the simple reason which is that -- let me rephrase that if you're a business you might want to use it to make important decisions. if you're an individual we really don't want that to be the way the decisions are made about us because guilt by association not in the case of being gay obviously it in the case of, you know, you can get some credit rating on their friends. now, should banks not offer people loans because their friends have bad credit ratings? that's not -- i don't think we want to get into that world. but we are -- actually because there is so much of the state and its what's done with it i feel we are more in that world and we think we are. >> host: . >> host: we had a scandal 15 or 20 years ago where the banks

would red line certain districts and the famously wouldn't give loans to businesses in harlem because that became a essentially the metric of each individual business. what you are saying in a way is that each of us is potentially in a kind of district of our own in that way and each could be a red line for things true of our network but not of ourselves. >> guest: that's right and there's a great book that details how geographically there's this self sorting happening. we live in neighborhoods that are increasingly homogenous. my neighborhood of four to green in brooklyn has people like me, that's true across-the-board, and, you know, you what kind of hope that at least when you go on line you would be stepping outside that a little bit. people like me but at least i'm going to be exposed to the broader world. increasingly i think we are

seeing that kind of situation where there is, you know, the information that we are exposed to get sort of as well based on what people would like. >> host: so that's an marmot of judgment about what should be the case, and yet people are moving into neighborhoods of people like them out of choice. this is not now forced segregation. it's ghettoization. >> guest: the second is about affluent people moving to where people like them who live, and it's about who gets left behind who are then a sort of self served by default in other words you have these cities where people basically detroit or the rust belt city where the people who can't afford to solve sort in some neighborhood and chicago or new york people like them

because they can't move. >> host: it sounds like the danger that you are focusing on here is somewhat to meet individual. you don't want people to make judgments on a us based on guilt by association but it's also about the society as a whole if i understanding you the danger of this is about what happens to the larger social groups not just the individual. >> guest: what would be wrong if we got a better and better at relevance? >> guest: there's a couple things. so you know, one is just this very basic thing of being able to empathize with, understand what's going on and someone else's life or what a different perspective might it look like and if you are exposed to lots of different diverse pieces of information or seeing stories

that relate that are not that relevant to people like you but are relevant to other people, then it calls your attention to certain problems. the thing i think about a lot is homelessness and it is very easy to pretend that problem doesn't exist at all. for if you live in certain neighborhoods you can imagine that nobody's ever homeless. that's not, you know, that's not a very good situation to be in as a society, and actually you do experience -- i was at a conference with a lot of fairly well off entrepreneurs and venture capital people, and the idea there was a crisis where a lot of people didn't have jobs in america, that wasn't the crisis for them because they don't know anyone who doesn't have jobs, and frankly i don't know of that many people who,

you know, out of work because of this recession, yet we know they are out there, there are a lot of them -- >> host: but it's just a statistic. only 5% of people but no 1i know personally. >> guest: right, and that's a dangerous -- that's what's concerning because, you know, you want to have some ability to track at a societal level here's the things that matter overall even if they are not relevant, and i'm very fond of this quote from mark zuckerberg where squirrel body in the front -- facebook news piece is a squirrel going in your front yard may be more relevant to your interest right now than people don young in africa. he's saying let's show people the squirrel, and a sort of points to the problem which is it wouldn't be good if we were just working in our front yard.

>> host: so we were going back and forth between -- a statement like that, it wouldn't be good if people were not looking at the front door and seeing you want to be in a situation where you are more aware of what other people are saying, but i think for a lot of fuss we actually don't want to be in that situation. we are actually preferring having relevance given to us. so, how do you get a around the kind of eat your peas field to the argument you are having? >> guest: we of obesity -- you know, the -- i think the challenge is the question of want, and saying in a way why shouldn't people get what they want? what i'm saying is we actually want a lot of different things. we want to eat cake and beef in

and healthy. we want a lot. >> host: watch 30 hours of tv and be fit. >> guest: we want to be entertained, and just have fun with our media, and we want to be informed citizens and to know about the world. there are few people who are not going to check that on a survey, yes i want to be an informed citizen, i want to know about what is going on in the world, and so then the question is given that internal tension how does the media institution or the -- how do these things serve different wants? and i think what's going on here is that because you can easily track the more short-term kind of compulsive want -- on this right now. i'm going to click this -- you get the idea that somebody is in

that direction, kind of a compulsive media. and this is a lot of what facebook is. it's very, very rare that i or anyone that i've talked to walks away from facebook saying i had the most revelatory experience on facebook. i learned about the world. it was a rich experience. that's not what happens and it's not built to do that. it's built to give you a little compulsive tidbits and flash of dopamine or, you know, feel good, but it doesn't actually -- it doesn't change the way that you look at the world, and it's almost as if next to the like but in it was hard at first but it changed my life. [laughter] >> host: [inaudible] in tsa world, there is now an

attempt to do that pre-commitment. i would like to say that i'm going to read that leader. there is an application that recently started returning the results saying in the past you said you wanted to read this leaders a you stretch out your time commitment. >> guest: that's the and, you know, where -- to be clear i'm not nostalgic for the good old days when there were editors who just -- >> host: i would not accuse you -- >> guest: but, you know, i do think that there are things that those media knew or learned how to do that not do the to have not yet made it to a lot of the algorithm like editing that is readily happened. >> host: that is certainly true also i spend a lot of time

in the media environment. a lot of what we attribute to those companies is the scarcity that was around them in a way they could be hit because they were not under quite the same competitive pressure. i remember when walter cronkite died, the outpouring of the motion in part because we were burying the dead men but in part because we were burying the world. host >> guest: you can have somebody say that's the way it is without fear of a contradiction. >> host: but here i think comes the big tension around this harnessing of the media entity force of the good which is each of them is just a commercial actor. so you tell the story of going on to facebook and friending people more conservative than you politically, and facebook not showing you what they were saying that not surfacing their activity in your news feet because facebook observed out of

funds observed you were not actually interacting with of those people. but is also possible from facebook's point of view if you want to see what someone is doing without interacting that you are spawning on them and facebook isn't in the business of showing you what other people are doing if you're not having a kind of real connection? >> guest: welcome to know, it gets to the question of revealed verses expressed preference. so, this is the should i trust what you say you want or what your behavior indicates that you want -- >> host: the line of serving you and -- >> guest: in this case i said i want to hear from these people and facebook said yes but you are not behaving as if you did so we are not going to show that to you and this is one of the interesting sort of power struggles in the kind of media. netflix does the same kind of thing now where it used to be all about expressed preferences. it used to be clicking on five

stars and therefore they will know that i like these kind of movies but what they found was it didn't work that well because people gave schindler's list five stars ethan -- how could you not? it's a great movie about the holocaust -- and so, now that they have the streaming -- now that so many people are streaming movies on netflix they can say to watch five minutes of that but then flip over to the fast and furious, so that is actually what you prefer and this gets me back to this thing out what is the preference? and should you trust me when i say that i prefer something or should you trust my behavior? i would argue that even though people be given the way they say they will find there is a lot of value taking them at their word and giving them that kind of agency that just as a general

principle you want people to make decisions about what they do. you don't want to have to be given a start with it than -- >> host: businesses don't want that. businesses want people to get -- again, the tension between the sort of commercial and civic value. it strikes me when you say you have to trust the when i say something but you are pointing to the idea that there's two different me's. there is the me that's going to discuss the next five seconds and the next five years. if i commit to wanting and next sense of the world in the next five years businesses that serve me five seconds at a time are undermining this longer-term. >> guest: i want to be thin and healthy person that burger king burger looks awfully good, that's the sort of attention and you can say it's people's choice they can choose to have burger

king or not and to change that behavior you both knew better signaling from i would argue the companies to say here's how kofi this is for you so you have some guidepost on which to know this and then you want to have people actually have options and then encourage people to use the options well. those are three things you might do to confront the problem. here you don't have any sense of what the other options or necessarily. you don't necessarily know what else is on the menu of, and part of what this is about i think, you know, sort of trying to make this more passage of the media in a way that -- >> host: what demon?

make online more passive media? >> guest: so he essentially what a lot of us -- what i love of the internet is it is an interactive thing. you have to be leaning forward or something, you can't -- at the same time if what you're trying to do is deliver ads or media that has ads in it or whatever that's not necessarily what you ought because people are going to click over to some other side or whatever so how do you provide a stream of things that sell that keeps you there, that keeps you watching youtube instead of quitting over to facebook or whatever, and that is fighting to talk about the competitive environment what people have realized -- how do you do we enough of the legwork that people stick with you finding the next thing. >> host: so everybody in

america has grown up in the environment where an advertiser can get so much information about us on our zip code than if you tell me your zip code i can tell you roughly what you drink and put toothpaste you use. we never lived in a world and relatively satisfied we may have preferred it to the world in which advertisers knew nothing because of the misfit. how is this different than that? the so-called prez some of colluding of grouping people together by zip code. how is it different than that and is it a bigger threat than that and if so how much bigger? >> guest: one way in which people it's different in that the grouping can be much more desperate, so if i'm a lawyer in lines of code, you know, you have the geography, the geographical soul-searching and

order personality self sorting happening flying apart. they are not happening in the same place some important consequences because there are values to be in actually plugged in to whatever is going on in your brutality. a broad way that it's different -- one important way is not a shared experience with the other people right around you, so if you turn on your local tv news, you know, then you are all in the same stories and narrative for better or worse. this is -- you can live in a different media universe than the person right next door. but the bigger problem is again, that you don't know, you can't

see the contour -- you don't know it's local news even if it is local news in a way because it's -- local news is news for people that live near you. we are all -- we are increasingly getting local news in a different sense but it doesn't feels that it's that. it doesn't reveal that. >> host: when i do a google search it's not that i'm in the locality of one. >> guest: in the same way on google news you don't get a sense of how -- timoney standard deviations am i off from other people in my area and the

average -- you don't get a sense of how different your experience is and that's important because the process of figuring out what is going on, what is true or real is the triangulation where you're comparing the study about liberals listening to rush limbaugh and it doesn't make to them change their mind. they don't all of a sudden realized that rush limbaugh was right. the start to realize a lot of the things come a lot of the beliefs and they held strongly aren't supported buy very much information our arguments that they have. that she said something and they say i think that's wrong but i don't know. and it calls attention to the limits of your understanding of the topic actually.

and i think that's a really valuable thing. if you want to have your attention called where are the places i know i know this and where are the places but i may not. the known and unknown parameter and you don't on the unknown unknown. >> host: i didn't even know people think different -- >> guest: right. >> host: . you call these companies a code word. is this what it's a code word for what is being -- >> guest: there is a simplistic idea of what relevance is i got into this argument with google and said okay, you say you want to provide relevance, but if i am a 9/11 conspiracy theorist and

googled 9/11 is the most relevant result most likely to clich which is actually how you are mostly defined in personal lines or is it the popular mechanics article that debunks that stuff -- it's relevant in a way that is uncomfortable, and i will get back to you. it didn't -- they didn't have an answer but i actually think that is a hard question. >> host: have you post it to the search engineers? >> guest: i did have the conversation with the search engineer and they said we tried to have a diverse -- we tried to bring diversity into the mix, but i actually think that it's not what -- i think when you hear eric schmidt or larry page say what we wanted google to do is provide you one search

results, the right one it doesn't really work in that. it just doesn't make any sense in a way for a certain kind of queries and ways people are using this. there is no right one for 9/11. >> host: let's take a break and then we can talk about what we do with all of this. >> we ask what are you reading this summer. here's what you had to say.

>> host: in the stories you tell about google and facebook and the filter will, you talk about the kind of political limitations of what it is we can see about the world, and for most people who think about the internet either think about it for a living or the interest of what is happening, that, anything having to do with the thesis of the internet effect on political conversations was introduced with the book republic about, cannot ten years ago there is a 2.0 version of it but introduced the idea of the

echo chambers and the selecting enclaves on the internet just as a general concern. what -- how would you position your thesis relative to sunstein 's republic.com >> guest: talking about sort of an explicit salles selection, i am a neo-nazi who is going to the news online and even, you know, i'm getting indoctrinated and ending myself up and i can exclude the whole rest of the world i think the first thing that's different about this is that it is this passive experience it's not that people have to seek it out at all or know that they are shaping -- making decisions that shape what they see. it's just increasing the they view the web sites and it's the

kind of step that seems to be relevant or that you don't see the editing at work and are not making a decision but i think that decision is important because i think, you know, understanding what the editorial sensibility that you're dealing with these important. if you are trying to kind of see what the world is knowing i'm looking at this through fox news, probably not going to have good stories about obama. i should watch that and make sure that i'm not -- looking at this on msnbc and, you know -- so, that part about selecting, choosing the filter to which you see the world i think he has that as sort of okay people are going to go out and do this i'm saying you don't even get to -- you don't even get to the place -- >> host: you didn't get to the

moment of choice because it -- >> guest: it just comes to you. when you hear people talk about the future of tv, you know, it is very much of this future of tv is show after show someone like you will like. i often think the because of the nature, the way that the algorithms are constructed this is where it is the parts more from sunstein's world in a way there are things that fallout sort of as unintended consequences of doing that, so again, to do in the future of tv, you know, presumably the news comes on very rarely for a lot of people in that future model and you know, certainly a half-hour block of news that has foreign news and whatever, that's just not -- >> host: it's already gone to celebrity news and much less of

its original focus. >> guest: right, and there's a political science marcus prior that is interesting work on this and tracked sort of all three channel tv world's versus the cable will reverse is the internet world, what is it due to the shape of how people -- of political knowledge and he basically says there's a good reason to believe that and that old world where there are three channels and if you wanted to watch tv at six you have to watch the news. you know, you had a bell curve essentially that you had most people grouped in the middle in terms of what they knew about politics, and as you increase the amount of choice you get more of the power curve where he have news junkies amazing the much. crusco straight into the vein. >> guest: but the average

moves downward actually on the curve because on balance people don't necessarily preserve should i watch another episode of -- >> host: housewives or whatever. >> guest: yeah. again it's a challenge of if you're going to look at the world through these algorithms to our trying to figure out an idea of relevance or of what you want, how do you balance those conflicting impulses and get it to do what i would argue is the great media which is yes the new yorker has these long articles and cartoons in the little -- and most magazines are like that. it's actually part of the magazine for is its entertaining enough that you will read it and you will get to the long stories. >> host: not all magazines are like that.

>> guest: but rolling stone is an example. a lot of pop nonsense and music reporting and whoever is doing the experts say on the political world that is a really nice format for having people know about the world come and as you get bundling and the different pieces available without having to see what other pieces they are grouped with you leave some of that and it requires the algorithm to be much better about building some of that back. >> host: so you are simultaneously saying that the large size chunks of media whether it was really had three networks and we had the news for an hour every day and magazines that had a little bit for everybody for whatever they wanted to reach in order to get everybody to be willing to support the subscription everybody had to subsidize it

reveals a little bit. but i'm struck by how much all of that is in engineering terms absolutely inefficient, that on engineering it is unbelievable much better and yet we lose the kind of cross subsidy of the bundling. >> guest: the question is how do you define efficiency. >> host: in defining it in engineering terms. >> host: if you say five-star movie is worth way more than four because actually what you want is only to have people have a good time on netflix, but have movies that stay with them for 20 years or whatever. this is where you see some of this. >> host: here's the question. engineering almost never reverses. we never get to the point we miss those who think this is overrated in terms of the combustion, so we've got --

we've got this explosively efficient network now that is casually unbundling and destroying and the engineering and efficiency that made the media environment will look the way that they did come and you are not eight nostalgist, so we can't use the old masses to give the old value but it sounds like you are saying is the old value exposure to the ideas and concerns of lots of people, cross subsidies of one another's interest still important, so starting from here and moving forward rather than backwards what do we do? >> guest: right, so i think there are -- i'm pointing to the two related but different problems, one is personalization isn't very good right now. >> host: it gets worse.

>> guest: it doesn't understand -- again, it doesn't have a very good view of the nuances of what we want and what we consider. >> host: personalization. >> guest: it's still very much you are interested in computers you're probably also interested in software so it's not drawing of a larger inferences because you can actually do very well in the narrow territory, so you could use it to produce people to more than they've been introduced to but because it's not very good you just get -- the second problem is yes there are some places you actually want to introduce things that are pulling in some other directions that are not going to be at least in the short term

the most efficient so you want to say as opposed to the market argument for that would be that that is actually with a satisfying media -- discovery and i didn't think i would be interested in this but it turned out that i was -- >> guest: this is why people stick with -- even the old media environment wasn't as optimized as it could have been, and magazines or newspapers didn't just -- there were other things there and part of it is that it reflects back to an idea that you are learning about things, that you -- it gives you that feeling of getting to know about the world and a broadening your horizon and people like that and the algorithms we have right now

are good at providing that feeling and i would argue some of that it's going to be awhile before we start. knowing we can get more people to read this piece it directly follows this piece about something else because they have some kind of an intangible symmetry or somehow working together. that is some of the art that needs to be built back into this so i'm not saying -- engineering doesn't go back but it can learn more from -- its like yes, we have cars now, we are not going back to horses, there are other things to learn about how to build better cars, and so how do we learn from some of the finer points of that that made the best media in the 20th century how do we take that and then figure out can you code this in? and then the other piece is how do you provide people with

better ways of -- how do you make this a more multi dimensional kind of sorting system? so to go to facebook for a minute, "light" there are particular things they're easy to like and particular things that are harder to like, the things that are easy to like move across facebook more rapidly and harder to like, you know, there's no inherent reason would be hard to say there's also an important button and disliked, and in a way this is -- we've learned so much in the sort of tagging about all the different ways that we can sort the media but we -- this comes up in facebook in the sense of facebook doesn't really know why

naturally friends with, who is my family -- it would be a better service if you could figure out a way to do that that was on intrusive, so there's a lot of -- partly i think there are -- you can get a big game by just doing this first step and getting it to your -- we are going to show more stuff about gadgets, but to get to the place that it's actually good it may not deliver as much profit, but it's harder. >> host: it doesn't sound like a winning mix for business. >> guest: i talked to them on facebook and it said look we just love coming up with clever ways of making people spend more minutes on facebook. that's what we love to do, and what you are asking us to do is mess around with these complicated social questions. this is all kind of light, you

know, we just want to do the fun symbol stuff where we figure out a clever -- we show people this and they will be on facebook more. you know, i think the counter balance of that could be people start to understand how this is all working and that there's some kind of literacy that emerges where people say i really actually want to be able to sift this differently or i want to have a better experience and then they start to provide the tools and find people are taking them up because they have the desire. >> host: that's the supply side and i am struck when you're talking saying we should do this and we could do that with algorithms and tune in and so forth and it's pretty clear that class of engineers who work on gathering the signals and making a predictive models netflix

recommendations taken to some kind of civic level. but most of us are not so become valley engineers, so there's also the demand side. so let's say i am a citizen who thinks i'm never going to be a news junkie. i'm not going to be someone who consumes anything on any subject because that group of people is in paradise, but i do want to use the network to get more -- a broader sense of the world than i am currently getting a and i can't go in and change the google search algorithm what can i do, what can an individual do to say i want a and its varied media? >> guest: i think different platforms do this in different ways, so you know, facebook does this a lot, twitter at least until yesterday hasn't done this very much. i lost my conservative friends

on facebook. i never see them anymore. i see them on twitter and that's because twitter has a very simple transparent rule which is to follow them if you're logging on to twitter now and they are treating you will see their stuff. -- what happened yesterday is twitter is actually starting to succumb to the same motivation for the same reason so when you search for something on twitter you get the most relevant for you -- and it explains a conversation i had where it's so great you are holding the line on this and they said yeah it's really hard, it's a hard question. you know, but at the moment any way it's a tool that does at least allow you to say i want to hear from, you know, i follow

karl rove and rachel mazel and "the new york times," and together i'm going to kind of get a sense of what's going on. >> host: so twitter if i follow someone it isn't affecting the stream but they are manipulating the search -- >> guest: which is a lot of the way people -- twitter, the phenomenon of global climate change and maybe i'm going to get results that are for people like me can now move to twitter and google in climate change. >> host: there are no climate change diman years. >> guest: again, it's i think useful to all of us to see that come to be reminded and to actually understand those arguments. we have and i really mean it but in. >> host: it sounds like you are suggesting backing out of

the alveringem design that my expressed preference i should be allowed to override my own personality with an expressed preferences i want to. is that right? >> guest: i should know -- it's even more let me turn on or off your behavioral reading of me. sometimes i just want to come home and a vigil in front of the tv and you do your best to show me videos i think are going to be funny and i will keep watching as long as you do. that's fine, but i want to be able to opt out of that some time, and i want to have to be -- i want to know that that's what you're doing. and i think part of the challenge where we are right now is that i think a reason that this is not transparent is it creeps people out when it's done explicitly that actually if you

track the history of this it started explicitly and it started with microsoft and these little personal avatars and like word, the little cartoon paper clip that would go out and do things for you and it's annoying and creepy and very bad. it just doesn't work that well when you think of it is a person who knows you who obviously doesn't know you very well and is in a very smart. if you in bed that underneath the surface of things and you are not saying to people i'm doing this task you're just showing them stuff we will then click on the links much more. the experiment is i'm not sure if you had a tag that says personalized google, regular google i think a lot of people would choose regular.

>> host: if google is bending things in that direction. i'm struck to the story there was perry mason about mit that work on personalization. the called it, but if children ferdinand developed lido was exactly an explicit you say what you like, and people hated hearing when they made explicit choices they hated hearing it was predictable because of the likes to believe that eclectic music. and they would see if you like metallica you may like flyer so how could you predict -- when the perception came along which is the company amazon -- no one

was ever asked explicitly to opt into the system and have the cognitive dissident. >> guest: that's right, the people are willing to be shown things they might like but they don't like being told that you're going to like -- >> host: so you say bring the cognitive dissidence back. at least tell people this is how the sausages made if you don't like it -- >> guest: right. and what that does then the argument, the other argument again and again with engineers is they say people aren't really demanding this to the different services that you want, and were the algorithm to be changed in a particular way, and so how can you have consumer pressure about a product they don't know exists? you can't get -- it doesn't exist. >> host: it would be like the cigarette company saying the additives we are putting in the tobacco nobody is complaining about them.

>> host: so let me ask you this again in a really explicit way. i am a chardonnay is willing hugging and globally minded liberals. i want to see a different slice of my own country, of the world i am currently getting at with google and facebook and so forth. so i want to break out of this bubble that you have convinced me exists. i go home and open up a browser and then what? >> guest: i would say right now the best and the simplest thing you could do is to go to twitter and put together a stream that includes sources that get you out of this and the nice thing about twitter -- one of the things the newspaper

front page does come in and you pointed this out, that you're not going to lead all the stories but at least you know the exist, and you can skip over them but it changes it from the unknown unknown to the unknown unknown. when i follow karl rove four bollenbach which i do follow, it doesn't mean that i read all of the stories, but at least then i start to get a sense of okay, this is what fox is talking about, and you know, i also follow the bloggers in the middle of the country that are not a part of the coastal -- it's actually fun exercise in a way to say what are the different places here that i need to touch in order to get kind of a pretty good view of what's going on? of course you never did -- is the sort of realism in the

philosophical terms like you see where you believe what you see above the world and we all do that to some extent as best we try to say i know things are being edited out. i'm going to pay attention to that we are susceptible to that, but you can if you are choosing and you want to get a good view of the world i think you can do a lot better than facebook or google or google news for world you can get a much better view. >> host: so you believe what you see what you are saying that the feedback now increasingly throop search and personalization what you believe is they can inferred your belief, what you see contrast, darbee leaves contract and what you see. so i go home and i find people on atwitter who are on the areas

i care about, but twitter is not yet a news feed all the wood has some of those characteristics, and i think about the source of personalization which is a google news, yahoo! which are doing these kind of filtering. if i were to change my news what to buy two? >> guest: i think twitter does most of the proxy for news and most agencies are on the twitter and actually i think that google news has wrestled with these problems more than other parts of google. they thought about this and because its -- people think of it as a kind of machine news, the dirty secret is a totally

won't work as soon as the newspaper editors go away. it weighs what the newspaper editors judgments are very highly and even the placement on the page and what the font size is to all of that, those are human judgments. it is aggregating of the judgment of expert human journalist editors not doing some sort of -- >> host: artificial. >> guest: right. so in a way i think the google news does some of these things that google search for example i don't know it's doing as much. - -- >> host: it does consume human judgment with a choice but what you're saying about news is it is consuming the judgment of a particular class of trained individuals, not just anybody that is --

>> guest: if newspaper editors are disproportionately weighted and google news as i and stand from barletta and marrec, so it would be dangerous i think for other people to look at the google news and say this is the future of the news because the future of the news won't be dependent necessarily on a class of people that google news, bumbling and alienating. i think the thing that i am excited about in a way is i think the solution to this is doable but not yet done at an algorithm alberta is mike level and i start to go out and talk about the book run into more and more people who are at journalism school or, you know, programmers saying we are

working on that product that does this it's like an exciting you view this as a problem for these companies or you can view them as grand challenges. >> host: so you outlined a supply-side issue which is changing the alveringem and outlined the demand issue which is i want my expressed preferences to be weeded maureen and i would like to people to know more of what is being done to me and my name. if you could change one thing, if you could pick -- as a place to start if you could pick one of your suggestions or hopes from filter bubble to see what's happening in the world in a short term what would you change first? >> guest: i actually probably wouldn't do facebook or something like that. i would -- i think there is a sore left low hanging fruit intervention that could be made with the facebook news feed that because facebook is