loan often have one choice, a cash advance on their paycheck. check cashing is very profitable. there's big money in small money. and that's caught the attention of silicon valley entrepreneurs who are rethinking the street corner cash advance store, creating websites like spot loan where you can borrow up to $800 for as long as eight months. the money wired to you you almost instantly. now, the revolution here is not just that the corner paycheck advance store is online. it's the ability to predict who will and will not actually pay the money back. one service uses thousands of data points about you to judge your credit in seconds. jeremy liu is a venture capitalist. he's been named several times to

forbes' list of top 100 venture capitalists. last time he was on "press: here," you'd helped sell a company to disney. jeremy, let's start with just the thing that i think is on the top of everyone's head. isn't this all a bit tawdry 1? 4 >> venture capital doesn't get involved in street corner check caring places. >> the hinge is that tens of millions of americans who don't have access to credit. so if their car breaks down and they can't get it fixed, they can't go to work. and if they can't go to work, they lose their jobs. and that's a terrible outcome. so for them it's really a question of necessity for these emergency loans. and right now many of the options that they have are difficult. >> from our perspective, any of us could walk into a wells fargo and say i need $8 right this very second and somebody would say you can sign the paperwork

later, here's 800 bucks, walk out the door. a lot of people cannot do that and perhaps we take that for granted. >> the great thing about if you have a good credit score and i imagine that you you probably do is the credited is relatively easy. you can just use your credit card and if you need $800, you can take care of itnd 80s's not a problem. but if you don't have access to credit, then it's much harder to get access to that money. about if you don't have credit history, then people don't know whether to give you credit or not. >> the next natural question which is, okay, so we provide this service to provide who ne

car fixed and over the course of the loan, it costs them $2,000. why do people need to pay such exorbita interest rates? >> the question is do you want to keep your job. because if you can't get a loan at all, you lose your job and it's a terrible outcomend at number one reason that people borrow money is auto repair and number two is emergency health care. so these are sort of -- >> they need the money. no question they need the money. and banks are giving to them. >> the issue is it's not available to them, so where can they turn. >> this is what the market will be bear. >> and when you're talking about lower credit, you codo see a lo of defaults. so the question is how can you cover that default risk. >> i. >> my high apr is covering martin when he doesn't pay back a loan. >> exactly. >> that's essentially right. and one of the fundamental issue

with that is that if you you can tell who is likely to repay better, then you can charge people lower interest rates. and that is the fundamental innovation of what we're doing and other company, as well, in that by trying to do a better job of who is good credit from who is bad credit, you can lower the interest rates on good credit so you don't have to pay for the sins. >> aren't we building up another subprime loaning crisis here by making it easier for people to get access to credit who perhaps don't understand the implications of having to pay it back at kinds of interest rates scott is referring to? how do we make sure that doesn't happen again? >> i think the key issue is one of transparency. it's like do people understand what they're paying. not just from an interest rate basis, but from a dollar basis. so if you make it transparent to somebody that to borrow this much money, you'll pay a total of that much money, this much

money a month over this many months and it adds up to this, and people say, you know, what i need it, and that is the fair thing because that is my best alternative. my alternative is i lose my job or i don't get that procedure done that i i need done. >> please tell us you won't cut this down into tranches and securitize them all. >> we end up putting our own capital at risk, so that means that we have to believe this is an economic investment. >> one of the online check cashing stores, the best sort of animalogy analogy, this is a company you invested in, it got awfully good at that system you were talking about, the algorithm of figuring out who is good and bad and sort of pivoted on that.

>> yeah, zest finance is taking what they've learned and scoring loans for other people, as well. and so -- >> are there credit rating companies do that? how is that different? what are they doing differently? >> they absolutely do. so if you look at fico, they take tens of data points. they look at your previous loan history repayments, and then they determine a score based on that. and one of the challenges for people who don't have a very good fico score is that they haven't necessarily taken out a lot of loans before. so in this world where we live in, whether it be marketing data that tells you what websites you visit or what magazines you subscribe to, all the different sources of data, if you take the tens of thousands of data points instead of just tens of data point, the question is you can

build a better idea with who will be able to get good credit. >> are you doing that thing where is your portfolio or company doing that thing where you look at somebody's facebook friends and if they're most of their friends are high credit individuals, that seem to have good jobs, they're probably more likely to repay than somebody who has a bunch of dead beat friends? kind of creepy. >> social is one aspect of how you can build one of these tense of thousands of data points. for instance you might look at court records and you might say someone who had a clean court record is a better risk than someone who has a bunch of convictions. or uniyou height samight say an example, someone who fills in an application properly capitalizing their address information, it turns out, and this zus jnt some hypothesis, but statistically looking at tens of thousands of data points over many, many loan, you can

see that statistically they're more likely to be a better credit than someone who is sloppier about typing in their address. >> and going back to the need for 10,000 data points is that all of the data points on the fico score are just so lousy. you're looking for something that gives you some sort of more -- >> more data is always better for better algorithms. it always gets you a better answer than trying to be smarter about less data. >> if he applies for a loan and is turned down, can he say why have i been turned down, which one ofs these 10,000 pieces of data have made sure that i can't get access to that money, will you tell us that? >> typically there's no simple answer. it's not like someone says -- >> hard to unwind the algorithm.

>> can i challenge your decision? >> could yyou could have bad da. >> i think that's true for any loan. if you go to wells fargo and you apply for a mortgage and they say i'm afraid i can't lend to you you -- >> under the credit reporting act, they have to tell you why. >> but they can -- i'm not pam with all the details -- >> and i remind the viewers as we get into this specific, you don't season or run the company. you are the invest to you. so i'll let you off the hook there. but continue. just want to make sure people understood that. >> so the key issue with the way that the lending decisions are made is that a whole bunch of data goes into essentially machine learning algorithm which then after loans are made you start to look for patterns that are similar between people who have repaid that loan, people who have not repaid that loan. and obviously you put a regulatory compliance overlay on that. for instance there are some things around race and gender

and some other things you can't use. you black line can't use for underwriting decisions, so we don't. but everything else goes into this machine learning algorithm which then starts to say people who are successful in repaying loans tend to have these characteristics. so we would like to have more people who are more likely to repay the loans. >> jeremy liu, thank you for introducing us to something that most people don't realize. up next, classes on create different at stanford. where ideas come from 37

my next guest is a neuro

scientist, author and college professor at stanford, who inspires young entrepreneurs.ne college professor at stanford, who inspires young entrepreneurs. she's the head of technology program, her book ingenious, takes you in to that exclusive class without the cost of the tuition. it's her 16th book which is -- have i got that right? >> you bet. >> which is just phenomenal. and rather intimidating all at the same time. what do you hope your students walk away with at the end of the semester, saying i learned this one thing? >> well, i hope they walk away with the knowledge, the skills and the tools they need to really face the problems that we have every day to tackle them. >> college students are really young. you wrote a book earlier what i

wish i knew when i was 20. something to that effect. that these less on thats that you're teaching them are falling on fairly inexperienced ears. >> but they're ripe, they're ready.teaching them are falling fairly inexperienced ears. >> but they're ripe, they're ready. they're basically saying we want to chankt worge the world, we w know how do we take our ideas and bring them to life. >> if you had to identified the sort of two or three keys to the creative genome, what would they be? what are the key three things that you say that's really a creative person? >> really good question. i've been teaching classes on innovation for the last dozen years. and when i started writing this book, i stepped back and realized i've actually been looking that the too closely. you immedianeed to pull back an at the entire system. so i created a model called the innovation engine.

you you need a base of knowledge, your knowledge is the tool box of your imagination. you need creativity tools, so you need to know how to come up with new ideas. and you you need the mindset, the attitude, the motivation to solve the problem. but will this isn't enough. you also need to be in an environment that actually fosters this.ill this isn't eno. you also need to be in an environment that actually fosters this.ll this isn't enou. you also need to be in an environment that actually fosters this. this isn't enough. you also need to be in an environment that actually fosters this. this isn't enough. you also need to be in an environment that actually fosters this.this isn't enough. you also need to be in an environment that actually fosters this. so the things you need to do as an individual, as a team, as a community to really unleash individual creativity. >> that's one of the biggerst things facing -- and i don't name any particular company -- where they say why can't we get anything going. and you say it's because nobody's excited about anything. nobody's feeding off each other. >> happy employee, good morale and freedom is one of the key things you point to in your book as something you need. you can't have a bunch of people who are desperately worried about keeping their jobs or outperforming bob next door or what have you. >> there are actually a lot of

things that go into creating a very robust habitat. the teams you're on and all the issues of who else is on the team and how you work together, but the rewards that can strike the incentive and even the physical space. if you walk in and it's filled with cubicleses, what sort of message are you get something if you walk into google or pax arre pixar, you get the message this is the type of place where creative ideas are celebrated.a pixar, you get the message this is the type of place where creative ideas are celebrated.x pixar, you get the message this is the type of place where creative ideas are celebrated. , you get the message this is the type of place where creative ideas are celebrated.pixar, you get the message this is the type of place where creative ideas are celebrated. >> there was a competition to have a good tasting but also nutritious cookie. and there were three different teams. and they were like the team of experts were maybe too competitive and the dream team where you've got different folks with different specialties. the winner is the one with the strong leader.

how critical is inthat? is it because only the supreme leader has the freedom where he doesn't have to worry about failing? >> here's the thing. people always want to have plotitudes. it's more complicated than that. you're referring to a story that i tell called the bakeoff. and they were trying to come up with the best cookie. but the hinthing is when you lo at the teams that were not successful, most didn't have the tools for collaborating in a really effective way. and one of the things i teach in my class is all the tools you have at your disposal. if you're running a team, if you're a member of the team, if you're managing a team, if you're hiring a team, to really put together a group of people that will really work together in a very effective way. for example, how do you effectively brainstorm?

brainstorming is very complicated and there's a lot of controversy about this recently. and people say brainstorming doesn't work and that's ridiculous. brainstorming is amazing. but it is a skill. it's like playing chess. if i gave you the rules for chess, would it that make you a chess player? no. but you would have to study and practice. there's knowledge and skills and motivation to make it work. and the same thing with problem solv solving, as well. >> are the young college students on maybe not the tail end of the internet, who knows where it's going, but i would be discouraged as a young person coming in where youtube and google already exist and facebook. these -- i can name three things they're not going to invest and that is the search engine, social network and whatever else i named. they came to the game too late. >> i'm smiling because those who

sold instragram were former students. >> that's an example. >> so a -- and in fact i have a story in my book about them and i wrote the story a year in advance of them actually selling the company. because what they did and what made them successful is they had an attitude of experimentation. they tried a lot of things. and basically kept trying and experimenting, willing to mine the failures. and i u use those as fertilizers for the next success. and they never would have been able to come up with instragram had they hasn't had that attitude. >> do you think here in america we're doing enough to encourage creativity amongst our young children in the educational system? because we often see them coming up, me, me, i want to answer the question. and as they get ground down through the system, it gets

beaten out of them. what should we do to encourage creativity? >> this is a really good question. there is so much emphasis on testing and on ranking kids. and of course knowledge is important. you need as i say, basic knowledge because it is the tool box for your imagination. but you also have to teach creativity tools. people often ask can you teach creativity. can you teach art, you can teach music, can you teach -- of course there are some people who have natural mathematician, but everyone not only benefits, but it's required. why are we not teaching these tools? just as we teach math differently than music or sport, we teach creativity in a very different way and we need to provide teachers and schools with the tools and permission to do that. >> she's the head of stanford's technology program. thanks for being with us. >> my pleasure.

well on back to "press: here." you started talking about those ads that i get in places i don't expect. sometimes i get texts, i expect that, but they're starting to proliferate. >> everywhere. what's happening is there's been this explosion of creativity in the app world. of course the developers have to make money out of it, so they sign up these ad networks who basically stick in the app and then serve ads in the app. >> that's fine. >> but how we have the attack of the aggressive ads. and there are some networks that have software that allows them to push ads into for example your text or they'll deposit an icon on your desktop. you think what's that icon, you

click on it and up comes an ad. >> and you think to yourself i didn't allow that, but you probably did. >> it's not entirely clear. so joe downloads an app, this is a real gray area because you give permission to the app to take certain pieces of information about you like your e-mail or your phone, but occasionally what happens is this network pops up and will take that information without you knowing and can do stuff with it that it's not entirely clear you've given permission for. >> worse on android than apple? >> worse than an destroydroid t apple. some develops have had death threats. >> i think it was path that first was uploading -- one these ad networks will go too far and

get the government involved and government will finally say that's enough of that. >> so what i've said is i think the industry needs to address this quickly. for example, it should make very clear up front what information is being collected, if it wants something like your phone number or your e-mail -- >> that would be nice. bad kid in the back of the class will get -- >> what if the industry doesn't. >> joe, you wrote an article about facebook actually monitoring our chats and our ims and things looking for those people who try to entrap teenagers. from a sexual predator point of view. i wasn't aware facebook did that. >> a lot of people weren't, but it shouldn't be shocking. google minds our g-mail to figure out what ads to show us and you'd like to think facebook is doing something on security to stop those horrible click here and now your friends are infected with something. but they've also taken a step

further and they're looking for sex predators. and it's something that almost nobody in the industry likes to talk about. and that's in part because of the privacy issue. you sort of wonder exactly who is looking at your chats. and secondly it doesn't always work. most of the time if there is a sex predator case, they find out after the fact that these people use facebook. but facebook is looking specifically at stranger on stranger interaction and when it does happen, it's typically that the victim knows. >> so it would be something like here is a man who lives very far away from and has an awful lot of relationships with young, really young women. 15 or 16 or -- >> that's the sort of thing they look at. they look at the web of relationships. here's a guy in his 30s with a whole bunch of friends or people he sends messages to that don't respond that are underage. >> and do they actually look at the messages themselves? >> they do. but it's all -- the staff is

fairly small, so this is all programs that are looking for this stuff. and not just key words, but patterns. and it's informed with among other things the schchats of convicted pedophiles. and if enough things trigger an alert, then it goes up to a human being that actually looks at it and occasionally kal calls the cops. >> it was your original reporting. and 35,000 people liked it. when they reported about your reporting. is that discouraging? you're the guy who figured all this out. >> i'm still trying to figure out the internet thing. i've only been doing this since '199. so the reuters.com page where i guess nobody goes,99. so the reuters.com page where i guess nobody goes, i spent three weeks on the piece and then just a couple hours -- >> nothing wrong, but -- but when you say i heard of that,

joe did it. "press: here" will be back in a minute.

a programming note. we'll be stepping out of the way for the next three weeks to make way potolympics. so our next episode will be august 19th. thank you for making us part of your sunday.

today on "class action," the philanthropist takes on the governor. at stake competing box proposals to raise taxeses for school i had kids. molly is here to talk about the high stakes election coming up in november and her plan to california schools. and then imagine that, an education uncould yencute bait s to revolutionize the way students learn and the way