tv Press Here NBC August 2, 2015 9:00am-9:31am PDT
♪ ♪ has airport security gotten so poor it's time to take the humans out of the loop? can computers find the dangers that tsa can't? security expert dr. lisa dolev, my guest this morning. plus everyone in silicon valley has an app, even the firefighters and the search for little green men gets a pile of green. astronomer, and our reporters from japan, and "fortune"
magazine this week on "press: here." ♪ ♪ good morning, everyone. i'm scott mcgrew. recently the transportation safety agency the tsa, tested its own effectiveness at airports throughout the united states attempting to sneak guns and bombs and other weapons past tsa officers. the results were abysmal, nearly every pistol grenade, ied and knife made it past a 95% failure rate. a failure rate so staggering maybe it would be more effective to let nine out of people get on willy nil willy-nilly on the plain with no screening whatsoever and check the tenth. how is such a failure rate possible and what should we do about it. dr. lisa dolev is a ph.d in engineering and veteran of the israeli air force and she's saying take the humans out of the loop with automated scanning
which has better than a 95% effectiveness rate and let me remind you in tests, the tsa allows allows 95% of the bad things through joined by lina rowe of "fortune". >> i saw a man that puts his bag into the capsule and the capsule itself is going in the bag and determining what's in there? >> that's a very small part of what's going on. >> to the english major. that's what's going on basically, and behind that is a whole area of sensors that are connect that are intelligent and machines that are talking to each other. >> are those machines out of sight? not necessarily and not necessarily other elements connected to it and the whole principle is taking a broader look and connecting together and even the human in the loop to make sure that you make this
automatic detection. >> a very good portion is automatic. >> a very good portion. >> let's talk about this should we have humans going through bags or should we have machines going through bags. clearly, you are at the point where technology could allow us that machines computers could better -- anything could be better than a 95% failure rate but that computers are ready for this job? >> yeah. it's like at more than 95%, to be fair being in a security space i'm not privy to how that report was done but something is a little bit of an abnormality for me in that number and that does not -- >> that does seem bizarre? >> that does seem bizarre and it does not represent the whole of what's going on with tsa. >> in these particular tests and let me be clear -- there are decent wonderful, hard-working well-trained people working for
tsa. >> okay when they ran the test that's the number they came up with. okay. now let's -- let me rephrase my question and feel free to jump in any time but let me rephrase my question. are we at the point where no matter what the results are, machines are better than humans at finding the bad stuff? >> no. you should never leave a machine without a human and you should never have a human doing things that machines can do better. it's collaborative. it's one of the things that since our inceptions have put into lace. one of the things do better with today's technology that you can have smart intelligence systems that teach themselves and teach each other and we're interconnected and certain thing, these machines can now do better and a human should never do. there are things that a machine would never do and a human must do them and that's basically the principle between it. looking at a bunch of images in this day and age is really not the correct thing to do but there are certain types of
things that you don't want to teach the machine and certain type of behaviors that you can't teach a machine and you have to put them together. >> do the machines have their own kind of vulnerabilities? you mentioned the machines aren't necessarily on site. are they connected? could they themselves become a target? >> everything can be a target and everything in the world is vulerror inable and that's why you need the best of everything and using everything at your disposal, the data the connectivity, the information, the machine learning and everything that's out there in the valley to use that to the best, but again, people too. people will close vulnerabilities from machines and machines will close vulnerabilities of people and policies and procedure and processes will close vulnerabilities no such thing that's foolproof. >> you mentioned the machine learning and the data and what are some of the ways that these machines get smarter and
specifically, could you give us an example of that? >> an example would be -- in every type of venue or place it's actually a little bit different than other and for instance you're walking through an amusement park and you'll have a bag packed with baby food which is highly organic and if you'll take a metro and deli it's very different to buy content and what a bomb or explosive is it can look very different and a system can learn what is that specific location looking like and then it can ask another machine maybe in brazil and maybe in france who is my partner machine and who looks like what i'm doing because machine learning is based on having more information and better and that's how you can improve and you can cross-information on certain things that may be in one location. there was a higher type threat and they became better at it and now what you're doing in your location that now became a threat for you.
so you want to transfer that. >> you recently tested these out at worldcom the statue of liberty. the idea is people walk through a sensor also and they put their bags in this honey comb thing. do they get it fairly quickly because you have to retrain us the traveling public? >> one of the fundamental number one on our list was to make it very easy intuitive and simple. you don't do security unless you are also handling the issue of your privacy and make it easy and good for the business and it was a deployment of ours and it wasn't just commercial and it was astonishing. if people knew universal, green and red doors and everybody knows what red is they stop green. people were taking selfies around the picture -- around the machine itself which is a bit strange for security i can imagine, people taking selfies around a security machine and
people asking to go through it again and it becomes not just the highest level of security you can provide and it also becomes part of that business going into a stadium or an arena or something that needs to be enjoyable and needs to be easy and quick. >> and it's our job to make it that way with the technology and design. >> i actually was going to ask, not do the people get it but do governments get it particularly when we're talking about security at airports and that sort of thing? it seem by the time technology gets deployed it is behind the times. are you finding that governments in general are receptive to something that is not, you know scanner image based? >> we are. actually, most recently we are and especially with tsa. a few years ago we weren't when we started this seven years ago with a clean piece of paper and wrote our wish list and wound up being at that time way far into
the future in what we were doing, but tsa actually most recently, if you look at some of the public things they're looking for in their proposals have moved up the step to requesting everything on the list and they're looking for the connectivity and looking to become smarter. will they embed it in the near future? governments are very slow and it's because they have processes and procedures in the private sector and other places who can do it immediately and those will be the doctors and they can't, and governments not because they don't like to. >> dr. dolev, i apologize, the machine learning and security system, we appreciate you being with us. >> thank you. up next in silicon valley even the firefighters are working. we'll take a look at that when "press: here" continues.
welcome back to "press: here." you know you're in silicon valley when even the firefighters are creating an app. a carpooling app helps you discover who is going your way. if you're going from san joss toe facebook headquarters each day you might not realize you have a neighbor who is a facebook coworker who would be happy to pay your gas. move spelled m-u-v just like
motley crue. and i'll apologize when we put your company name it will not have the umlat. muv, there are probably people with companies like facebook and google, there are hundreds of people who are in my neighborhood that i had no idea that worked in my office. >> muv is a peer-to-peer ride share platform and what we've done is we've taken an idea and vision that we had and created and we put that into a mobile application. you have companies like uber and they are an on-demand service and we looked at muv and when we say peer-to-peer you can have neighbors or be employee to employee or student to student and in a mobile platform we needed to look at a way to
connect you and as carpooling consumers, we thought what makes it interesting and dynamic. >> is this a one time only thing where once you find somebody in your neighborhood you are permanent carpooling or do you do it on demand i'm ready to leave for work now. who is going my way? >> that's a great question rich. >> we have the ability to do on-demand or schedule in advance and our niche is carpooling and commuting from point a to point b. >> you're not going to make money on this. apparently, you can get a little gas money, but this is not an ubery sort of thing where i'll just give guys people that work at google a ride. >> that's correct. when you go into the true ride sharing model you'll recoup the cost and not make an extra salary on top of your commuting. what we look as an attractable fare that makes it economically available to passengers to afford that trip to and from the
workplace. >> so i heard that you are still an active firefighter and so how do you have the time to have a start-up and also have a full-time job? >> that's a very good question and also balancing home life balance, as well. so with the balance when it comes to it and the person spirit is what drives this and move is a brand new challenge about a year ago with myself and two other challengers and there is so much development and open opportunities out there. it's driving our energy and keeping us up late at night. >> this is a side of silicon valley and he's not the first firefighter entrepreneur we've had on the show and pulse point was one of the other ones where it was not unusual to have -- >> if you're in los angeles, a screenwriter. if you're in silicon valley this is what you do right? >> so barry, i will ask that i was looking up you know you guys seem to be experts in lots
of field and one of the co founders is a philosophy major and this will not lend itself to writing an app. how did you write an app if you did not know how to write an app. in the beginning we had our vision and we came together as three co founders and we had that vision and we are going to go with our gut instinct and push it forward. co-founder stephanie, she actually did a lot of research and we found a mobile app development company based out of sunniville. >> there are people out there who are thinking i have this app in my head and if only i knew how to write it. how easy is it to find an app developer and have them execute your vision? >> it could be challenging. especially in silicon valley there is a lot of competition and a waiting list and we were fortunate enough to find our company in sunnyvale and through research and interviewing other companies along the way is how we were able to rule out some of the challenges of offshoring our technology and keeping it inhouse and we were happy to
keep it here because you're able to go to meetings quickly and you also have a great communication platform. >> when you got interest from the actual companies and schools to maybe partner with you on you know getting the students together and getting people from the same employer? >> yes, you have and the timing right now is impeccable because we just actually became available on the app store and consumer side and we've done a lot of networking and met with high-level sustainable people in the transportation space in the corporate world. >> go ahead, i have the same question. >> speaking of competition, ways which is owned by google has begun offering a similar service and you're a couple of guys in an app development shop how do you go up against going up against google. >> they're operating in tel aviv right now in the first market and they're testing it out and with muv, what we have is a
proof of concept. so we have a lot of test ams out there that say there are some corporations and higher educational institutions that are very interested and i like to use the word salivating for a mobile app to connect peer to peer. and so we look at if google is doing it and there are other companies starting to pop up and 12 months back when people say you are absolutely crazy when doing a ride sharing app, not going to happen and the big players are now doing the same thing. >> and the brave guys and women who run into the fires are the sort of people who are brave enough to take on google. thank you for being with us. >> thank you. up next what does a russian billionaire do with his money? we don't have the russian billionaire, but we do have the answer when "press: here" continues.
welcome back to "press: here," recently the russian billionaire announced he would spend $100 million to help the extraterrestrial intelligence. a underfunded effort to search the skies for alien civilizations, underfunded because there are plenty of ways to spend the money here on earth and underfunded because some people see the whole idea as silly, and some of those people are in congress and people stopped funding it 20 years ago.
recently one of the most famous astronomers shostak made set yi's case. he was sure we would discover extraterrestrial life in our life time and he was talking to congress people who tend to be older and we were talking soon real soon. >> yes we are. i feel like it's not a bad bet to say we are going to find some evidence of beyond earth. >> that would be one of the biggest discoveries of human kind ever ever ever in 20 years. >> it would show earth is not a miracle and we like to think that we are special because our moms have told us we are special and it's only natural and we think we're special and there are a trillion planets in our galaxy and it's hard to believe we're all sterile. >> i know the drake equation and i understand all of this. to be spike, you're not getting 100 million, but as a representist of seti researchers
in general, the science, what does all of this money do? what can they do with this money that they didn't do before? >> you are quite right, scott. this is going to the university of california berkeley. our competition across the bay and what they're going to use it for is spend the money over ten years. it's 10 million a year and they'll buy radio telescope that's just a big antenna. if you put this satellite dish in the backyard the neighbors would complain because it's really big, it's 100 feet 200 feet, 300 feet. so to get some of that by paying for it and what they'll use the money for is to develop new kinds of receivers that can speed up the search by looking at the radio dial at once than has ever been done before and they hope to eavesdrop on e.t. >> how much is uri miller who gave that money, the biggest consumer technology companies of the recent time. how much does technology have to
do with this if we think about apps or websites or things like that? >> it has a lot to do with it. it will be made public and if you will a big data project and maybe you can find e.t. on your smartphone eventually and that kind of thing, but the real point is this what you're trying to do is a technical challenge and we know there are a lot of planets and we don't know where on the dial to look and all sorts of parameters we're looking for and to the extent that you have more compute power at your disposal the whole thing goes faster. and it's like giving captain cook give him a steam ship instead of the sailing ship and he might have found the islands of the south pacific quicker. >> given that you're looking at a needle in the cosmic haystack why are you so confident at this point? >> it's a guess, and i'll bet a cup of coffee would do it and not more than that and only because indeed and it is the technology and its leanest point and the facts are that if you
take sort of the estimates of how many societies might be out there in the milky way that are sending signals that go through our bodies as we sit here and that estimates 10,000 or 1 million depending on who you listen to. if it's any of those numbers we can look at them if there are a million star systems for signals and we've not been able to do that with this investment and new technology we can do that. >> do the math for me again. we need to look at a million star systems to do statistically, probably pick up the thing, the radio signal. we can now do that because we have the funding to do that. >> yeah. we can build the technology to do that. >> is this new technology or is this more of existing technology. >> it isn't just more. we have three computers and if you had 300, it would require new technology and the receiving technology and e.t. didn't say
we have it on the dial we don't know to do that requires specialized receivers and no commercial market for the receivers. >> you're talking about something that hasn't been invened at all. the ideas are not radical and it doesn't require new physics. can you bring those in to telescopes and we will put it on the radio telescope. they'll cart it to a big radio telescope in west virginia called the green bank telescope and that's where it is and very imaginatively named and it's bigger than a football field in size and they'll take it to an antenna in australia which is also large. >> a lot of this has been about kepler, the wonderful space telescope and the lead principal scientist on it finding earthlike planets. you already knew those existed mathematically right? kepler doesn't change the way
you think of anything other than it kind of gets the public's imagination going. >> i wouldn't sell kepler short. >> oh no! it's a miracle! it's a miracle, but it doesn't change where your point in the dish is. in a sense, it doesn't, but on the other hand you tweet the an 10 as at some point, but do you think there is a planet up there, and you wouldn't know and it's not that there are lots of planets out there and what fraction do you have better sorted sort of like the earth and maybe one in ten, one in five one in two, one in three and it's not one in a million. that's the point. just in our milky way galaxy there are 10 billion planets like the earth and that's a lot of real estate. >> one of the things that scott mentioned in the intro is the fact that seti does not get congressional appropriations.
you don't get government money and is there any prospect that that will change at any point? is there any increasing receptivity in the political system to participating in the search? >> wish i could answer that. >> if i could answer questions like that rich i would go to las vegas. >> he's only looking for alien life form in congress. >> it's certainly not ruled out. call me pollyanna, but the chairman of the house, science based science and technology committee lamar smith who invited me to testify, he's very interested. if it were up to me personally i would put it back in the nasa budget but it's not up to him it's up to the committee and i don't know how to predict that. >> do you think it will just be more private individuals like uri or even elon musk and putting money into these initiatives? >> i certainly hope that elon is watching us. yes, i think at least for the foreseeable future it's not going to be government money it
american pharoah has won the triple crown! >> intercepted at the goal line. >> the following is a presentation of nbc sports. welcome to the honda nbc sports desk. >> hi again, everyone. jimmy roberts here. coming up mecum main attractions. first, we start with major league baseball and a big series in the national league east. nationals and mets. nats in first, and mets bottom seven. washington up 2-1.