refuge where they can escape emotionally from the stresses of society. paradise bungalow, which is what they call it, becomes for them a tiny oasis of freedom and life energy, and i won't ruin that story for you. you'll have to read that one. [laughter] >> you can watch this and other programs online at booktv.org. >> c-span, where history unfolds daily. in 1979, c-span was created as a public service by america's cable television companies and is brought to you today by your cable or satellite provider. >> host: rahul telang, in your new book, "streaming, sharing, stealing: big data and the future of entertainment," you open with: for the creative industries -- music, film and publishing -- these are the best of times and the worst of times. what do you mean by that? >> guest: so the best of time is

this is a tremendous amount of content proliferation. so a lot of people are interested in watching on variety of media, variety of channel. your ability to reach international market is never, like never before. the worst of the time is that same technology which is facilitating all sorts of expansions also means now you're facing unprecedented competition that you never faced before. especially from some of the same technology firms who are very good at creating technology. now they have decided they want to get into the content business, and they have some significant competitive advantage that the traditional firms don't have. so now it's a combination of both; you have great opportunities, but then you have to navigate all this competition as well. >> host: who are these technology companies, and what are their advantages? >> guest: so think about amazon, netflix or google.

netflix is a perfect example. netflix started as what we call a distributer. so it was in the business of shipping dvds to people. it was able to then navigate the internet and broadband world and said we could be a big player in streaming, so then it started streaming the movies and the content and the television shows into the user homes and your tablets. and then it decided that it has enormous amount of customer data now. it knows exactly, you know, what their customers want, what they don't like, and not only that, they have a very large customer base too, you know? i think now netflix is close to 80 million subscribers. so, and these are technology firms. so so these companies are built up with affinity to data, if you think about it. so they make a lot of day-driven decisions. -- data-driven decisions. so now these companies say,

well, i have all this user information. i can now use and exploit this user information in both how to price, how to promote, how to distribute the content to an individual user versus a collection of users and then maybe use that information now to actually start thinking about producing the content and produce the content that i believe my users are going to like. think about netflix from an online distributer to becoming a pretty major player in content production. and they have access to this data that the traditional television networks or the studios don't have, and we believe that provides them a significant competitive advantage. and that's the sort of disruption i think the book talks a little bit about. >> host: how do you use that data, and how do you collect it? >> guest: so for netflix it's a one-on-one relationship between the user and so-called server. so every time you log into the netflix account, ever time you

watch a particular show, every time you click a like button, every time you maybe even quit a particular show, all of that is information now netflix has. so one of the example i think we give in the book is the house of cards. so when the house of cards was being pitched to netflix televisions, including to the netflix, netflix had data about, at the time they had, say, 30, 35 million customers. so they had a pretty good idea how many of its people actually liked david fincher, how many people might like kevin spacey, how many people like political drama or even some combination of that. now you have all this intelligence, you feel pretty good about making some predictions on whether the show is likely to be successful with my subscriber base. compare that with maybe some, you know, traditional firms who have aggregate information so

you get nielsen rating, you know so many people from, you know, 15-24 age or 15-34-year-olds are essentially watching the show from 9 to 9:45. so compared to such an aggregate day, now you have netflix which knows exactly, you know, who watches what, and then it can actually even promote the content appropriateliment so the house of -- appropriately. so the house of cards, they had created, i think, six different trailers which, you know, they could use it to promote you like this sort of thing, so let me highlight this part of house of cards. so right there you can see having access to this detailed user information and having the skills and willingness to use the data, you would think that provides them an advantage, and you can see where the industry is now going. >> host: well, professor telang, wouldn't a comcast or a cox have that data through their set-top box as well?

they'd know if somebody is watching c-span or watching a football game or renting a certain series. >> guest: so some of it is possible for an nbc or cbs, i think that data would be very hard to get. they would know that my show was watched by 18-34-year-old during this prime time, but most of the television industry has sort of been driven by these standardized ratings like a nielsen rating, for example, which are based on, you know, very aggregate. that's how your ad revenues are generated, that's how everybody sort of standardized against. beyond having the data, which i think some of these studios obviously don't have -- maybe comcast might have, i don't know how detailed they have their information as well -- then the second part also comes in as whether you have the culture and the skills and the organization to actually use the day in a mooningful way. finish meaningful way.

so when you have an industry which has operated based on, you know, its gut feeling or instincts to decide which project is likely to work, which project is maybe not likely to work, you know, now you move to a world where there is this enormous amount of so-called what we call as big data. and, you know, sometimes there is no culture to use the data. sometimes there is no willingness to use the data. sometimes there is no skills to probably use the data. so, you know, having data itself as a challenge, then consolidating and use it in a meaningful way is also a very, very significant challenge, in my view. so, again, that's one of the narrative in the book, is that not only you have difficulty gaining access to some of this detailed data, you don't have the right sort of culture and people to make decision based on that. >> host: so what does that mean

for the legacy corporations in publishing, in movies? paramount, simon & schuster, comcast? what's the long-term trend, in your view? >> guest: so, i mean, one of the things that we try to avoid is predict the future. it's hard to do that. one thing i think we cannot discount is these people are very good at whatever they do, which is creating content. so they are excellent, they have a deep experience in creating the content and able to sell it to the end users. what i feel is going on, and i think michael and i kind of -- >> host: your co-author. >> guest: my co-author, yes, what we talk about in the book is the ground realities are people's consumption pattern has changed significantly. that is, the linear television from 9 to 9:45 prime time slot, i think that consumption pattern is rapidly changing where now

netflix offers an opportunity for you to do a binge watching. a lot of people don't want to wait every week for your show to come on. so all of these changes are, is a ground truth. that is what's already happening. there is a lot of data that's being utilized by the amazons and the netflixs of the world especially for pricing, promoting and distributing the content. i don't know, you might have been aware of that netflix had this recommendation challenge a few years ago. so they said we will give you a million dollars if you can design a -- to recommend a movie to its customer which is going to delight them. so you can see these companies are very much into the space where they want to provide the content to the user that they know that the content is going to be liked by the user, to offer them at a piece that the users are going to -- at a price

that the users are going to enjoy and be willing to pay for. so this is the reality, that we are already in the space. now, whether the traditional firms will be able to adapt to this reality, whether they will be able to exploit and be willing to use some of this data in their decision making is probably going to play a very significant role in, you know, how, you know, how long they keep sustaining what they have. you know, these companies have been dominant for a very, very long period of time. how dominant they remain will depend on their ability to then change to this new reality on both how consumers, you know, are consuming the cop -- the content as well as how all of this information can be utilized into making content that's appealing to end users. >> host: what if the content providers simply pulled their product from digital services saying, okay, you can only get it here? >> guest: that's a very good point. in fact, you know, michael and i

were teaching a class, and we had a senior executive from one of the studios come and discuss, you know, give a guest lecture, and his argument was somewhat similar. that, you know, we are the content creators. we know how to create the content. these guys are basically distributers, and they rely on our content. and if we choose to not provide content to them, what's going to be the future for those online platforms? there are two challenges to it. one challenge we actually provide a nice example in the book. the content providers' ability to control the content itself has been kind of degraded a little bit in the world of internet and broadband. especially when the online piracy and infringement has become so common. so you say so we have this example where nbc -- this is, i think, 2007, 2008 -- nbc had a dispute with apple's itunes, contractual dispute.

so nbc just said, all right, i'm going to take the content away from itunes and put it, i think, on the nbc web site. in the hope of maybe trying to kind of influence apple into signing a more favorable deal. our research at the time, we were actually collecting some data. interestingly, what we find is what people did as soon as the content was removed from itunes? they started flocking all the piracy networks. the piracy for nbc content soon after -- basically next day, shot up by about 15-20%. so on one hand, it's very hard for you to control the content. secondly, these, you know, the amazons and the netflixs are now an extremely powerful with tremendous scope and tremendous reach. like one of the book publishers said to us, the best thing that happened to book publishing was amazon came, and the worst thing

that happened was am vonn came -- amazon came, you know? [laughter] amazon makes it so easy for you to distribute and promote your book all the around the world, but at the same time, you know, amazon takes advantage of its tremendous power in negotiating with the authors the terms that are much more favorable to amazon. and finally, i think these online platforms also now realize that they cannot rely on content to all the studios alone. so what is netflix doing? we all are reading the stories. i think this year their budget for content production is, like, $5-$6 billion. so they said why do i have to rely on the studios to supply me content? i can hire the similar talent. i have the customer base, and i know i have the data and ability much more than anybody else, so why not me produce the content? and directly compete with these, the traditional players rather than they dictating what sort of content can be or cannot be

available? so i think the reality is a lot more complex than saying i won't give you the content and then you will just go away. >> host: rahul telang is a professor at carnegie mellon university, and he's co-director for the initiative for digital media analytics which is what, sir? >> guest: so, again, the book at some level is a culmination of the research we have been doing for the last maybe 10, 12 years. and in the process as we did variety of research both whether research questions dealt at the firm level, but also at a policy level. so we have done a lot of work on copyrights and infringement and how effective various policies are. in the process then we established this center called the initiative for digital entertainment analytics. the purpose of the center is to, you know, hire faculty and students who will commit their

time doing research relative to the media industry. >> host: professor, has, have the legislators and regulators here in washington kept up with the changes in this world? >> guest: so i think much of the focus over the last maybe ten years as far as i think when it came to regulation was basically protecting the content; that is, how to stop the infringement of the content. you know, 1998 when the napster came in, you know, the music industry is not what it was ten years ago. and, you know, obviously the content industry definitely wants to be able to protect, you know, the intellectual property and the copyright. protecting it gets very challenging because now it's the individual users who are indulging in infringement, it's very hard to go after them, it's very hard to stop them. you know, sometimes we are playing this technology whack-a-mole. you stop one thing, and another

thing comes in. you stop that thing, something else comes in. so i think a lot of the regulation has kind of focused more towards, you know, how do we make sure that the content production is protected and not misused. i don't know whether there is a lot of discussion on the market structure and who gets to dominate what. i mean, i think that's something for, you know, for future. we'll see if one of the firmses becomes too dominant and starts using its power in some way, maybe those are the discussions that will happen. >> host: do you think that the focus has been right for the legislators and regulators? >> guest: i think so. i think, i think, you know, piracy is something that we probably all agree is not beneficial. maybe it's beneficial to the consumers in the short term, but in the long run, it hurts everybody. it affects, obviously, the content creators' ability to create the content and, you know, it's a huge industry which

we export to the rest of the world. so i think that focus is probably very relevant. i think it's an issue that needs some attention. you can leaf it to the market -- leave it to the market sometimes. the media companies sometimes really cannot do anything. you know, once they create the content, it's like the cat is out the bag. they have very little ability to protect it. so some sort of legislative action or at least some sort of attention to the details is relevant. now, what exactly the form of regulation has to come about is always contentious because there are always different parties involved. but for the rest of the part which is, you know, who is going to produce what content and distribute what, think, i think it's reasonable to say let the market play out. i don't think that somebody needs to tell somebody you should be doing this or you should be doing that except as long as it's a fair competition. i think that seems reasonable. >> host: will we look back in

10, 15, 20, 40 years and see thatster as a real seminal -- napster as a real seminal point in our history? >> guest: i believe so. i mean, napster probably already has become like the seminal point that, you know, now when people talk about music industry, for example, i mean, it's usually what was before napster and what is after napster. so definitely. i think, you know, it was a culmination of both the growth of internet, it was culmination of also that the content providers maybe controlled the content a little too much and, you know, that led to all this proliferation off various web sites following napster. but i think, yes, i think napster was very important even in my view for the media industry. >> host: has the napster generation come to expect things for free? >> guest: i feel like lot of people are gravitating towards

legal channels as long as it delights them. in the sense that you offer a product at a reasonable price, at a platform and channel of your choosing, it's high quality, i think, i think people are my grating towards that -- migrating towards that. you know, some of the legislative action also has been to degrade the quality of privacy products which has also been a little bit helpful, you know, when people have to choose between getting the content from a network which has now been significantly degraded because of, you know, variety of actions versus going to netflix or amazon or hulu and get it, you know, at a reasonably, relatively, you know, reasonable price be point. i think, you know, at least in the u.s. i would say that i think it's a fair, it's a reasonable fight against piracy. you make the content available at a reasonable price, i think there is a chance that you can fight it. >> host: chapter nine of your

book, "streaming, sharing and stealing," is called moneyball. >> guest: uh-huh. >> host: what's the point? >> guest: so this goes back to the discussion we had earlier about how to use the data and information into making decisions. so moneyball, baseball is a perfect example. if you look at thely of the baseball -- the history of the baseball, there are the scouts and the managers, you know, they look at a pitcher, they look at his delivery, they look at some aggregate statistics and then make a decision whether this pitcher or this batter has potential or not. oakland and billy beane who was a small market team, they didn't have the $100 million budget of the yankees and the red sox. he said, what can i do? how can i make my team better? more importantly, i think, the book and the movie finish if you have seen the movie -- kind of makes it clear that he was, the word i think he used, there is a general fog of misjudgment. there is this echo chamber, then

everybody starts believing in whatever they see, and he felt that's not the right thing to do. there is a better way to evaluate people. and, you know, he started working with -- i forget the name of the person -- a statistician, you know, who was deeply interested in looking at the data ask the statistics and saying, well, but this pitcher which everybody thinks is not that good is actually excellent because, look, when you give him the right defense, when you give him the right ballpark, when you give him the right environment, he's outstanding. you know, doesn't matter whether he's throwing underarm or this way or that way. so what billy beane and the moneyball did was use the data to figure out which players are overrated and which players are underrated and then make the decision based on that. and, obviously, they were very successful. so moneyball was the moment in sports. where analytics became an important component.

and i think now it's very widely used. i feel like in the entertainment industry, the netflix and the amazons also have entered and said we are going to look at day and look at the data to make scientific decisions rather than make decisions based on our gut feel or our instincts and so on and so forth. >> host: so, professor telang, what does this big data look like? netflix collects it all, apple with music collects all this data. what does it look like, and how do you parse it out? >> guest: well, you know, it's a big data -- it's totally big data. i mean, imagine if i have information about every user's activity on my web site, and i have an ability to qualify and store that data and then extract the data and make some meaningful use of it. so, you know, amazon's recommendation. we all familiar, amazon people

who like this, what is it? it's exploiting the big data. i don't know whether you have heard of this very recent story that came out on the spotify, i think they call it the discover list or discover playlist. so what spotify has been doing, if you're a spotify customer, it recommends you every week a list of songs that they believe you are going to like. and the story was that it became so popular with people, and i was reading a little bit about it how spotify does it, and i won't go into the technical detail, but that's a very big data exercise where they have detailed information where they use this sophisticated algorithm called deep learning to figure out, basically, recommend to you. so when i'm talking about the big data, we are literally talking about both the scale and the scope of the data at a very micro level is so enormous and

now, you know, given the technologies and infrastructure, and they'll figure out how to use it more and more effectively. so now, you know, that's really where the trend is going. >> host: a question or a statement you write in your book and then you go ahead and answer it: you can't use data to make creative decisions. if you do, you'll interfere with the creative process and destroy the business. >> guest: so that's probably, i guess, one of the criticisms sometimes you hear from the people who are in the creative industry and say, gee, are you going to tell me that i should put bunnies in the scene because your big data predicts that people like bunnies? and are you then going to destroy our creative ability to create the content? and i think it's a little bit of a, i guess, hyperbole when somebody says something like that, that the data creep is going to hurt the creativity that way. i don't think that -- well, we'll see how things go, but i

don't think this is how this thing is going to play out. many times the data is essentially used for green lighting successful prompts. so at some -- projects. so at some level, you're using data to figure out what sorts of shows or what sort of content is likely to be successful with my audience, and that helps in green lighting -- which itself is a very uncertain decision. and then you let the content creator create the content as they see fit. there is also probably i'll add that there is also this element that somehow when the big data wasn't there, people had this up fetter ored creativity -- unfettered creativity without any commercials aspect in mind. probably never true. the commercial nature of the business meant that whatever content that was created was always created in mind with whether that's going to be successful with my customers.

people just use their gut feel versus the actual hard data. so actually what we feel is when you have good information, when you use that good information to make appropriate decisions -- especially if you can make, you know, you can, you know, on a more successful product -- i feel that the creators are more likely to be, or you know, successful because they're working on projects that are more likely, you know, have a higher potential of being successful. so i feel, you know, whether big data is going to kill creativity is, i think -- i don't think that's where we are going. it's going to complement the creativity. look at the awards, the netflix shows are winning, you know? or some of the statements i think kevin spacey made that it was the creatively most wonderful experience working on the house of cards. so we will see, you know, how the world is going to look ten years down the line.

right now i feel all of this information, just like baseball didn't tell a pitcher how to pitch, but i think simply, you know, this is meaningful and useful for the content creators. >> host: how big is the big data industry? >> guest: it's, i mean, there is no such word as big data. there's a big data in health, and there's a big data in education, and there's a big data -- what i know, i tell students, i teach students if i look at the demand for the courses that deal with the data, data science, big data, if that's any indication, then it's very enormous. everybody wants to have the skill. you go to the employers, and they want to know, can you deal with it? do you have an experience with it, you know? because even if you are a grocery firm, even if you are a hospital, everybody realizes that this information has lot of

intelligence embedded in it. do we have an ability to extract that intelligence and use it in a meaningful way to either ins hasn't our productivity or improve the quality of our product. and so on and so forth. and why not? i don't know how big it is, but i think it's very big. >> host: hah rule telang of -- rahul telang of carnegie mellon is the co-author of this book, "streaming, sharing, stealing: big data and the future of entertainment." >> c-span, where history unfolds daily. in 1979, c-span was create as a public service by america's cable television companies and is brought to you today by your cable or satellite provider. >> here's a look at some of the

current best selling nonfiction books according to wall street journal:

>> many of these authors have or will be appearing on booktv on c-span2. you can watch them on our web site, booktv.org. [inaudible conversations] ..