Skip to main content

tv   Press Here  NBC  January 25, 2015 9:00am-9:31am PST

9:00 am
. >> we will sit down and find out what happens when apple copies we will find out more from the financial times, this week on
9:01 am
"press here." >> good morning, everyone recently two of the smartest people on the planet steven hawking, came to similar conclusion artificial intelligence was something to fear. one donated to safety features. hawking predicts a future where ai realizes it can design itself and it does to the need humanity. if artificial intelligence does take over the field, it will be this researcher's help. he has a doctorate from yale. he created the university of chicago's laboratory. in northwestern he developed science which uses artificial intelligence to write stories about da that. for instance the computer can write a play-by-play of a
9:02 am
baseball game. it's fought so scary. it's fought as if you are funded by the cia. so let's get to that? a second. >> you are funded by the cia? >> only. not the cia. >> some people what you saw, a venture firm that fund companies with strategic interest, whether the cia or intelligence agency has a strategic interest if what they do. >> so i these about that. but is there some reason to fear the movie, we know our series can't understand what it is we want the thing to do should we be fearful? >> i would say ask not be careful for your children. they might grow up and be part iser tan you. and decide that they're going to turn on you violently. >> the odds are pretty slim and
9:03 am
the way you actually avoid that is you tell them not to do that. and you give them an environment if which, in fact it's very hard for them to be back and with ai and ai if plarks i think people don't realize how much of ai there is in their lives already and it's really a bet against ai or argument against ai an argument against intelligence. why would you want to argue against intelligence? >> sewer around the time already. there are lots of companies and ai companies how do you define that? what makes you a company ai you know. >> what do you have to get over? >> the coronation was the modeling or the sim lakes of human behavior that separatings us fromdogs from tree from
9:04 am
rocks. so planning problem solving. language understanding. communication. all of those things and part and parcel of that is reasoning. and to be able to respond to knsz in the world which are both changing and often ambiguous and the reality is is that you might have narrow ai so i would actually arc that you amazon's recommendations is a narrow ai. i would say that eharmony that suggests people you should date that's narrow ai. so there are slices everywhere. and ten there is the push towards a large scale learning systems, which are often seen as ai, but they're actually learning systems. the question is what is the tigs information they're using. the information they will be used for. so these are all narrow pieces
9:05 am
of ai. >> so they only have human-like details. >> that's it. it's almost impossible to answer. but i think that i will see in my lifetime and i'm old and dying. i think i will see in my lifetime systems that can genuinely understand language that can genuinely communicate with people by expressing things in language, which is what i do that can read about the world, learn about the world, and solve problems that we can fought solve. if at that time going to be one monolithic ting i don't know but all of those components will be there. >> thinking of a future where computers cannot write stories but also the front page stories and the ones that require understanding beyond numbers. >> well, in general, we don't do them anymore.
9:06 am
>> that's how you got your start. >> baseball why not? >> for a bunch of reason one we were the stories that we're not working covered. two, we had a moment of an epiphany. a moment of realization that in fact the kind of stories we were writing, that is taking data in the world and recording on that that aside from data driven news stories there is an entire story lovingly referred to as big data and no one understands it and we pretend we understand it and we don't, it generates an understanding and communication off of it so you might want to take tension like a company's
9:07 am
financial report or insurance statistical tables and find the story and see the story in there. your community can do much the same. >> oh, absolutely. we can do it without error, without scale some we do it much faster. we do a commentary, not the most exciting thing in the world. it's tremendously. we have a mutual fund. you have a benchmark. the guy that runs the sign and his team can write that report and its painful. >> i was afraid you were going to say the news covers didn't pay? >> there is that as well. we primarily are focusing in financial services. >> right. >> and it is because the data is there. there is a tradition of
9:08 am
analytics. there are tremendous reportings some of which came naturally him some of which were imposed and they can write checks. they can pay for this stuff that we do we can cover things that nobody else can cover. >> you teach in northwestern. you sent us a lab at university of chicago. is this a place where the chicago could take where is the center of ai? is this chicago in. >> oh no. i think the center of ai is nicely distributed right now. i think there is a tremendous work of ai work going on. pacific, google, facebook there is no way around it. gook him went on a buying spree in terms of a lot of ai in particular from the neuro -- deep learning work but i would say that if you take a look at what's going on in m.i.t. cmu, the stuff at northwestern yes,
9:09 am
there are different branches of ai. so people doing natural language process, understanding systems. that's still fairly well distributed. i think that we have yet to see an industry panel jenl genuinely form around ai. instead, we have silence within the particular company. they have the most powerful kansas in the world him thaifr they're still silent. . >> as ai becomes more pervasive, will we all realize water happening or continue -- >> i think it will be a part of the environment. but we will notice that the things we don't like to do the reasoning we are interested in will be the machines take care of for us. there will be an evolution as it should be because it's fought like tomorrow we will wake up
9:10 am
and there will be a sky net on our front door. but there will be a moment we walk out our front door and a uber central car shows up we will work today. >> it's getting us there. there is great. >> let me point out as a side note, looking back you will have invented it ask but it works. because this robot will come kill you sometime soon. >> oh no besides the fact i think the kind of technology i work on which is explanatory tech knoll, they tell you what is it doing, what it is thinking. what it knows, that technology is one of the bullworks against the machines doing things without 100 scannings. >> that communication layer, we will stick wit. we work on, will be a part of
9:11 am
developing a relationship with a machine. >> thank you for being here. >> thank you. >> up next did you quit your job to blog? how much would you need to make a buck when we continue? .
9:12 am
. >> my next guest is someone you love. they turned into a career
9:13 am
online. michelle pham was a pharmacist bo loved to cook and a pharmacist. so she quit that job and became a full time food bloger. her expertise is paleo. sheing coulds takes pictures and posts paleo is the so-called caveman diet. it's more than diet. it's a lifestyle. with that in mine they say she is something of a martha stewart of pallio. you need a cookbook she has that. non non paleo food for humans. the "wall street journal" calls at this time worst design cookbook of the year. then they rem it as a christmas present as well. i don't think it's a terribly designed cookbook. it's number of cartoons and great pictures. you are a car toon as well. >> yeah. i actually love that that's what they said when they recommended
9:14 am
it. it was sucky yet awesome. >> that could be everybody's headstone, yes? >> and i adore. i was like as long as you've seen it. >> america's testament. who, let's face it. is a bit of a precise fellow? so this is kind of not his speed? i always think he wanted to say like i'm hoping he loves it. it's so. >> he clearly loves it. >> how many hours a week do you work on this? i get the impression s this isn't like in a pajamas and an occasional job? >> no to be perfectly honest it's a joint effort between my husband and i. i always joke we're like
9:15 am
millihave a but i have a total mom and pop orioles. we work on it all the time. he actually has a day job. but we would work on it all the time, on weekends. at night when the kids are asleep and so it's many hours. i don't think as soon as i get out. >> you find out later you are have some search engine authorization or do you create content? >> no. i don't even know how to even make it seo, you know perfect. we can't put in all those plug-ins. it really is hopefully content. >> you really stumbled into this? a stumbleer is not necessarily the best word. it's clearly the way to go. >> right here. >> you at the time didn't know that? >> yeah, yeah there are many
9:16 am
things. i mean there are many things that kind of fell into place. plus like our hard work so it all kind of worked together. >> i think there are a lot of people who know blogging, i understand work friends. i understand all these things they don't get a third of this as you do so it obviously mufb must be about content. >> i hope. so i work hard to make sure rest is work. henry's job is to make sure everything is beautiful and i work on engaging writings so people want to come back and they trust me. i think, really, we've put our heart and soul into this and i know it sound super on tv you be we're successful because we love what we do. >> so how much do you know about your business and what kind
9:17 am
there are and how many there are. >> so we have analytics and they're kind of all over splooiz surprisingly i think, you know the united nations, they are like the bay area austin portland colorado and new york. but then australia is really big. and the u.k. is also very big him so i think it's kind of pockets if places. >> do you have enough to give up your day job? >> i think that's the key. >> how do you make money at it? >> those are all good questions. in terms of numbers, i don't think we ever reached a fair number, eek, okay i'm going to quit. but i think that we had to make sure that we're making money off of it so that i could quit and by 2012 we were making if you have money for sales and afilliates and instead of advertising on the blog so it's
9:18 am
-- >> so like amazon so basically, my restaurant or something that i like and if someone buys it through my link then i get a small percent annual off it. they don't get charged anything represents a higher ticket value you will get a bicker competition. >> right. that's something which i'm very careful about. i only propose things i love and i don't want to be seen as like a -- >> you would guest. >> yeah. no. i mean they're of course if i have these great ticket item you are just careting to people who can afford it and so, you know, i really try to find things that people can afford and i find super useful and i promote it. >> i hate to be i think that's
9:19 am
going to happen. it's not that yours, you won't be honorable. i think at some point people will say oh i remember when you were small. now you are this big industry et cetera et cetera i seen it happen. there is a point, i guess it's a big problem to have. it's so big that your readers begin to see you more of an industry than michelle pham. >> that is potentially a problem. i guess like the mpr model. you put out this free content. once in a while we will create book or our ipad app. if you want to after you test drove everything for free you may want to. . >> we are out of time. paleo, itself, some viewers, if you stick around, we will go to the web. press here, tv.com. we will talk a little more about
9:20 am
that. >> of course. >> in the meantime let's continue our television. up next what if you had an idea for an app that was so good apple took it? more on that when we come back.
9:21 am
. >> welcome back to "press here" we had one person on who said you should never take one to explain the o. that said the
9:22 am
easiest way to explain the app flare, it will send a flare to a friend. you give them a momentary glimpse of your location so instead of texting you are stuck in traffic or shoot a flare from the bar to let your friend know they should join you. the guy that came up with that idea is a former exec at sony. we want to start with the latest ais. i know you will text your location. best said there must have been a certain amount of time darn it? >> if i had a time for every time i had to hear an investor what that meant that apple had included that feature. >> so i'm paying my dime. it's on my dime. what does it mean? >> i think what it does show is annual really believes in location and using location in a different way.
9:23 am
we start out using location five or six years ago with companies like 4 square where you got to check in for badges. now you see if messageing location being used and tagging of the messaging, in the case of ios 8 it's into details and in a shredded conversation. at some point you have to know where somebody is. in flare it's almost an inverse of messaging. we use everything visual in nature. so you begin with a location because you are usually on the road or traveling. so we found the user scapes are particularly different. but it does show you i believe that apple is a big believeer in the location and using it and a shuj supporter of flares so far. >> some of the things has the potential to appeal to a lot of people? >> i do. and i think what interesting is
9:24 am
the original thought or context is for businesses. the way it came about. in fly wheel i broke down on highway 10 and my truck stopped. we called aaa and we waited and eventually i figured out i didn't know if it was going to come or fought. the one thing that crossed my mind why don't i know where that truck is? i didn't know what to do. if i should get out of my car or not that was the point where i was going to leave the company i was at and build a really a platform for user location and panelry that let people communicate. >> you talk about the location trends. the ought thing it reminded me of is the policity and contexttural trend we saw last summer, if anyone remembers.
9:25 am
that's right. >> afc an app that had a great moment you know, we didn't know what the content of the message was, people had their own kind of signs. right? what do you think of the futures contesting that? >> i think what's going to happen, you see more apps operating. it's operating in another station and it showed you as long as you are interested in context, it was usually in the case of friends, it could be powerful operating on a layer you will see officially. you might get a yeo and fought getting a yeo. that's the problem. i think flares is really powerful. we found it most powerful in families. you are looking where is your sweet spot? we have been out only a few months. it could be kye, super viral t. next most viral groups may be
9:26 am
millennials and we found it's incredibly strong they are constantly trying to draft it. >> you say viral t. location-based app is going to be viral because i send a flare to you. let say you don't have a flare. you whether get some sort of notification that hey, you should download this. in other words, everyone whops to use sit sending out to friend who will say, oh sure i'll give that a try. it must be a cool app. >> that's right. >> they have a power that other apps couldn't have. a video player or it's fought forcing the next guy to download the app? >> that's right. i think in terms of emotional, how are different apps or products viral? you have to have emotional residents messaging does it partly because if you really like what it essentially can do with that app. there are millions of them. a lot are incredibly successful.
9:27 am
you have to pair it. it's immediately viral. in our case, it's usually around the table on the flare immediately. like with kids it's a little different. teenagers may open the app. the first thing they do is see if anyone else is on it. they often won't use it because tear friends aren't on it immediately. >> hess the head of flares. it is usually great fun. >> we will be back in just a moment
9:28 am
moment. >> that's our show for this week. thanks to my guests. thank you for making us a part of your sunday morning.
9:29 am
9:30 am
>> hello, welcome to communidad de valle. today we have a newly elected councilwoman and cash for college for you. >> nbc bay area presents communidadcommune del is valle. >> man was burglarized and he was robbed of sentimental value. he is a d-day survivor. not a day goes by he doesn't relive the horrors of d-day. >> you wonder it's scary.

301 Views

info Stream Only

Uploaded by TV Archive on