 Okay, so I'm going to talk to you about big data. Now our reveal that you need to know what a cassette tape is. Do you know what a CD a compact disc is? Have you actually used the two simultaneously? Is there a generation here? Then you're probably born in the 60s or 70s. Is there anybody who is ahead of technology who's born in the 50s that know both what a cassette tape and a compact disc is? Is there anybody here born in the 80s who does the same? You see, one of the things when we talk about technology evolution is that we often mistake a clear view for a short distance. We see things that are going to happen and we think they're going to happen much sooner than they actually will play out. So if I were to draw you a curve of that, it would be something like an S-curve, something like this, but our expectation is fairly linear. So if you are in the 1970s and are holding one of the cassette tapes and listening to music and think the world is dandy and you just fitted it into a Sony Walkman that runs on one single AA battery, you're a happy camper. It lasted, the battery lasted about one cassette tape. Those were the good ones. And you see this strange disc coming along and you think, why on earth would I ever need that? And what happened there was the largest, no the second largest technology shift in the last hundred and fifty years. And whoever can mention the first one will actually get a coffee break. You know what the first one is? Or was it? The introduction of electricity. The largest technological effect on humankind in all of our history. Why? Because we on average shorted the time we sleep with almost two hours. So if you're tired now you know why. Damn that electricity. So what happens when the world moved from analog to digital is exactly that. The industry mistook the clear view that we were going to go digital for a short distance. And there were two companies on the planet that actually held on to it and managed the transition. Philips and Sony. So why am I telling this story? Well what happens with technology is exactly this. It takes about 30 years for technology to make it as an overnight success. And we are in one of those fluxes right now and that's what I'm going to be talking about. I will basically talk about three things today. I've started with never mistake a clear view for a short distance. I will talk about two waves of evolution today and why we're talking about big data and why that is completely uninteresting. And then I will actually end up not talking about stuff but actually showing you what big data does today. And I know some of you will be impressed because you know what's behind it. Some of you will not be impressed because you think this is what we should have had a long time ago. So you are both mistaking a clear view for a short distance. Anyway, is this is this an afternoon sort of? So let's go. This is actually what the graph looks like when you have PowerPoint presentation. You know, overestimation, expectation, diffusion of technology, underestimation and then you know what happens. There they are. Invented in 1962, released in 1964, death in 1992. Invented in 1979, released in 1982 and boom in 1992. And the larger shift underneath is from analog to digital. So we are living in this digital world whether you like it or not. And we are living in it so well that we produce so much data that is coming literally out of our ears. Some of it useful, some of it not so useful. So let's talk about the definition of big data. And this is, you know, thank God for Wikipedia. Big data is a broad term for data sets, so large or complex that traditional data processing applications are inadequate. When we talk about big data today, we talk about 80% of it being unstructured, which means text, videos, somebody recording when I'm speaking. And the other 20% are our DBMS data, which is relationship database type of data, which is structured things. They're nice and dandy and we've had them for a long time. So one of the sources of origins for this first wave of big data is actually us getting connected and getting online. Sharing God knows what from what I ate yesterday to what I'm doing in the toilet today to what I'm thinking about tomorrow. There's no end to the expression in social media, which I'm sure you have experienced. So we are actually the source of most of the big data today. Sorry to say. And it is not, doesn't look like it's decreasing. It looks like it's increasing. This is the device we use. This is sort of a historic view of a telephone that goes into a smartphone. So which is basically packing a very powerful computer connected in the palm of our hands. This was science fiction in the 70s and 80s. And I mean real science fiction. You will have a Star Trek where they flip over the talky thingy device and we think wow. And you take a look at us today and we don't say wow. We're trying to find ways to get away from it. When we shift from a telephone to a smartphone, another thing happens too. And that is we shift from talking to people to sending data to and from people. And it's illustrated by this graph. I'm sorry for the poor quality. But this is actually the monthly petabytes. And if you don't know what petabyte this, I can show you later. Monthly petabytes. And this is years 2008 to 2016. And the little purple thing, that's voice. The increasing voice. The blue thing is the increase in data mobile handheld devices. And the yellow is PC tablets. That's what we call an iPad. So look at that tremendous explosion of the production of data. That's what we do. We leave little traces behind us. Unsurpassed information. This is another slide of showing it. Where we're starting to go into internet of things and machine data. I'll talk a little bit about that. These are some of the companies that operate in that space. It could be stopped for a little bit. Because if you have put out so much data out in cyberspace. And I can take a look at that with very powerful machines. Analyzing what you're doing, what you're saying. I can pretty much form a very precise opinion about you. What you do. What you like. How you do it. When you do it. With whom you do it. Just about anything. So I just need a powerful set of computers. And some sophisticated analytical software. I'm sure you heard about the NSA. They have both. But they don't use it the way you think they do. They're very little interested in just you. The other thing that happens is that as this data gets produced. We are starting to look at things a different way. We're starting to use things a different way. I'll just briefly give the example of Uber. Who's used Uber here? Who knows what it. Do you all know what it is? It is basically an attempt to take over the world of taxis. But the fun thing with Uber is that they have no cars. They employ no drivers. They own no data infrastructure. They're built on two things. A sophisticated core engine that's data driven. And a very sophisticated analytical engine that analyzes and runs on that data. That's what they do. And a little secret to all the Uber examples you're going to be hearing about in the next year or two. Because it's the grandmother of all examples used by just every Joe Schmo and Schmack. When they started Uber, they went after the people that produced the taximeters. You know the little machine in the taxi cab that tells you the price. That's a global oligopoly. There's three manufacturers on the planet that manufactures those machines. They cost about $10,000. Manufacturing cost is a little bit less than $100. So the profit margins are enormous. That's what they went after when Uber started the software version of that. Which they give away for free by the way. Those types of business models or business examples can only be used if you have some sort of sophisticated analytics together with your data driven core. You can't do it. It is impossible. So it's all in fine and dandy up until now. We're on social media. We're producing all these great amounts of data. We're doing all those things. It's basically machines talking to us. And we are taking over a lot of work that was previously done by bank tellers or who have you. How many of you pay your bills through your smartphone? It's a good chunk of you. You're basically doing the work for free for the bank. And they keep charging you. And I'm saying that a little bit cynical because this is fixing to go away with the banks too. And I'll get back to that. Where actually technology now starts to allow us to cut things off. Perfectly good example is Elon Musk's launch a couple of weeks ago with a battery powered house. Did you miss that? Your environment people are. You should pay attention to that. Good. Because what happens is that they are actually by default cutting the power cable to the energy companies. Most places on the planet is self sufficient solar power charging the batteries consumed during the evening and the night recharge the next day and there you go. So in a world where the value chain gets scrambled it's now things are cut off. Literally cut off. These are one of those technologies. You remember I talked about the S curve that we're in one of those fluxes. This is the new entrance. The Internet of Things which is actually machines talking to machines before they talk to us. Now that might sound simple but this will have a profound effect on an organization like SEI. Let me explain. The number of connected devices is set to increase to some say 50 billion, some say 100 billion. I don't care billion. There's going to be a lot of things. There are a lot of things connected. There are going to be a lot more things connected and the worst thing of all is that they're simply machine smart. What do I mean with that? Well, this is a picture of a General Electric series 90 engines. It sits on most of the newer produced aircrafts in the world including the new Boeing 787. This is virtually a computer machine because it sends sensor data from everything that goes on this machine including the air around it to the aircraft and then some of the data is passed on to the ground analyzed with a bunch of other GE-19 engines flying around the world. So literally you would have a virtual climate and land assist machine on every airplane in real time. This is what I mean that somebody could actually tap into this data, probably get it for free from General Electric because what do General Electric care about climate data? You will get temperature, you will get air density, you will get quality of the air, how much things are in the air, how much of those little nuclei, everything. You have no idea there's about 110 different measures on air coming into a jet engine. And then you can pretty much calculate what comes out the back. You start networking those together and you will have a virtual real time, any time of the day machine that can analyze and portray data for you. Take a normal little thermostat. You know what a thermostat is? Have you ever used a thermostat? You know what a thermostat does? It does one thing very well. It keeps the temperature at a set rate in a room or in an environment or whatever. That's what it does. It doesn't think twice if the temperature is right. It doesn't think twice if there's people in the room. It doesn't think twice if it wanted to alter during the day or during the week or during the month. It doesn't really do that. Somebody else has to go up and go like me. Along comes this. That tiny little thermostat that happens to be smart. They can actually compute things, store things, it has a camera so it will know things. So when you walk into the room, it will actually know that you are you eventually. I'll come back to that. It keeps track on what temperature you want, what time of day, a little warmer in the morning, cooler in the afternoon, and it stores that and remembers that. And for most all, it sends it back to Google because they are the owners. So you'll wonder why on earth would they buy a little thermostat company for 3.2 billion US dollars? The little thing costs about $11 to manufacture. It retails for about $300 and it's selling like crazy because Google wants the data. Because by harvesting that data, they literally will be the predictor of energy consumption on the planet. Not kidding you. That's what they will do and are doing. So imagine this then, that you have your little smart thermostat, you have your little smart house, and every little iPad or device becomes literally a climate education center in every smartphone or smart device. That's kind of your line of business, isn't it? Are you sitting on any of this data? Not yet. This is actually what's going to cause people to start to imagine what they can do in climate data in their house, for instance. When you talk to an energy company like Wappenfall, you know Wappenfall in Sweden? It should be big enough. I'll do a similar presentation like this and I will ask them or they will ask me, so what should we do? And I say, well, you should buy a home security company. And they go, what? I said, you should buy a home security company. And they look at me, it's like, what are you, you know, we're an energy company. And I go, yeah, you should buy a home security company. What is it by this point you don't get about that? And we don't understand. Well, let me ask you simply. Do you have a device like this that runs on my iPhone or iPad or whatever that allows me to look into my house and control the temperature? And they go, no, yes, we have one you can buy. No, no, no, I said, do you have one for free? And somebody said, yeah, we do for the first year. No, no, I said, do you have one for free? As long as I live and own. And they go, no, we don't. And I said, well, then you can consider yourself this intermediate that somebody's going to steal that space from you. And it's easier for the one who has this already to backward integrate into your energy business and tell you what to do. That's what happens. And I can keep going. Smart grids work the same way. Smart sensors sending information, somebody analyzes and reaches that data with something else and pushes it further. And you trigger manufacturing processes, you do whatever. Smart home, I just talked about that. Smart cars tells the car to slow down, because you're burning too much fuel, or slow down, because you will not get there as quick as you think because there's roadblocks up ahead. So you can slow down and conserve energy. Now you're going to start to see those and already am seeing that type of communication into the car. BMW is the perfect example, because they will automatically send the signal to your car to reduce the horsepower under certain circumstances. Smart cities with little cheap devices in the roadway that can actually tell people to change lanes, because it's going to get slippery or move over completely because there's an accident. It's going to get completely different sets of automating things around us. When machines talk to machines, talk to people. I wasn't going to show this, but I have to, because the technology that is coming around the corner so quickly, that is replacing sensors, are cameras, because cameras are the one technology now that evolves the most. Cameras can actually see what air quality is in front of them. Who is in front of them? If it's a person, man or a woman, child, they can start to see what people are doing and analyze that. They can start to see if you're coming in or going out. They can see cars. Cameras can do just about anything. And you know they need about a quarter of your face to analyze that it's you. And it takes the camera about one tenth of a second to do that. And the smartness is built into the camera. And all the computation things goes on in the camera, and then the signal gets sent up. And I wanted to show you this, because this is something that is coming so rapidly around the corner that it even has us technology buffs a little bit puzzled, because cameras have gotten smart very quickly. And it's much thanks to Google and the self-driving car because they've developed the algorithms that actually recognizes things. And it takes up no space whatsoever inside the camera. It is tiny. So we talked a lot about big data. And big data is fine and fun. But the whole thing is about analytics. So now I need to ask how many have read at least some little bit, you know, the university course on statistics? Oh my God. The university, you know, the more complex course on statistics. Anybody with a PhD in statistics? Anybody with an almost PhD in statistics? So when I say analytics, do you know what I'm talking about? Probably not. So if I say statistics, you know what I'm talking about. Good. So imagine statistics is a lot about talking about what happened yesterday and making sense out of that. Now imagine if you can take that and also figure out what's going to happen forward. That is what statistics is often used for. Now let me explain. You see Uber, I talked about the taxi thing. What it actually does is that it keeps looking at where the cabs are that has Uber. And it looks where people who have the application Uber are. So if you have it installed on your smartphone and hasn't told the smartphone not to share geospatial information, it will ping Uber all the time. So as you move about, it will start to look at who's going to be needing Uber and where are the cars. You know, they guarantee you a car within 10 minutes. Try to get taxis.com to do that. Good luck. Because a sophisticated analytical algorithm is computing this all the time. So it moves about and it moves around. And as you start using Uber, it starts recognizing you and your behavior. So it starts to keep a little bit better track on you. Because most of the taxi fares and taxi trips you do are the same. I'm sorry to say, I know you don't believe me, but they actually are. You have a taxi behavior that is very similar over time. Especially if you arrive to an airport. And it's global and they recognize this. You can try it out a couple of times and see how it gets smarter dealing with you. That is analytics. So I'm using historical information to try to figure out where things need to be today and anticipate what you might want to need tomorrow. So instead of talking about what this looks like, I want to do a demo for you. And now bear with me because all the technology gizmos we've had. And hopefully this will be fixed the connection problem. Yeah, let's go there. Let's go here. See, what I am going to do now is that I've done a very simple thing. And that is that, you know, the WTO, the World Trade Organization, how they gather data from all the trade that goes on in the world. Up until about three years ago, there was no way you could analyze that data without loading it in increments, building funky cubes. Cubes is a way of doing analysis if you don't have enough data power. You have IT involved, you have to program things and so forth. What's happened in the last few years is that technology, the development of technology with Moore's law that you've got twice the computing power tomorrow than you got yesterday, and new infrastructures has allowed us to take large amounts of data, shove them into a memory and run queries to it really, really quickly. Structured and unstructured data doesn't really matter. And what that allows us to do is, for instance here, this is World Trade data from 1988 to 2013, updated now in January of 2015. This is 300 million rows of data with most of the trade data you can find in and out of a country. Now take a look at this, because what I'm doing now in real time is that I'm loading all this data up into memory and I'm running an analytical engine on it. And I'm looking here at the top importers of all regions, the volume of money for that and what the top five commodities are. So the top five commodities imported into any country are minerals, fuels, oils, distillation products. Second one is electrical equipment. Third one is nuclear. Reactors, boilers, machinery, etc. I would say that the world runs on energy. I can look at the top exporters in the same way, 300 million rows of data pushed into memory, analyzed, and there we go. I can pretty much conclude now that being the nuclear reactors and boiler machinery business is fairly profitable because it's leading in exports and it's one of the top things in import. So this is actually what it looks like. This is about the simplest big data example you can run today. This was impossible to do three years ago. Now you can imagine if you were to get access to this World Trade Data and you load up any climate data as large as you can find it and push it into the same machinery and I will ask it once, what are some of the correlations between the climate data and the World Trade Data and it will actually load it up and show me that as fast as I did now. And I don't know the last time that the climate data was updated in the world but you have to prepare yourself for that being updated every minute of every hour of every day of every year. That's how fast this is going to go and it's not a question if somebody will do it, it's a question when somebody will do it unless it's already done. Now I think you are in that business. So let me show you something fun and I'll end with that. You know how countries lie? I'm sure they don't do that when reporting climate stuff but they certainly do it when they report import and export. So what I've done now is I've actually asked the computer to tell me what are the discrepancies between what imports are said, oh you don't see that, sorry I'll push it down a little bit, what the imports are said to have been and the import mirror what actually other countries are reporting as exports into that country. That's pretty shrewd isn't it? You know it's very simple. So you look at China who've reported this amount of import in billions and this is what people have exported to them. So and I'm curious actually to see what it's made out of. Electrical and electronic equipment, nuclear reactors, articles of apparel, apparel, toys games, furniture. So let's take a look at the export statistics. Same thing, mirroring effect, oh see the US is lying too. Who would have known? That's what they were exporting the most of. We're getting the most money out of. And lastly if I just want to do something fun I can see the top trade partners in the world how they have developed over time and if you ever seen Hans Rosling this is the real shit not his fake shit. And I can pretty much say that China with a trade deficit is counterpart with the US with a trade surplus. And now I actually need human power to figure out that those countries are interdependent of each other in a probably very dangerous way. This is what the database looks like if I were to pull it up. This is all the data. I can pick out just about any country and ask it any trade question you can possibly imagine. And I'll actually end with that. Thank you very much.