interlinked by data satellite to create a global interconnected network of cloud data centers. peter: why have clouds developed? >> well, there are a number of reasons. one of the reasons is that they provide an efficient storage facility for the rapidly growing store of the world's data. certainly as, as individuals it's difficult to contain all of our data on one computer and for companies, there are enormous potential savings. zoom are being realized. so we're in the early days of the cloud. because as one person described it, if you want another definition of cloud computing it is an opportunity to outsource all of your i.t. operations. in other words the cloud company becomes the i.t. department for corporations around the world. so there are certainly storage savings and cost savings for individuals but especially for companies. and governments as well have been moving to the cloud rapidly, especially the u.s. government. because it offers the kinds of savings that corporations have begun to enjoy in, in i.t. operations. >> well you write in your book, "to the cloud," tha

data centers. these are an enormous buildings located all over the world that are filled with in some cases tends of thousands of servers that are storing the data, and these data centers are located all over the world. i mean everywhere. one of the best places to locate a data center is in a cool climate. canada and scandinavia have data centers because the servers need to be kept cool. if there is a physical location, it would be in data centers that themselves are interlinked by cable satellite to create a global interconnected network of cloud data centers. peter: why have clouds developed? vincent: there are a number of reasons. one of the reasons is they provide an efficient storage facility for the rapidly growing store of the world's data. certainly as individuals, it is difficult to contain all of our data on one computer. for companies, there are enormous potential savings, and some are being realized that we are in the early days of the clouds because if you want another definition of cl

so there is great debate over big data. i call for greater oversight of the cloud and big data is in our excitement to embrace the quanatative bits in the clouds is we tend increasingly to see the truth embodied solely in that can can be quantified or organized in correlation. as a trained socialologist i learn there are limitations. and i fear we are loosing sight of the importance of equalitative, historic theoreticle information, of capturing the subjectivity of people whose data is located in the cloud. we have numerous cases of errors made in big data analysis. largely because of the misuse of correlational analysis. as a result i think we need to be careful, to be cautious and to recognize that our society has benefited enormously from multiple approaches to understanding data and information. and that we should increasingly emphasis the need to supplement quanatative big data analysis with those from humanities and association social sciences that are more equalitative and historical. >> host: and finally, it struck

we combined the lights data with the population data as well as a few other data sets, to make sure that we can confirm each light is actually a population place. narrator: using a combination of various population and geographic data sets balk is able to determine how many people live in low-elevation coastal zones. balk: this is southern vietnam. so, we see the white areas are administrative boundaries. the red areas indicate the nighttime lights -- the urban areas. and this blue data set is the elevation zone that we've constructed from the srtm data -- shuttle radar topography mission elevation data set. so, we have ho chi minh city here. that's the big red area. and we'll now add the elevation buffer. the areas that remain bright red are above 10 meters, and all other areas right here where the blue has overtaken it is below 10 meters and it's evident that the entire city is not below 10 meters, but much of it is. this notion of combining these different types of data at these fine resolution is an innovation and it's an important one for effecting policy because without data of thi

each pitch in a baseball game is a data point in a season of pitches become big data. the more information you have the more you can learn if you ask the question properly. ann johnson may not have all the answers, but she is an expert on how to ask the question. her company in toronto helps people ask questions of their data. i think sometimes you're going to have to put it in the perspective of google for me so that i understand exactly what it is you're doing. i go on google. i ask it a question. i want a fact of some sort. it returns that sort of thing. is that what happens in big data? people ask questions of their big data and they get answers? >> so what happens right now is that when you ask questions of the data tools people have you have to be trained. there are special languages you ask the questions in. and when you ask the question it takes minutes, hours, sometimes even days to return the answer. so imagine if you were using google and you were searching the internet. and you knew it was going to take ten minutes to return the answer. >> that would be rid

data from a.l.p.r. devices, including photos, a time stamp and location is retaineded in thousands of databases across the country and shared by various agencies and sold to private companies, with little or no regulation. he says it's unconstitutional. >> if you don't have a probable cause against as citizen or individual, why are you keeping data. there's no reason to do it. >> use of al pr has interaffed. >> jennifer limp, a lawyer with the electronic frontier foundation says with a.l.p.r. affordable. >> it has a position to collect data on a scale we have not soon. in los angeles, the l.a.p.d. has 250 squad cars equipped with cameras, and 35 satisfactionry cameras. each records up to 1800 plates per minute with a capacity to collect data, they are very, very large. >> large and according to limp pay to invasive. >> if a licence plate camera pecks up the licence plate many team during the book, it pinpoints the location, charting the pathway through loif. that reveals spsive information, telling who

in one place you need data warehouse to extract and analysis the gap between data and information. >> so the patient surveyed our most used tool for quality care. >> that aspect the quality care that survey is what we used to determine patient experience. >> i maybe missing part of our question. >> i'm not sure i got the are answer i was looking for i'm trying to determine you talked about all those manufacture and sort of indicated start it is compleshthd but quality you have to have manufacture given all the sources of data you collect including the patient survey how do i get to determine your meeting the goals of clinical care. >> the last 4 slides is representative of the kind of measures we're trying to track and move the needle so if fewer people die from an infection or an infection from their wound after surgery or more people are getting mammograms or vaccinations we feel those represent the quality of care and dr. chin i was going to ask you commissioner that it sound like when i looked at the patient status you have to standard device those because they're asking a questi

this data you get is one data source but it's a lot of information.nds of things have you learned that you didn't know you were going to learn about locations? mark: what we have really learned is that it is a good source to determine if it's a genuine transaction. we know in general when you are traveling and we can compare that to merchant locations. cory: how long does it take you to understand an individual client? do you have certain types of people that are similar in some ways but different in others? do you have a handful of prototypes? mark: we have a profile for every single card in the system. we create a profile for every single card. that's what we used to calculate it. cory: when i first had kids, i got a lot of dings because buying diapers was not a big part of my life for a while. mark: profiles change over time. we're constantly motor fine dice modifying -- were constantly modifying the profiles. cory: at a certain place, you're going to do different? inc.? mark: we're looking at more patterns of fraud. where are we seeing fraud hotsp

cory: this data you get from phones is one data source. there is a lot of information.sa is tracking similar information to what kind of things did you learn about location? mark: it can be some really good source to determine if this is a genuine transaction. we know in general when you are traveling, we can compare that to the merchant's location. cory: how long does it take you to understand the individual or do you have certain types of people? do you have a handful of prototypes? mark: we have a profile for every visa card in the system. we create a profile of what's normal for that account number and that's how we calculate the score. cory: when i first had kids, i got a lot of dings because i started buying diapers. mark: we are causally looking at your spending behavior and modifying the profiles are your card. -- for your card. cory: do you recognize people at certain stages in life? mark: a little bit here from a front perspective, were looking at patterns of fraud. where are we seeing fraud hotspots. as we see your patterns become unusual is when we start to

up to e lack and look at the data they first become mrimd limited in english language they also have clarified as proficient in english and we had an average so it is determined it takes 5 years for an english learn to be fully reach english proficiency and this is actually a really good number if you look at 5 years ago it takes 200.5 years but 2013-2014 they only have 3.8 years the district is doing a job in kraifr the english learners, however, if you look at the beaked by the division level so some of the basic the at the secondary level it takes along these for to be proficient in english especially, if you look at the latino group it takes a substantial amount of time so let's move on to the next slide okay. so this slide we trying to look at their language tests basically the state test the star test and it is really hours that the english learners will not be english proficient but that is what we want the idea we want to train here actually, if you look at the percentage of the above let me go to the rating of what this star is about it is based on the rating you basically 3

disappointing data out of france. disappointing data out of germany.composite out in just a couple minutes. equities just off session lows. the dax still down. euro-dollar, 1.0684. that is trading weaker as well. the markets just starting to roll over a little bit. euro-dollar down by 0.4%. if you want to talk about these markets, i'm on twitter and you can follow me there. ♪ guy: germany's business blues. will will bring you the eurozone manufacturing data next. francine: tsipras on the sidelines of a european immigration summit. guy: and, it could be the biggest libor fine yet. deutsche bank may pay two point -- $2 billion to settle a probe with the u.s. and u.k. welcome to "the pulse." we are here in london.

is showing now and probably at the next quarterly meeting i'll have the exact data of each specific district and station and again, we're all focusing on those five -- are we obtaining the goal today? not necessarily so there's a mere i there's an miriad of reasons we may not reach the goal and you know that and whether it's our presence being the crime prevention factor things like that. >> i want to acknowledge captain silver man's efforts after the killing on wednesday of the killing of mr. yee in a crosswalk outside of a senior center and so there there will be more stings in that area after the tragedy injuries of a a 3 and 4-year old old on folsom street the police station was right on it with efforts to really work with our office and pedestrian safety community groups to make sure that we're showing the city's efforts for people to drive much more safely in those areas and i think the citations are another part of it besides the visibility and it's unfortunate that the culture of speeding continues unless we take more efforts with engineering and continue enforcement on a re

california to see if they might prefer another plan whether or not they've challenged we do have the data 61 percent never respond they were passively enrolled into their current plan the majority of the people reenrolling or passively rereportedly on 6 percent choose a different carry in 2014 so unfortunately, we don't have county level data we've hoping through california we'll have more information as it goes on but if we apply the statewide it is a 72 percent renewal rate of the 41 thousand people that enrolled last year would have reenrolled by san francisco out paced the state of people enrolling i think we have more people more willing than statewide moving on 0 medi-cal that is restricted to the 2014 calendar year so you may recall we shared our projects 2014 last year and 81 to 95 thousand people statewide had been eligible for medal we expected half of them to enroll and actually from the hums agency reports that our projectors were exceeded actually enrollment preceded we have 56 thousand people enrolled in medi-cal and other thousands enrolled in non-magazine i did refers to p

sealock get the data and data protection that could be used to help them intervene but it is also worth reshaping the believe some of the kid so we all though the when teachers given permission there led to believe if they are gifted send students achieve more. such as of this comprehensive dossier the readout says this kit is a potential dropouts. that is the kind of question and even in the court of public opinion when rely cover kids you have a break? what about the value of that information as may talk about reducing dropping out in general. it is very complex and totally fascinating and scary as well. >> 25 years from now you were all in a book tour with the sequel what are you talking about? >> a great question. i would love to be talking about how the next generation of assessments opened our eyes to each person's capabilities that allow us that each student graduate of of robust record that they could years to qualify themselves to the above warehouse -- to the employer you could see students using information to improve how to learn. if i could call up that sum total first grad

with other data. is that correct? >> yes. well, aggregating it with itself and then with other data. you can have a buffer that only stores three seconds of information at a time in the drone, and to the extent it detects any illegal activity, that puts you in the jacobson world of -- >> this is a month long surveillance. >> for the analysis, i'm going to think of aggregation as taking the drone data, adding to it other data, and then drawing conclusions from it. versus persistence which is taking the drone data for days and days. and my view is persistence alone is going to be enough to trigger the fourth amendment protection. and i base that view on the jones case and the five justices who seem to think that the trespass didn't really matter so long as there was persisted surveillance even though the case did go down on trespass grounds. i want to flag to folks a really interesting discussion about exactly this at the privacy law scholar's conference on june 24th, 2012. just search on this from jone

spanish data improves. and then, the french data does not. the survey is 52.8. the survey is 52.0.e disappointing data out of france. they seem to be shrugging everything off. the dax is up. after the break, we will bring you some final four with guy johnson. jon: good morning and welcome. the pulse is coming up and we are joined by the host. the data out of europe is there a story to be told? guy: there is a story. bad news for greece and good news for spain. it continues to get stronger and stronger and the spanish recovery gains. we are starting to gain traction. you see it for another term and that is really bad news for greece. if spain can do it, why can't the greeks? they are looking and saying, this guy can do it. why can't you? >> we have seen purchasing to get the tax and version. -- the tax inversion. guy: you point it out to me and this looks cheap. it looks cheap. six month ago, it would've cost you -- jon: 1.2m in dollars . things are starting to improve. guy: the merger makes sense. there are a few concerns about the regulatory story. they seem unconcerned. the quest

data. this is the official manufacturing pmi number for april. that's going to be a key test for the china markets. back to you now. >> all eyes on that pmi number. for now, thank you so much. twitter shares plunging on tuesday as the company's first quarter erngarnings leaked early and showed the weakest since the ipo. they tweeted the results after finding a link on twitter's investor relations paid. twitter fell by 18% in regular trade and another 1% in after hours. twitter ceo will be on squawk alley today in a first on cnbc interview at 11:00 a.m. eastern time. you won't want to miss it but is it time to dump twitter stock or is this a good-bying opportunity? we are active on social media, twitter and facebook. e-mail too at cnbc.com. our personal handles on the bottom of the screen. yesterday you and i were discussing how we've seen this big run -- disappointing photo right there. >> us during the break. >> we have a little fun during the breaks. keep it light but on a serious note tw

♪ ♪ emily chang: what kind of data are you gathering about these users, and how do you use that data?zuckerberg: i do not think it is anything different from how people use facebook normally. the biggest thing that we have had to do to make internet.org work is connect with the different operators in these different countries. so for example, airtel in zambia, to make it so that people have a very easy way to go buy data when they want to do more things. so for example, you might be browsing facebook and see a link to news, or you see some video that you want to watch. and you know that is rich media, so that cannot be covered for free. but we make it so if you tap on that, it is very easy to just right in line pay, and that is good for everyone. it makes it so people can discover why they would want to consume content on the internet. it makes it so that airtel and our partners can make more profits, so they can continue investing and building out a faster and broader internet for everyone. and it gets more people online. emily: google is working on project loon and google fiber. wha

hosting platform shouldn't list anything that if have a hosting number and they're required to valid data that number. >> maybe open to interpretation but the listing at the time the listing is posted it is a valid number. >> mentioned they need to be monitoring the unite in compliance is not really what the ordinance says it says the hosting platform should not providing provide the listing unless the listing includes the number and it is valid. >> deputy city attorney those are two separate issues in the recommendations it not called out i recommend you consider it it creates a new going on the host mrafrpdz and the obligation the hosting platform has to verify the fabulous is valid this is included in supervisor campos ordinance his ordinance also amended chapter 41 a for 0 right of private action against the platforms the private right of action against people that are legal representing the unit out not against the hosting platform that is a change you can consider as part of that recommendation a do you want to make it an obligation of the hosting platform to verify that the registr

northern border would be i'll suggest the goal of this conversation is for us to do that take the data and the antidotes and see where we end up so my suggestion here we hope to next week have a minimum of two maps we can vote on we're not voting tonight only discussing but prepare a map this is my proposal we prepare a map to polk either to polk and one to haven we can decide after we hear more if the community and check back in but make that decision we'll heard loud and clear larkin street was not there we need a map colleagues, any suggestions around that does that sound right yes. >> inspector it doesn't capture the severity we're talking about a large incident. >> not to open old wound one of the things we've found when we did the redistributing the whole issue you're talking about there was a lot of finally around the issue of larkin polk and van ness and what to do as you do that and try to wrap the tenderloin into one piece it comes down to what it does to northern and honestly the 09 districts in order to capture those tonight you're having to capture a bigger piece it is int

meta data's really easy.you can put it into an al gor rhythm and the more you have, it seems like content becomes less and less important. i think it was maik michael hayden that said we kill people based on meta data. whether the panel compels that sort of distension or whether it's something we have gravted on to it and can reverse. >> well, as most people here probably know, the fourth amendment has been interpreted in different ways and at different times in our country's history and has really struggled to keep up with technology and has gennarilyuineerallyÑi lag edged for a long time. in 1927, the supreme court ruled there's no fourth amendment right of privacy in a phone call because the government didn't have to intrude on your property in any way and it wasn'tçó until 1967 that the supreme court revisited this. privacy isn't just in your privacy. it can !ube, it fl lows the person, not the property and it can be in a phone call if you go into a phone booth and close the door. many doctrines have chan

data that will drive euro-dollar. mark: did that mean the data should be put aside?ane: given the german data, i am pleased i have not. just as the german data has been surprising on the upside, a lot of the u.s. data has been surprising on the downside. given that story, if you were to take a weighty qe plan, -- take away the qe plan, i think the dollar would be higher. if germany does carry -- that takes parity out. caroline: dollar strength has been the story, but a little bit of weakness in the u.s. data. the fed trying to talk down talks of interest rate rises. jane: that is exactly right. we have been a little bit nervous, maybe putting on the data we might have to -- but right now, looking at inflation data in the u.s., it is benign. wage inflation is going to be important tomorrow. looking at the average earnings data. also, if you look at core cpi. they are benign. in our view, right now, that suggests the fed does not have all those boxes ticked. for us, it is december september, not june september. mark: how do you read from all the noise from fed official

i will get that data. >> thank you. i think it would be great and then to follow up on another question of the coordination of city services, so associate superintendent how are we partnering with sfpd on this? >> okay. just an off the cuff thing, right. okay. no, i think our partnership with sfpd is very, very good. okay. we have an mou with sfpd. we have a point of contact now with sfpd. we're meeting with our sros on a regular basis. we need more of them. the superintendent and i met with the chief of police in january to discuss this because we really want to build the capacity and strengthen the relationship between our officers and our school staff and our administrative staff, so when necessary this is a violation. it is against the law. it must be reported. we always have a staff member present. it's actually written into the mou that we ask the students for any of the sensitive cases. you have the right to have a parent obviously first and foremost but if it's a situation you would like a staff member from sfus

is still there, and then download streams of data and just as you heard, data is matched up and synchronizedockpit voice recorder the radar data and anything else that would have an input to telling us exactly what happened. it will corroborate what went on in the cockpit as far as the controls and the engine the altitude all of the control surfaces and tell us exactly where and what that plane was doing while he was manipulating the controls. >> paul that plane hit the mountain at 400 miles per hour or faster. would that necessarily damage the second black box to the point where we couldn't extract information? >> well hopefully not. the actual data is on chips, small chips, just an inch or less on each edge. they are well protected within a number of different shells. they know exactly what they want and how to protect it. it's got several different suits of armor surrounding it. hopefully if the chip itself is intact then the data should be there. >> i want to go to you, david soucie. the information, just to put it simply that investigators will want to find out is if this pilot was manu

francine: it is data dependent. guest: i think the data has been coming in much weaker.hat weakness will now creep into employment data him up maybe tomorrow or next month. francine: what is your take on u.k. politics? a major debate today. are we looking at pound weakness on the back of a labor -- guest: labor andur and torries are virtually no and neck. if it is that, that will be tough news for the economy because we don't know how progressive they will be. it is a difficult. outcome -- difficult outcome. the market is kind of ignoring it. we could see some kind of sterling downside if we focus on it. francine: thank you so much. i think the markets are ignoring too much. just my opinion. coming up, deal or no deal? iran's nuclear talks set to get back underway. ♪ francine: iran nuclear talks are due to resume at any moment. our middle east editor joins us. elliott, deal or no deal? they are still talking. elliott: i guess it is slightly less clear than it was yesterday. the fact they are still talking believes they believe they can get 1 -- indicates they believe the

if we do not get everyone to agree to share data, there -- we are not asking for internal data, but we won't be able to get the things we're looking for. i made an analogy about -- men helping an elephant, we will not know what we are touching. mr. clinton: josh, this insurance model that you described to us, just as a normal person listening, from a naked business model point of view, it sounds like you are saying if i do this well, i can ensure people for less money. that is what you are trying to do? so what we want to do is ensure for wellness -- insurer for wellness. since i'm a presumably -- since presumably the problems for medicare and medicaid and the federal health insurance program , if they follow your lead, if you can do this, that way you can be truly disruptive -- what do you need to happen beyond your control for your business model to have a good chance succeed -- to succeed? joshua kushner: i think it is about incentives. think there is a lot of things we are doing that are very differentiated. to go back to the question you asked before, how does technology impact ou

. >> this is insightful data for the kraigdz list this was the only data i accounted get rather thansing the board summary statistics for airbnb i was hoping to look at data i could classify that right the only rental approach i had data available from 5 different points of the time and commissioner antonini. >> yes. thank you for your report i have a couple of points i think you said something like the average week r was 43 days a week and average income $950 per month obviously some represents are making a lot more is that accurate to our report. >> as and mentioned at the end of my comments i really hope you take those numbers as guidelines i was trying to estimate revenue i don't have booking data from airbnb so it is not possible for the exact number of revenue but you're able to get a general sense so the assumptions i made is that each liz was rented for the average days that's been 5 days per booking then we do know that reviews are under recorded by some instances and reviews for the way i tried to proxy to estimate houchltd their rent and inflated the number i ended up with

collect that data? >> we have automatic passenger counters, about 30% of the buses and circulate in the system so we get rich data this counter is essentially two laser beams. if you break this one first you're getting on and the next one you're getting off and not whether you pay or not but tracking everyone that gets on and off the bus and gives datsa how many are boarding and which stops and crowding and we defined crowding on the buses based on a square footage per passenger so we assume all of the seats are full and there are some standees but it's not packed to the gills situation, so on a 60-foot bus for example we can serve a bus if it has 90 or 100 people on it is full and for a 40-foot bus it's full if there is about 60 people on it so it's a combination of seating and standees. >> sorry for 60-foot bus how many seats? >> there's about 60. >> about 60 so you're talking about all of the seats are taken and 50% more are standing on the bus? >> yes. >> actually i am surprised and impressed there

the solutions to bad data is better data. sometimes the solution is literacy and curiosity and asking hard questions about the nature of the information of the judgments. honestly, there are fair-minded folks inside all the companies who are scientists who if you asked them, they will freely admit the nature of something like a fish be that proficiency had never been legislated for 100% of student. we defined the proficiency with reference to a bell curve than the idea that 100% of students will pass that point. you know, it is a laudable ideal, but an absurdity of practice. >> i am going to reach her boat. i wonder what examples for the younger grades and how to need that seems to be one of the plays to stop top of families about this is where your child goes through. one way to show what they know. there are other ways. so that's a good question about the pushback from the youngest grade. at my school we have the yearly regular outcome and core standardized testing. we feel a different way in terms of what expect it. >> yea

from it -- health care data? at the other end, you are combining all of this with the big data sets and you want to protect this and then you get tremendous power and you can go to other people but you might not want your health insurers to see parts of this and your employers, but we can take care of all of that stuff and we can get to where we are going to become triple on that spectrum but we have to get past the loud yelling that everything has to be ultra-private and we are going to make more laws to make sure that and it hinders instead of advances the field and everyone wants to believe that data has to be protected but new regulations and new penalties about how it needs to be private hinders things. we have not had a natural -- naturally known -- national discussion on this information exchange which is one of the biggest problem and causes of that because we do not know who is who because we do not have a national database. all of the countries with a good health care system has a national database, but

they'll have a lot more data but everything they have seen so far from the data confirms their initial suspicions. >> what do you make of the fact that it was on auto pilot? he kept reprograms, speeding up the descent. but it was on auto pilot. why? >> that's pretty typical. auto pilot delivers a smooth flight, delivers the plane exactly where you want it. it's because it's very easy. all you have to do is dial in additional descent, where you want the nose vis-a-vis the horizon. auto pilot just makes it easy and quick for the pilot to be able to fly. i'm a little surprised because i would have thought he would have hand flown it. that did surprise me. it's a very quick and easy way to fly, auto pilot. >> that's kind of why i asked. i don't know if that's the picture that's being painted his intent and the fact that they're now calling it premeditated murder. i think i was hearing from our correspondents on the ground that at the point of impact they were going 420 miles an hour. you can just imagine what the passengers were seeing outside the window and what andreas lubitz was saying

i collect data scientists just because the fact we sit on so many records of data. back in 2014, we were processing about one million records in hours about his raw voter contact. we also realize our volunteers do not want to talk to democrats. we actually do it on our own to make sure we are talking to democrats. when we were talking about the fact of talking to local republicans, we make sure the democrats were also not turning out. we have to keep an eye on both sides of the aisle to make sure that it is actually being prefilled -- fulfilled. we were getting a broad base. when they were doing survey sample sizes like 500 a month we were doing 1000 week in every single target state, plus the voter contact. we have so much information. >> what do you do for cell phones? when i worked on a gubernatorial campaign hours and hours i spent making phone calls and they were dead hits. the majority of my heads were dead. you don't get the cell phones, you are losing the population. jesse: every single sample is at least 30% mobile. we will make sure we have an online audie

in terms of finding assisted living as well as having caregivers notify in with them we need more data it is fair to assess for different reasons and for access to hours for us we would like to be part of the hope when people come to our doors to provide ways to show them what's happening over time and what happens in the future so they have at least some assurances they're part of - they couldn't lose hopeful e hope essentially it is important to have data something you and your offices can bring to the table when people come to your doors where's the planning housing for me breaking down that is very complex for someone that needs housing right now today and it is for us people who are service providers to break down and make it clear this is coming up in this neighborhood and this neighborhood but we know this is coming up in the pipeline we need your support in that thanks. >> thank you. >> thank you. next speaker, please. >> good afternoon my name is paul and i'd like to address the board as a long term resident of the mission i've lived here for 35 years and served over 20 years

with oil crashing, you need to know the data now.traders and derivatives market, this is the third riskiest country in the world after ukraine and greece. mark: thank you. that is today's latin america report. still ahead, lessons from fashion icons. we will look at what determined to makes the cut in the highly competitive fashion industry. the creator of new york fashion week joins me when "bottom line" continues in just a moment. mark: welcome back. the national football league has one final approval of a $765 million settlement of explain or head injury claims. a federal judge ruled the revised deal covering more than 20,000 players was fair. the ruling is expected to be appealed by opponents who argue the money falls short and deal terms are unfair. a group including billionaire tony ressler and grant hill has won the bid to buy the atlanta hawks according to a person with direct knowledge of the matter. the sale price was $730 million. that is a look at the top stories we are following. fashion is a trillion dollar global indus

centers and google data centers.are the private networks between these data centers taking all the information that was sent over those links. it's a lot of information and when yew hoo pushed out an update to its internal file system storage, the way they store data they had to sync this stuff back, nsa had to turn it off because the essentially all the stuff from everyone all at once over the period of a day or a few weeks. that is to some people and the first thing we look about from the 215 metadata program was pretty shocking to a lot 6 us when we realized that, you know, some amazingly large 99.9 something fraction of these people are entirely innocent in their date is being collected. i like to say the nsa and its partners were converts to what we call big data way before industry caught up. they were basically collecting all of the hay in the haystack to get access to all the needles in the haystack. finally this kind of surveillance is gone from sort of relatively unintrusive. xd part of this is just how soc