the retirement income they'll need to enjoy every one of their days. ♪ prudential. bring your challenges. good morning, wrvrn ooshg i'm scott mcgrew. the latest iphone is an amazing gadget. you can talk to it. the best thing to happen to a telephone since touchtone.

>> the best touchtone -- with this indicator you see how many seconds you save the new way. >> technology has broughts you us a lot of changes, but just as revolutionary are those changes in the way we interact with technology. from the first mouse -- >> has a ball at the bottom. if corowe takes. >> the graphic cal user interface to the touch pad swipe and pinch, to the touchless. interacting with computers using our bodies. the latest using the most natural way to communicate by simply asking. >> do you think it will snow today? >> sure looks like snow today. >> matchable calls venture capitalist gary morgan. the father of theory. he was the first investor found in the apple iphone. he is also an investor in

nuance, the company that powers those voice activated phone systems you hear when you call the cable company or the airlines. joined by eric savidge, forbes, and man of the financial times. it strikes me as the most remarkable thing about our technology is the way that we have begun to interact with it. it seems to understand us. is that a fair assessment? >> it is. first, i want to thank you for that, and say really the fathers are three extraordinary founders. >> dads often have money as well. >> i was the dad with the money, yes. i was the one -- sri international also was the founding source of the telling nolg. an amazing team altogether. this is the third generation of user interface in human computer interfaces. the first being the major

advance in the mouse icon. the second being gestural interface, and the third being a conversational user interface. none three developed by apple. all three xhashlized and popularized by apple. really all three led by steve jobs. it's really quite remarkable history here, and this is an interesting stage in that progression. >> mike, i can say, i press a button, and i say music, and then i can say the track. that's voice recognition, but this is something far beyond. this is where i don't have to follow a certain set of rules in order to get the thing to work. >> exactly. there are three ways to think about the theory. one is that this is a style of user interface conversational user interface. the second is that it's something new. it's an executive assistant that is virtual and takes your commands and understands them and does something with them.

the third is ins bringing a layer of artificial intelligence on top of everything that has been built here to foreand that will become your personal agent, your agent for the internet the way to -- to actualize your demands and to make things happen on the intbt. >> so it seems to me this. voice recognition has been sort of 95% there or 96% there for a long time. microsoft and others have poured ungodly amounts of money into getting to that last 3% or 4%. what was the problem with the last bit? was it scottish accents, irony, and what brought these companies over the hump? >> so great question. voice recognition has been a long time in coming. it was coming in the late 1960s, early 1970s. it took nuance communication really to create a breakthrough

and a multibillion dollar company in the category. our firm was an investor in nuance communication. i served on the board of directors for seven years. what nuance did was it created large speaker independent no training large vocabulary voice recognition systems, in i think 51 lainksz today twlash knee aunss the platform on which the theory technology arrives. theory -- nuance provides speak recognition, but what theory adds to your question, they add a layer of semantic understanding. it adds a layer of intent recognition and action upon that intent. those three things are key pieces of artificial intelligence. that's the breakthrough. that's what everybody intuitively says, oh, this is different. this is exciting. this is the future. i think in some sense we have crossed a threshold of people recognizing that and are excited about it. >> how accurate is it? does it go beyond english? can this technology apply to multiple languages? is one language more difficult

for theory to learn than another? >> i have a problem with chinese, personally, but i think -- so apple has announced three languages. english, french, and german, i believe, in the initial release. the nuance speak recognition engine accommodateed, again, i think about 51 languages, and you can expect that over time apple is an international company, and they will introduce the theory technology in multiple markets. it's been such a hit here in north america. you can only imagine that people will demand this in other countries and that it's coming. >> is there one language that is just difficult for computers to handle more than another, or is language language? >> i think language is language, but there are very big differences in terms of what do words mean, what is -- what constitutes intent in different languages, and what is funny in different languages, and -- >> i'm also curious about -- you talk about the world of artificial intelligence here. >> yeah. >> is theory learning from me as

i ask it questions, if i tell it i want to do one thing or another thing, or i want it to give me directs to one thing or another thing? is it learning my speech patterns? is it learning other information? is the software actively updated so that i can do more things with it as time goes on? >> another great question. so theory is learning -- the answer is yes. theory is learning your speech patterns and is able to better respond to you personally with -- and that really comes out of the nuance technology where it's speaker adaptive. theory is also collecting and apple, therefore, is collecting massive amounts of user data, unprecedented amounts of speech data, and what people don't expect here is just how fast the technology will move forward because of the massive amount of data that's being speaker independent, that's being collected, that will be used to improve the recognition rate and the accuracy with respect to

understanding. >> gary, our viewer at home is going to have the question. wait, it's collecting my data. show me the closest alcoholics anonymous. it's understanding the way i speak, not what did i just ask myself. >> again, this is anonomized data. apple is extremely conshe enshus about this. with any data in the internet, one imagines and i think there are real risks over time to privacy. having said that, if your data is going to be handed to anyone, personally i'm happier that it's in the hands of apple. i think they have strong corporate ethics about this, and stroth every strong -- >> it's more about the collected data, not about the individual data. not my data. it's about the way the population of users ask their questions. >> yes, but -- so theory did collect data and your queries that are anonmized and it's not

collected, but theory will also learn about you. can you get information as to what's my home address, who is my mother, what would you -- what shall i call you? so forth. all of those things become part of your personalization. you want that. you want your executive assistant to understand who you are. >> yeah, yeah. >> because they cover apple, and i think apple has been fascinating at being not the absolute innovator, but the absolute adopter and popularizer without spending gobs of money to do it. does theory have an ongoing licensed relationship with nuance and is, therefore, a bit of every phone that has theory in it is a small check going to kn kn nuance in the end? >> neither apple nor nuance has announced that, and that's not for me to say. >> is that -- >> look at him smile. >> he adds speculation every time that maybe apple might want to buy nuance at some point.

>> there has been that speculation. >> they would have to spend real money to do that. >> that's not their general habit. >> the cheap way -- >> let me get gary off the hook with a serious kind of smart alec. that was intentional. >> yes, it was, and so what was most intentional was that theory should not be boring. we saw the robotic presence of tell me and google voice action. frankly, wonderful systems in their day, but they were -- they recognized your voice, and they responded to precanned words. then big advanceses. what is different with theory is intent recognition, and response in that regard. >> do you indicate sarcasm. >> he is sashgassic.

>> again, we didn't want theory to be boring. we decided that theory would emulate a real world executive assistant. we wanted to be the best possible executive assistant you could imagine. probably she's female. although this can be customized in different countries. she is professional, intelligent problem-solve, and she doesn't take any guf from you. if you throw something at her that's sashgassic. it's going to be right back at you. we decided she should be cheeky. intelligent, cheeky, very professional, but, you know, back at you if you -- >> how many easter eggs have come out? proportionally. go ahead, gary. >> we know there are some if you can her where are you from she'll give you a smart alec answer. >> so, yes. i think many have been found, but by no means all. furthermore, why assume that this is not an ongoing easter egg hunt?

>> you can call it as you go. >> and when somebody asks a clever question, they'll figure that out, and add it to -- >> from the corpus of data, they are collecting all these responses, and you can imagine that they're saying what would be a clever response, and this is a great way to get media attention and apple has figured this out, i'm sure, so you can expect more. >> they have that down. >> have you heard it here first. >> all right. gary, the money father of a smart alec young lady. up next, a guy who whose story does not end there. we'll be back in a minute. look, every day we're using more and more energy. the world needs more energy. where's it going to come from? ♪ that's why right here, in australia, chevron is building one of the biggest natural gas projects in the world. enough power for a city the size of singapore for 50 years. what's it going to do to the planet? natural gas is the cleanest conventional fuel there is.

we've got to be smart about this. it's a smart way to go. ♪ welcome back to "press here." my next guest sold his start-up to ebay for $75 million. now, normally the story would stop there. except stumble upon's founders bought their company back from

ebay one year and 11 months later for far less money. after ebay appeared to change its mind. stumble upon is a peculiar mix of search engine and serendipity using your preferences and tastes to steer you to web sites you may like. >> for the most part we stayed pretty independent. >> reporter: its found eer garrt camp is a serial entrepreneur. he grew it to 20 million users, and he launched uber, an instant gratification service for people looking for a ride. garrett camp actually gained one of his investors in stumble upon from stumble upon, the investor used the service, liked it, and said i want in. so you buy your company back from ebay. why? why not just move on to the next thing? you made $75 million. move on. >> yeah. i think the reason is -- basically desire for both flexibility and incentive. when you are use aing smaller

company, you can move faster, make decisions very quickly. that was one kind of reason. the other reason is for engineers to have equity in a small company with a left upside, means they're more motivated. those are the two biggest factors. >> you came out of ebay and brought it so that you could give the engineers that. >> exactly. so it could be realized its full potential. >> do you think they were buying something they didn't get, or did you make a decision that this was not a business they wanted to be in? >> i think it was they were more focused on customers commerce. we wanted to pursue more media discovery. >> was there a point when ebay -- under megawhitman as she bought skype and a lot of us went what? then they bought stumble upon, or maybe stumble upon was first, and we said what? and then ebay said what? >> that's right. >> you know, i realize that you don't want to speak ill of somebody that gave you $75 million. did you wonder why you got bought by ebay? >> no. we were going to explore products, and our technology could be applied to anything. we can do photos, video.

we can do products. at the end of the day it was to explore more in the way of products. >> it didn't work? >> we never got that far. >> how often were you in house? >> about a year. almost a year. >> one year, 11 months. >> were they going to get around to it? >> we started exploring it, yeah. >> seems like a long time to not get there. i don't know. >> we explored it. >> what have you done with stumble upon now that you couldn't do with ebay or that you are most proud of post-buying your company back? >> the biggest change has been the ship to mobile. before we were actually during ebay we were pre-ebay we were basically a tool bar. we were a tool bar for firefox. the biggest shipment that happened afterwards is we moved to the web, and we could be viewed on any browser, safari, ie, whatever. then the shift that happened is developing clients for iphone, android, ipad, and that's our fastest growing one. we're really more -- >> maybe you could step back a second, because i'm trying to think of a little bit about what is the average user doing. when they come to stumble upon? what is the goal they're trying to english? >> they're basically trying to

find new and interesting things on the web. things you never think to search for. with a search engine, you know exactly what you want. you go and type in a couple of words, and you find hopefully something good in the links, but with stumble upon, you go and click a button. it knows what you like. you have set your preferences. it knows who you are following, and you are basically going to be surprised by an interesting piece of content you wouldn't have thought to search for. it's the surprise of ser endipty. it's something we try to deliver. >> i have seen it -- i've read that the filter will be the future. there will be so much information that it's not the information provider, but the one that can make any sense of it. i don't know that stumbling makes sense, but it brings something that i am interested in that nobody else would realize that i would want. is that a fair assessment? >> that's what we're trying to do. we're trying to basically find things for you that are in the web. maybe something was posted a few days ago about someone that you know or maybe it's just something that's very popular in topic that you are interested in and to bring that to your attention. normally there's so much on the web, how are you going to know to go search for that new band?

>> what's interesting about that is a lot of the approaches to presenting people content on the web has been -- you check off a lot of boxes. we're going to deliver to you a story that you like. in sort of a neat package, but what's lost is the serendipity, but you are providing that, and not the whole package of all this other stuff. >> i mean, we're trying to make it a little easier, a little more fun. we really want you to kind of like as you go give us a thumbs up or thumbs down if you like content, and based on that, we'll figure out more what you like. we'll figure out who are the other people in the system that have very similar preferences to you, and then not just people you follow, but there could be someone who likes all these same things, lives in a different city, maybe they should be the one showing you your photos because you have similar taste in photos. >> how important is this social networking part of it? is that -- will that be increasingly important as gu forward? >> it has been increasing. it starts off with any social importance really. it was clever filtering, and then over time in social networking it came where we started features. we do a facebook connect now so

you can find who your friends that are already on the site, and then follow certain people. it is becoming more important. it's not required. like, can you go in and just by signing up for a few topics they'll have a great experience. >> you mentioned facebook. mark zucker berg recently said that he had done it again, and it's very quick quote. he wouldn't have come to silicon valley. he would have stayed in boston. there's some nuance to that, but you brought the company in from canada. would you have done it again? is silicon valley still the place to be? >> i really like it. i think the synergy here of kind of people who, you know, want to take risks and start companies, people who are willing to take bets on risky companies, i definitely feel like silicon valley is still there. definitely new york is starting to happen a lot more. you have tech stars, you know, london, berlin, and i dwinl think that silicon valley still is -- maybe i'm biassed because i'm here, but i found that it's really great -- i had a great experience. >> and it's warmer. >> way warmer.

>> go ahead. >> what's the business model here? are you selling advertise sng how are you making money? >> we have our own ad model. basically one in 20 kinds will show sponsor results. those basically are called paid discoveries, and a little bit like paid search, if people want to get their products and svrss skord, they can use us as a platform to take page and put it in. the cool thing is it's not a text ad or banner. it's actually a full page. it has 100% click-through. >> i am notified of this, right? or i'm aware of it. >> it just says it's a sponsored result. if i'm signed up for photo gear and i'm in california, someone who is doing a workshop on how to do some type of photography could basically take their own web page, put it in a stream, and that would become a thing i stumble. >> have spammers figured out how to game the system as well? >> a little, but the nice thing is because since everyone is -- if any bad content arise, you --

>> are you -- they are smart. they'll curate the content, and it's self-regulating. >> we're out of time, and it's my show, and i get to ask another question. the metrics have done pretty well. >> we just crossed 20 million users. mobile is going great. 30% a month. >> terrific. >> garrett, stumble upon. up next on "press:here." the cyber shots fired in a new cold war.

sfwlimplt welcome back to press:here." the chinese made a rare public statement about hacking. this after someone attacked american sated lights with malicious computer codes. the chinese said wasn't us. >> reporter: we may never know who attacked the satellites, though china will always be suspected. china is thought to be behind

hundreds of attacks on american computers. most recently attacks on american chemical plants and on google. nothing new. what is new, though, is the willingness to point fingers. semantics ceo enrique salem. >> countries from eastern europe to asia. >> we better step it up. one -- >> and house intelligence chairman mike rogers, accused china specifically. >> josephman, one of the world experts on cyber war. he has written a long series of articles for "the financial times" as well as an entire book on the subject. the u.s. counter intelligence service actually said china out loud. they said russia too. this is new. i mean, it used to be some entity of some sort is -- we all knew they meant china. now they're poking fingers. >> it's fascinating. people have been waiting for this for a long time, because it's increasingly widespread knowledge in the security industry, in the -- among defense contractors, among many

companies that have been hit that it's the chinese. the code is in chinese. the ip addresses are in china. it's very hard to get definitive attribute use, and i think what's happened is the u.s. government has said, okay, 99.5% certainty is enough. we need to go public. >> is there a risk when we say, yes, it's the chinese? >> yeah, there's a risk. you know, they can be offended and they might, i suppose they could invoke some sort of sanctions against us, but i think it would behoov them not to do that. i think it's a sort of an opening -- an opening statement in what needs to be a more public debate. >> the government said china and russia are persistent threats to american security. parse that for me. it's not clear and present danger. it's not quite that strong, but that strong language for diplomats, persistent threat to security. >> they said persistent threat both to economic prosperity and national security, which are actually pretty well linked.

so persistent -- well, persistence is an interesting choice of words because that's the middle name of the advanced persistent threat, which is when a company gets hacked, they say it's an advanced persistent threat, which means we couldn't figure out how to stop it. >> sure. >> if somebody else gets hacked, you say it was sloppy security or something else, but if you get hacked, it's an advance the persistent threat. >> we're talking about hacking threats here. are we talking about malicious destruction of data, or is this really an -- this is more about economics, right? >> this is about theft of intellectual property. >> this is what former nsa chief mike hayden said is the greatest unprecedented theft of intellectual property in history. they're taking libraries of congress worth of data from us. they're after trade secrets. military secrets, but largely it's economic know-how. whatever companies increasingly depend on the intellectual property of trade secrets, they have. sometimes it's customer

information. sometimes they're just looking for bargaining leverage. there are law firms that get hit. they're trying to negotiate a deal in china for a joint venture, and those guys get hit all the time. >> is it that we're not -- why aren't we stopping them? is it we're not smart enough that it's a constant back and forth. one guy gets an advantage that they're able to steal that much stuff? >> i think one of the reasons the u.s. government has finally gone public is we don't have a fix. we have no technological way to stop this from happening. the government's advice includes keep your crowned jewels off connected computers. >> didn't we invent the internet? how do you escape from the lab? it's still in beta. ben stern said we were going to put in security, and then the thing just took off. it's kind of -- it's kind of hard to go back. that's why some people think we need a do-over. we need a different set of protocols for security. >> i have to stop you there. we'll have to take a break, and we'll leave that with the audience frightened, which is what you want to do with

television. we'll be back in just a moment. look, every day we're using more and more energy. the world needs more energy. where's it going to come from? ♪ that's why right here, in australia, chevron is building one of the biggest natural gas projects in the world. enough power for a city the size of singapore for 50 years. what's it going to do to the planet? natural gas is the cleanest conventional fuel there is. we've got to be smart about this. it's a smart way to go. ♪

>> my thanks to my guests this week. there are dozens of interviews like you justway watched here today. a special thank you to william sabboth on his first appearance on the show. we hope he and you will return. i'm scott mcgrew. thank you for making us a part of your sunday morning. 6