and are the algorithms being tested for bias. we've had a hard time finding studies and we interviewed companies about the issue and neither of the companies could point to a specific test for racial bias. and finally where does the bias come from? you could guess the demographic for the engineers, if they are testing on themselves i'm skeptical, the complication used for the training and some people are harder to recognize which certain studies have speculated and one company said this make be the color contrast in the face being reduced for people with darker skin making it harder to distinguish features. so i'll hand it back over to clair to talk more about the study. >> to sum up, facial recognition is beginning to challenge our assumptions or expectations of anonymity in public spaces. and while it has positive implications, it is a powerful policing tool and critical to understand the risks that it poses, particularly if it has this risk, which we're finding in our research to disproportionately impact, particularly african

race is removed from all of the algorithms. the results has impacts on racial justice, communities of color. and that reason is simple and elemental, the police data comes from police. and all of the implicit and explicit biases policing in america today show up in that data. in this way, predictive policing may not be any worse than traditional policing the question is whether it could be better. want to sketch out the reality, scope of objective policing, give you terms and we will discuss it in the question and answer. to give you the sense of scale, predictive policing is only used in major cities like new york, currently being used in major cities like new york, philadelphia, seattle, kansas, and a whole host of smaller cities. there are three main types of predictive policing, ideas that you can predict where burglaries might be, where car thefts, thefts from autos. you have your violence, gang shootings, retaliatory shootings in bars and clubs, and you have your person-based, we can identify the people involved in crimin

in 20 they took the pool of algorithms and separate them into two groups. those developing east asian those that develop in western europe and the united states. they want to see a packet these were on pictures of east asians and pages of caucasians. take a look at the graph on the left and juicy black line higher is better, more accurate. the black line represents the east asian algorithms and they're more accurate on east asian faces. the graph on the right the opposite was to get on pictures of caucasians the algorithm developed in western countries was better than the algorithm developing east asian countries. this is a well documented phenomena. we call this the other race affect what people tend to be better at distinguishing members of the own respect transit of algorithms have a similar issue. this is setting in which an algorithm is develop can't affect its biases. knowing the american software development industry tends to look a lot like me, asian, male, probably in 20 or 30. this raises concerns the effect this technology is more likely to be em

we don't need big data and new algorithm to be able to figure it out. the question is why and to whose profit are we using these tools to answer questions we already have answers too. >> as you noted, there are certainly some benefits to whether it is the health context or understanding someone's medical history, someone's, you know, education needs and those sorts of things. i do think a lot of times the discussion is cast in the negative and i think sighting the benefit from thinking about the potential uses of big data for the science benefit overall. >> in the corrections system, sometimes very similar algorithms or significant algorithms are used for something called needs assessments. it is basically a purpose in which, you know, essentially the profile of a person is used to match them to surfaces they might need. now, you know, my problem with that is the insight the data tells us in this case are things that have been known you don't actually need to stay up a number score on a particular individual to know, right? we know -- if it is true, for

if you have algorithm or a police department has an algorithm that makes predictions correlated with rates, supposing for a second they are actually accurate is that category something that's buys aed or not, according to your metrics? what metrics do you use? >> that's a good question. i think one of the things underlying in your question is sort of this notion that use of race should be a factor and there's been policies for many years on the use of race was in some ways a misnomer because it's intended to not use race as basis for any particularized suspicion on an individual. i think one of the things in the department varies parts of the department there's the office of justice programs and bja and cops and those entities that give out grants. but they are interested in training and working in the same way that they do to deploy these technologies working especially in the aftermath of ferguson with police departments to specifically understand and identify that some of these officers whether or not they do it it intentionally or not, may have preconceived notions about particul

algorithmically, what do you want? you algorithmically doing things to encourage more personal sharing? it is like an existential problem, because you actually have power over what people see. algorithmically, how are you handling that? mike: there is lots to do, all the time. you measure the results of these and all sorts of ways, including user settings, where he asked people if they like the stories they saw, do you want to see these things instead? --t of the goal is to get give people what they really want. here are all the possible things you want to read, what are the things you actually want to see? products't build people like, it is a competitive market, and people will see other things. emily: doesn't that change all the time? mike: people did not know they wanted video and mobile, and now they do. you're seeing immersive the video and the are -- vr, something we could have only dreamed of. emily: do you worry that this could hurt personal sharing, and what makes facebook rate? mike: i don't think so. to wat

algorithmically, what you want? are you algorithmically doing things to encourage personal sharing? like an essential problem, because you have power over what people see. algorithmically, how are you handling that power? mark: i think there is a lot that we do in newsfeeds all of the time. we measure these results in different ways, including user studies. that is where we ask people, do do like the stories that you sell? would you like to see these things instead? the end goal is to get to the perfect mix of what people really want. if i could show you, here are all the possible things you could read, what do you want to see? that is the end goal. if we do not build products that people like, then they will go use other things. emily: what people want -- does that change all of the time? mark: people do not know they wanted video on mobile five years ago, and now you can. we are seeing the rise of vr, that was something we could only dream of years ago. now, you can do that, dol. too.ou can do that, emily: are you worried that the rise and broadcast -- in broadcast could hurt pers

algorithmically, how are you handling that?e: there is lots that we do with news feed all the time. where we askes people, did you like the stories you saw, would you like to see these things instead? part of the goal is to give people what they really want. i said on stage here are all the , possible things you want to read, what are the things you actually want to see? if we don't build products that is the end goal. people like, it is a competitive market, and people will go use other things. emily: doesn't that change all the time? mike: things change. people did not know they wanted video and mobile, and now they do. we are seeing the rise of immersive video and vr, something we could have only dreamed of. emily: do you worry that this could hurt personal sharing, and what makes facebook great? mike: i don't think so. you want to watch something on tv, or go to a concert, humans are pretty natural about finding different groups they want to interact with. i think we can build products that hit all of those interactions. em

>> reporter: the latest machines use ct scans borrowed from the medfield combined with complex algorithms to examine 1,000 bags an hour. the technology is so good it can map out the densities of various materials in 3d images. the result is a very high resolution 3d image that the algorithm categorizes by density and then it compares it tablier of known threats and alerts the technician if there's a problem. the days where a worker has to stop and open a bag to physically inspect it are mostly gone. the ceo says as airports around the world intercept luggage bombs they can add images to the threat library and update the databases around the world. >> the bad guys are always doing something new so we need to keep up with the intelligence. in a developing story: todaa hearing happens on whether allow live music at the army in san francisco's mission district. the building at 14th and min streets originally housed weapons for the national gu now it's owned by two men w stage electronic dance music events. and they want to hold frequent concerts with promoter "live nation." but neighbors say

whose results but to he can predict. >> when an employee figures out a way to hack the random algorithm and push them on to relatives multiple times is ingenious. >> and criminal. >> the two brothers out of jail and you have to wonder if there is more incidents in other places, crazy. >> you credit than pulling the track back up with a chain and pulling the atm machine. >> and we are done with the thunderstorms and light showers that col roll over nats park and will not prevent them from playing. we did go for a high of 65. despite the clouds and the showers and we'll let you know how we did at 11. live look outside. it is our live cam. beautiful evening. a couple of sprinkles southwest and down to 53 and winds out of the southwest at 7. they will pick up and turn more westerly tonight and tomorrow. here is the radar over the past hour. a couple more showers had developed on the east side of i- 81 between haggerstown and winchester. these are the heaviest showers into southern maryland moving to calvert county and we have showers south of 66, the west of colpepper and we will zoom in. r

reporter: is that down to the technology, the algorithms to be able to help personalization that muche actually making significant investments in artificial intelligence to make the algorithms even smarter, to determine what content should we serve up. at any given moment, there is literally over 1000 possibilities to serve anyone who is pulling up their facebook or instagram newsfeed. what the algorithm is trying to do, and a lot of it is being done in conjunction with our ai efforts, is to determine what is the best content we conserve up, be it from friends, family, or brands. it is fair to say it has been a busy week with lots of earnings news. here to break it down, editor at large cory johnson. start with amazon. you remember it all by heart, right? let's start with amazon. numbers weres impressive. we had two companies reporting this week firing on all cylinders, going through big changes with their business model. amazon was one of them. that topline number re-acceleration of revenue growth from amazon, into one of the biggest quarters they have ever recorded, the first -- big

suggest on math requirements we put in calculus when we were loosing foot neck and it is great but algorithms should be part of the math. >> we require four years of math. doesn't have to include calculus. you don't have to have any particular score. but statistically the single strongest correlation to success was four years of math. something about the discipline. so kids could dial it up if they are not there already and you might see improvements. >> the gentlemen here and then here. that way keep the microphone here and we will pass it around. >> thank you. this is for president daniels. i am sorry. i am nick farmer. >> if i can interrut, i think president daniels has a nice ring to it. -- interupt -- >> [laughter] >> what percentage of your students are u.s. citizens? and what percentage of your graduates are working into the united states? how does this inform your thinking about policies that are working and not working in the country today? >> 85% of your undergraduates and only 55% of your graduate students are domestic. we think -- i will talk about the undergraduates. we have 30,0

. >> we can assure it's because an algorithm. algorithm. >> reporter: download the app and they'll send a genie your way. you can call them directly, too. the idea for genie mates came from their own struggle to work, take care of their three sons and keep their house in order. >> we wanted to free up our time. >> it's important. we like to play basketball. we play with our sons. we're very active. we work out. we're not the type that want to stay in cleaning for three hours. shoestring budget, less than $500, and kept it hyper local, testing their cleaning service and algorithm on family and friends in the brooklyn area, doing a lot of the work themselves. they now use independent contractors in the five boroughs and nassau county. scrubbing bathrooms on demand and offering competitive pricing. >> our basic cleaning is a minimum of two hours and $40 an hour for the service. that service is basic dusting, wiping, cleaning, mopping, what's already expected in your normal cleaning. >> reporter: need extra elbow grease? not a proble

the codes directed the computer numbers, but instead, to use an algorithm whose results could be predicted. eddie tipton, a former security director at the multistate lottery association, had helped to build the computer. >> you know, the employee gets a security job to help protect a lottery system, figures out a way to hack the random algorithm, so that he can win the lottery, and he can pass those winnings on to his relatives, multiple times, it is pretty in ginous if you think about it. >> eddie tipton's lawyer told cbs this morning that his client is confident he will eventually be exonerated while tommy tipton's attorney says his client will plead not guilty. josh elliot, cbs news, new york. >>> now police are continuing to investigate those brothers who are right now out on bond. they suspect there could be even more instances of this alleged fraud. >>> still ahead on the news, it is just business, right? coming up, why presidential front-runners from both sides are taking their defeats in wisi refugee using a baby to explain how dire their situation is? yvette: i was running for my

i think there is a problem with the allegor algorithms that people don't know. we don't have control over what is being shown to us. the computers are doing that. >> jane newton, tell me, what do you see as the role of what we'll call new media now, and how it can perhaps become better? >> it's a challenge. microsoft did an experiment with artificial challenge where they created a robot that existed on twitter, and it responded to what people tweeted at it. and it became a racist sexist creep because that's what twitter is, and that's what the internet is, it's full of trolls and horrible people. if you let ratings drive news and let ratings determine what people consume, you're going to get not much more than puppies, kittens, rainbows and donald trump. that's a huge problem and it's leading to clearly a bigger problem in politics, which is a whole other story. there is not a lot out there that can self create for this in media. there isn't enough that creates balance on both sides of the story unless people self select to go out and find it themselves. >> we a

for now, let's say they do moved to this algorithm style. they say is to improve your experience and show you what you're most likely to like. the content, based on your inner josh your content and interest. there are definitely ulterior motives. change into the algorithm, the precursor to this system like what happened on facebook. this is an easy way for instagram to make money off of brands and businesses. they pay for their facebook post to be delivered to more users. all the marketers and retailers in anyone selling anything who dishes are content on instagram could lose the free right here. their content will get lost. when you think of it that way, it will be hard for instagram to avoid switching to this style. they will have to find different ways to make money. for now, it's not happening -- happening. >>> a teenager in kansas city is getting it just giving a generous gift to you and cancer patient. >> the 17-year-old handed over three tickets worth $900 for the kansas city royals home opener tonight. here is the story. >> it starte

. >> i suggest to we put in calculus, algorithms and algorithmic thinking. >> doesn't have to include calculus. you have to have any particular scores. statistically many of the scholars will notice. the single strongest correlate, four years of math. one simple thing is dial that up they are not there already. >> keep the microphone here. >> thank you. this is for president daniels. >> next line, independent citizen. >> behave yourself. what percentage of your students are us citizens and what percentage of your graduates end up working in the united states? how is this improve your thinking how policies that are working or not working in the country today? 85 percent of undergraduates and only 55 percent of graduate students are domestic. and we think we have -- talk about the undergraduates, 30,000 on the main campus. the rn about the right balance. one of the most diverse universities, 15 and in recent years of low iron that. the 1st foreign trip you take, make the effort on this campus. so we talk about that a lot. i think we had a recently good balance. i understand it is held h

natural language processing, reasoning algorithms, statistical parsing.easoning algorithms, man 1:man 2: i am.ng. woman: ex-military? man 2: four tours. woman: you worked with computers? man 2: that's classified, ma'am. man 1: but you're job was network security? man 2: that's classified, sir. woman: let's cut to the chase, here... man 1: what's you're assessment of our security? man 2: [ gasps ] porous. woman: porous? man 2: the old solutions aren't working. man 2: the world has changed. man 1: meaning? man 2: it's not just security. it's defense. it's not just security. it's defense. bae systems. he is a 31-year-old belgium criminal. a u.s. official telling dow jones authorities believe he may be the third march brussels suspect in that airport attack. four men who are charged with having enlisted to isis appearing in denmark. three of the four denying the charges. the four men are all in their 20s. they were arrested in a police raid yesterday. a fire that ignited this morning at the refinery in houston has been put out. the plant refines nearly 265,00

>> the algorithms, we don't see everything that's there, twitter and facebook, you're not going to seeverything your friends post. it is going to show you the things you like. what you fall into a trap of, only the things of interest or the things you like. jane talked about this earlier. you're not expanding, not getting a verse style view of what's going on in the news. i think there's a problem with the algorithms people are not aware ever what's going on, we don't have enough control over what's being shown to us. the computers are doing that. >> jane newton, and ms. small, what do you see as the role of new media now and how it can perhaps become better? >> well, it's a really big challenge. microsoft interestingly did this experiment with social media, within 24 hours it became a racist sexist creep because that is frankly what twitter is and that's what the internet is full of trolls and horrible people. and if you let ratings drive influences, if you let ratings determine what people consume you are going to get not much more than pullies and kittens and rainbows and donald tru

>> i cannot see disney moving into the algorithm business and fortunately the main consumers are peoplewe are story telling animals. that's what disney is great at. you would absolutely want your ceo to have a real appreciation of that. that said, bob iger is great. i'm not sure how much he contributed on the set of star wars. he knew how who to hire. >> abc background. >> but there's great creative people out there. what you need is a leader. bob iger is known for this. appreciating that in a way that may have brought steve jobs and pixar into the family. >> well they don't have problems any time soon on that front. >> there's lite of runway there. it's good to see you as always. >> meanwhile the markets have given up wednesday's rally and then some. the dow is down 173 points. major averages down for their third day in four. the s&p is down by 1%. the nasdaq by 1.3%. apple shares falling after btig cut it's price target and it's earnings estimates. the firm no longer expects apple to return to revenue growth this fiscal year. that's a bold call and taking the stock down by 1.5%. mean

reasoning algorithms, statistical parsing.ow you are just showing off. great time for a shiny floor wax, no? reasoning algorithms, statistical parsing. not if you just put the finishing touches on your latest masterpiece. timing's important. comcast business knows that. that's why you can schedule an installation at a time that works for you. even late at night, or on the weekend, if that's what you need. because you have enough to worry about. i did not see that coming. don't deal with disruptions. get better internet installed on your schedule. comcast business. built for business. >>> good morning. yesterday jpmorgan lit a fire under the dow. more financial earnings on tap today. bank of america and blackrock set to report. this hour we'll bring you the numbers and reaction from the street. >>> meanwhile, regulators setting their sights on theranos, proposing a two-year ban for the blood testing startup, including its founder, elizabeth holmes, a two-year ban. >>> and passing the ball was not in the game plan for kobe bryan

so although this is now based on algorithms and presumably, the whips on the flora can get the floor can get the information at particular delegates fast and cater their pitch. host: this will be a first for a national convention. any precedent for this on the state level? evan: there have been technology used at the state level used successfully in virginia with gillespie and in utah for senator donnelly, senator lee. panicthe technology was so much more primitive. what they were basically trying to do was get a mobile app corralr where they can all the information. they did not have the resources to engage in data mining. there is exponentially more money. the stakes are higher. there are all kinds of things that will come into play in cleveland that have not at these state-level conventions. host: it will not be a brokered convention because they will be there are no brokers. it will be a contested convention. evan: i cannot promise no smoke-filled rooms. i am sure there will be a lot of lobbying and persuading just because there is all this technology. when it comes down to persu

it uses an algorithm. according to the software, he has a 97% chance of kicking the bucket.f the dragons, gets a 95% chance of saying the big old bite bite. i got all the names right. >> are you a "game of thrones" fan? >> no, i don't want it. >> how dare you be different. >> i'm moving out of movies before you chew me out. >>> charlilate show, stephen co gave a tribute to the late great prince. >> beautiful artist, not many people as influential to be known by one name, let alone by one color. but he was. so in a small gesture of respect tonight, we would like to change the set in his honor. >> that was beautiful. >> a nice touch. we were talking earlier like itunes and everything, someone was saying 18 of the top 20 songs were all prince songs. >> all about prince. >>> i'm alex witt, and this is "early today." ♪ ♪ headache? motrin helps you be an unstoppable, let's-rock-this-concert- like-it's-1999 kind of mom. back pain? motrin helps you be the side-planking, keeping-up-with- your-girlfriend- even-though-you'll-feel-it- later kind of woman you are. body pain? motrin helps

global kinetics' algorithm processes the data and presents it in a series of graphs. >> you take it at 7:00. that's these bold red lines. >> reporter: the graphs detail all of harrington's movements over six days. today she and the doctor can see which were periods of exercise -- >> you were probably at a gym working out. >> reporter: and which were unwanted body movements. and at one times she wasn't moving at all. >> maybe you're sitting down reading. >> reporter: dr. dahl decides to keep the medication doses as they are for now, and to consider other options down the road. he says he can see more cleary how his patients are doing once they leave the exam room. >> it really provides us another window in to what your experiences are on a daily basis. >> reporter: harrington has learned something too. >> i was surprised that it is showing my lifestyle has gotten slower in the afternoon and evenings. >> and she's feeling better about the future. >> it creates an optimistic feeling for me. >> reporter: elizabeth cook, kpix 5. >> by the way, global kinetics is making the pkg available to

they're algorithms. so defend your homefield advantage and argue that in accurate law which is still adjudicated so far by human judges. and then you stand a chance. this is really all we're doing. then you stand a chance of fighting against the regulation that prevents people from engaging in commerce on equal or better footing than the corporations that have usurped it. >> i'm happy to talk about it. a lot of good examples to learn from that a lot of -- how you manage the partnerships to allow it to be owned by the community. with that i think we are going to close. which means everyone can come ask questions individually. >> i'll be here. >> or talk to me about municipal fiber. thank you. >> thank you. thank you so much. [applause] [inaudible conversations] >>> here's a look at some books being published this week. josh king former direct record of production for presidential events at the white house recalls some of the most unsuccessful examples of political stagecraft in american history. and recal

an algorithm scans for recurring purchases, flags them, then helps you cancel what you no longer want. truebill says it never shares data with third parties and has read-only access so nothing can be changed on your account. for now, the service is free. >> how did you feel when you realized you were leaving $3,600 a year on the table in forgotten expenses? >> i felt bad about that. >> reporter: now he's got a subscription for savings too. olivia sterns, nbc news, georgia. >> things do add up. >>> we're back in a moment with how fast food workers were tricked into vandalizing their own restaurant. >>> days after bruce springsteen canceled a north carolina concert in protest of that state's so-called bathroom law, what he and others have called anti-lgbt bigotry, canadian rocker bryan adams is following suit further south. adams has canceled an upcoming show in mississippi, citing that state's new law that allows religious groups and some businesses to refuse service to gay couples. the mississippi law will take effect july 1st. >>> caught on camera, panicked workers at a minnesota bur

sir avery: i've done the algorithm analysis of somebody my age born at my age in my time, so i know whatth the next 10 years of my life. failure is really just for people who are not prepared. you don't waste time on things that aren't going to get you to where you need to go, things like clothes. ♪ ashlee: sir avery is well known for taking complicated medical devices and making them simple, cheap, and durable. did you make this in here? sir avery: yeah, yeah. ashlee: this is your workshop/lab. sir avery: yeah. ashlee: currently, he is working on this incubator. sir avery: it is almost like something that superman would be happy to be in, and the shape of it is much more about getting even air distribution than trying to make it look cool. ashlee: the life pod costs $2000 versus $40,000 for typical incubators. the hope is that it can save thousands of lives each year. sir avery: i will show you, this is really cool. this is pretty ugly. ashlee: he also has this giant block of frozen chicken slurry as a way to amplify the nutritional value of trash meat. sir avery: after kentucky fried h

in fact, the dying thing is such a "game of thrones" constant that fans created a computer algorithm ice and data" that calculates how likely characters are to die. ooh, ooh, do me, do me!

. >> the key to the system is that all of the price changes are determined by algorithms died into the point of sale system. it leads customers to pick the discounted drinks. >> they make a high profit margin on tequilla, so peel will start buying tequilla. we created a system for bar owners to get the most out of special products. >> after examining the first few months of sales records, they discovered something odd. >> patrons purchase items not part of -- that are part of the drink exchange, and ignore everything else. >> temps and freelancers could be exactly what your small business needs. one, upwork. post a description of what you're looking for and candidates can apply for the job. try to include specific instructions. two. wonolo will put you in contact with immediate hourly or daily workers for times you could use extra help on sight. shift gig is a great resource to find service industry professionals. fourd, thumbtack is location driven. and five, reach out to local universities. they may have students interested in your field and eager to work. you can get affordable help

an algorithm can't do that.w do you see the hierarchy between apple music and spotify and tidal, like how does that play out? troy: so, you have terrestrial radio that is stronger than ever right now. this is another one that will not be a winner-take-all market. it is very, very complicated. the streaming market is going to be very, very competitive. emily: what about tidal? troy: i think the verdict is still out. i think the intent was fantastic. -- i want artists to own something. by the way, i would rather see something that is artist-owned, than something that is venture-backed. you know, so i think the statement was powerful. and you can't underestimate jay-z. emily: artists are now making the bulk of their money from live events and tours. right? do you foresee that changing? troy: there will be multiple revenue streams. i think we're going to see revenue streams that don't exist, right? emily: like? troy: ai and vr. like when artists can scale themselves by doing -- instead of having to go out and beat y

it uses an algorithm to predict the show's next victim. here is what the software found. the king has a 97 chance of kicking the bucket. queen of drag gets 95% chance of saying good-bye. >> that is almost everyone! >> i know. really big characters. >> name me one big character. not all of them. >> you never know. >>> if you're heading to the theaterses this weekend, . >>> on the late show, stephen colbert gave a tribute to the late, great prince. >> besiding being a beautifule artist but not many people are influential enough to be known by one name, let alone by one color, but he was. so in a small gesture of respect tonight, we would like to change the set in his honor. >> i'm betty nguyen. this is "early today." this is "early today." oh, yeah baby you're much too fast t headache? motrin helps you be an unstoppa ble, let's-rock-this-concert- like-it's-1999 kind of m om. back pain? motrin helps you be the side-planki ng, keeping-up-with- your-girlfri end- even-though-you'll-feel-it- later kind of woman you are. body pain? motrin helps you be an unstoppa ble, i-can-total

i did the exercises if we took the algorithm and applied absolute equality to all 3 programs. it was a slight adjustment that the city planned in this exercise. so, how does this happen? well, we have lots of tables. >>i have one question. just a moment i will take the floor. if you would slow down in your discussion of this material number 1 and number 2 if the commissioners have questions if you get my attention to the chair. >>yes sir. >>the three-year plan does not describe the current rates, right? >>that could add bigger amounts of money that we are discussing. we have disclosed at the february 11 and since we do not want to say anything about what the potential rate increases would be if it was okay with the board they said use 2016 rates to demonstrate the impact and therefore that's why was done this way. >>commissioner breslin. >>there is an increase in the city plan i mean it's not a huge increase but it is still an increase. >>yes there would be an adjustment but we will do the exact same thing. i will go slower i'm sorry about that. please turn to page 18. what yo

who they he wants to push his algorithm on to whatever platform he uses.don't support that yet. of someonethe realm who has a day job and wants to put their money into use in the market. carol: time will tell how successful robin hood will be. they have gone mobile with this. >> it keeps them clean and simple. you wake up, you get on your phone, and it feels like your money is in your hand is how he wanted to present it to customers. carol: photographs say a lot about the people the magazine is profiling. tell us what is involved in doing something. you have this one, build a chair in three minutes. the visuals are compelling. , they are trying to take a regular chair and make fits in a shipping box and arrives and there are no tools. it is supposed to come together in three minutes. myself, tracy, and a few other members of our team got a studio in brooklyn and assembled the chair. carol: how many takes did it take? >> only four. carol: pick another one. i like this one, the architect of highline. what i thought was interesting he was talking about was ho

reporter: the latest machines use ct scans borrowed from the medical field combined about complex algorithms to expect 1,000 bags an hour. it can map out the densities of materials in 3d images. the result is a high resolution 3d image that they categorize by density and then compared to a library of known threats and then alerts the technician if there's a problem. >> the bad guys are always doing something new. so we need to keep up with what the intelligence is saying in regard to what they are trying to do next and stay one step ahead. >> reporter: in newark, kiet do, kpix 5. >>> we are getting our first look at two men who police say stole a scooter and other items from a san francisco make-a- wish office. 37-year-old and 26-year-old men are behind bars at the san francisco county jail for burglary and other charges. officers are still looking for a third suspect. over the weekend, security cameras were rolling as thieves swiped laptops, electronics and a scooter signed by giants outfielder hunter pence. mayor lee announced this morning that scooter has been recovered. >> this is on the

google's algorithms can -- the calendar is going to tell you when to do these things. >> right. >> this calendar was designed to compete with apple, microsoft and facebook getting into, darlene, the life management business. your phone is going to tell you, calm your mother. >> telling me what to do. that's what i need. one more. >>> it's 4:51. still to come, what you don't want do at a car wash that could cause an explosion. >>> you may not be able to get ats to hamilton, but hundreds of local students have still the hottest tickets in town today >>> a car wash owner in dallas admits he's neverieeen anything like this. a vac ul exploded in flames. he was trying to suck up spilled gasoline from the trunk. that's a no no. the driver doused the flames with water. that driver could have been electrocuted. who knows what could have happened. luckily, no one was injured. >> you can't be doing that. >> sucking up gasoline with a vacuum, an electric -- >> probably not good idea. >> we're all learning. before. now i know. don't vacuum it. we have a nice a day ahead. it's going to be sunny and d

google has massive compute power in the cloud and all kinds of algorithms and research around where weeaded. it is a huge part of server technology and software that we would need to be able to pull it off. emily: you continue to run nest as a semi-independent company. how has your role changed? how have the goals of nest changed since this transition? tony: more, faster. that is it. we laid out a roadmap, a two to three-year roadmap. we saw eye-to-eye on all of this stuff. larry just said go implement it as fast as you possibly can. it is like no, i'm not going to go changing it. we believe in this, go. emily: how do you think about the next new product. how do you decide what the next new product will ultimately be? tony: sure. i think that, first, we each, every day, run into frustrations. things around the home, why is it that way? the next thing, piece is, hey, why don't you guys make your roadmap and i will make my roadmap, so we have, like, five or 10 different groups making roadmaps of what they would think it would be the next two years and then we compare notes. emily: would

are you algorithmically doing things to encourage personal sharing question mark it is like an essentialblem because you have power over what people see. how are you handling that power? >> i think there is a lot that we do in newsfeed all of the time. we measure these results in different ways, including user studies. the end goal is to get to the person -- the perfect mix of what people really want. if i could show you, here are all the possible things you could read, what the you want to see? that is the end goal. want -- doeseople that change all of the time? >> people do not know they wanted video on mobile five years ago, and now you can. we are seeing the rise of vr and now you can do that too. emily: are you worried that the rise could hurt personal sharing or hurt what makes facebook great? >> i think we all have different modes we operate in. you want to chat with friends, watch something on tv, go to a concert. humans are natural at finding the groups they want to interact with and i think we can create products that hit those interactions. let's talk about ai, what is the hol

reporter: here's how it works, ive truebill access to your access or cendit card statements, an algorithm flags recurring purchases and helping you cancel what you no longer want. truebill says it never shar s dataker h third parties and has read-only access to nothing can be changed on your account. for now, the service is free. how did you feel when you realized you were leavin $30 on the year on the table in forgotten expenses. >> i felt bad about that. >> reporter: now he's got a subscription for savings, too. olivia sterns, nbc news, georgia. >> things do add up, don't they? >>> we're back in a mo whoa. what's going on here? oh hey allison. i'm val, the orange money retirement squirrel from voya. val from voya? yeah, val from voya. quick question, what are voya we're putting away acorns. u knhe, to show the importance of saving for the future. so you're sort of like a spokes person? no, i'm more like a metaphor. okay, a spokes-metaphor. n tho, i'm... you're a spokes-metaphor. yeah. ok. see how voya can help you get organized at voya.com. (brian)i'm briiean. i was in the military for

. >> the key to the system is all the price changes are determined by algorithms.h realtime sales data driving price changes, the game often leads customers to choose the bargain drinks, whose prices may drop but whose profit margins remain high. >> it will drop the price of say a tequila shot. they make a high profit margin on tequila. all of a sudden people start buying tequila. we've created a system for bar owners to get the most out of profk products. >> and that isn't the only way this game can influence customer choices. after examining the first few months of sales records, todd and frank discovered something odd. >> patrons tend to purchase items that are part of the drink exchange and ignore everything else. >>> maybe you're not ready to hire a full-time employee yet or maybe you only need a couple hours of help. if so, temps and free lancers could be exactly what you need. smallbiztrends.com gives us five ideas of how to hire temporary employees and contractors. one, upwork. post a description of what you're looking for on the site and candidates can app

i do not think the nuclear scenario, forcing a big change to the search algorithm is likely. think google would rather just a lot of the eu then give away their secrets. i imagine they have a compromise somewhere in between. in some respects, google is a victim of its own success. share on the smartphone in europe is to blame for where they are. they have to be very careful with how they manage business. mark: thank you. sebastian of robert w bear and co-. coming up, what wall street wants to know about the young prince reshaping the saudi arabia and economy. but as next, right now, mark carney's "it new york at the united nations. he is his case for a new financial system that is needed to deliver on development goals. we will continue to monitor those headlines. you can watch that unfold on the bloomberg at live/go. ♪ mark: we take a look at the biggest business stories in the news right now. they filed for chapter 11 bankrupt. it was down $3.1 billion on paper. it also missed deadlines, raising questions about the quality of the company and send investors running for the e

. >> ensure that a cleaning professional will be there in an hour, because of our algorithm, we had matched client within 5 mile haze outof a cleaning professional. >> download the app, send a text they will send a jeany on the way. -- jeany on the way. >> hello. >> idea for jeanie mates came from struggle too work, take care of their sons, and keep their house in order. >> we just wanted to free up our time. >> our time is important, we like to play basketball, with our sons, we're active. we're not the type that' to stay in the house cleaning for 3 hours. >> steve and sade started on a shoe string budget, kept is hyperlocal testing their cleaning service on family and friends in brooklyn area. doing a lots of the work themselves they now use independent contractors. >> our basic cleaning minimum of 2 hours, it is $40% hour, that is basic dusting, and mopping. reporter: need streak? >> a deep clean? not a problem. >> time too go out, this service is also giving a lot of families back their time, a lot of busy professionals back your time. >> jeanie mates are running special through mother'

i tweeted yesterday that because the algorithm had changed on twitter you also had fictitious moves one people tweet things that came out two to three weeks ago. how do you deal with that? howard: i don't. i don't use twitter for that, that is why we started stocked which. i need a couple hundred thousand people with my list of 40 to go to. since the days of area well, i think the world has been about smaller groups of chat. bloomberg is smaller in a sense because you pay your money and you are in the network. it is not how many you follow, it is who you follow. i do not think i ever heard of this until yesterday and i don't believe it is true. machines are trading up, that is old news. i don't believe this is happening. tom: where is marissa mayer's next job? what is the -- where is the best place for her to go? howard: alaska state governor. tom: be nice. howard: i am not impressed. from the first hire on down, she did the alibaba thing. i'm not impressed. the site stinks. they let the core assets die. i have not been forthcoming with the real plan. the foray into digital magazines w

necessary to be protected against attackers, and it automatically encodes every message with an algorithmthe sender and the recipient. not even the employees will be able to look at the customer's data. the company will have no way to apply with a court order that apple received in connection with the san bernardino, the shooters phone. the fbi's top lawyer says that the bureau has extracted information from the phone and it is putting it to use but no other details have been explained on what kind of information was on the phone. they want to find out where they went after the shooting. they want to know if the couple had any help from outside the united states. >> the fbi's layer did not say if the information would be made public. >>> and asian airline is doing better, the annual -- finding improvements for airlines in key three areas. >> landing on time, fewer baggage is being lost, and fewer flights are being overbooked, passenger complaints, they have increased and reached the worst level. >> joining us, a professor at golden gate university. >> let's talk about passengers patients.

, and marry that with this advertising technology that uses algorithms to show digital advertising, and that was a mobile video product which, her verizon has already introduced, that seems to be the plan, because her businesses are so similar, you can reduce a lot of cost. there is synergy with verizon that you might not get with any other company. i think that's why verizon is taken such a hard look at yahoo!. mark: what can verizon or google dotime or at&t or comcast that marissa mayer could not achieve? paul: the real challenge for any owner of you who is to try to get that core operating business growing, yahoo! has been a business that has seen its revenue decline every year for the last six years. it has a large user base. users,e billion monthly they have not been able to make those users a compelling by for advertisers. the real challenge for yahoo! is aat is it is no longer credible player in the search market. google took over that business. shiftsocial media, they they miss that shift for social media, they lost that to facebook and linkedin and snapchat. is a company, it is

he wants to push his algorithm on to whatever platform he uses. they don't support that yet.realm of someone who has a day job and wants to put their money into use in the market. a couple hundred dollars a week or something like that. carol: time will tell how successful robin hood will be. having said that they are smart because they have gone mobile with this. >> mobile first. it keeps them clean and simple. they really wanted to tap into what is become second nature for a lot of people in the 20's and early 30's. you wake up, you get on your phone, and it feels like your money is in your hand is how he wanted to present it to customers. carol: photographs say a lot about the people the magazine is profiling. i sat down with photo editors. tell us what is involved in doing something. you have this one, build a chair in three minutes. what does it involve? the visuals are compelling. >> thank you so much. this chair company, they are trying to take a regular chair and make it so it fits in a little shipping box and arrives and there are no tools. it is supposed to come toge