 And out of everything that you felt that you saw, was there anything that surprised you? Good question. I don't know if anything surprised me. I think it was, you know, a further discussion and development of some of the big themes that have been going on out there. You know, what's different, I think, and I did this last year as well, is understanding how these things have evolved. One of the panelists I had this morning was on IoT, of course, a big topic this year. But, you know, and he kind of put it well, he's like, you know, this is my third Dell world. Two years ago, he didn't even talk about IoT. He wasn't even a subject. Last year we talked about it, but just barely, and then this year it's sort of the central theme. So it was those, that was the kind of things that happened this year. So, and again, is that surprising? No, it's not really surprising, but it's interesting to mark the evolution. And sometimes you've got to do that. When you're in a very fast-paced environment, you know, the changes seem to be not that big of a deal until you take a step back and go, wow, wait a minute, you know? Like if it was, yesterday was back to the future day and you look 30 years back, you look 30 years ahead, it's crazy, but even, not even that much. A lot of times it's 30 days, right? 30 days you look back, like, wow, that just seems so different. Certainly 30 months, which is a realistic timeframe for a lot of businesses, is that's a big deal for them to kind of mark where they are versus where they were 30 months ago. And even 30 months from now is just far enough out that they can kind of start to think about, you know, the things that will be happening that. I was just talking with the gentleman before you that was telling me that since they announced the hybrid solutions, hybrid cloud yesterday, that overnight they already have customers on and using it. Wow. So it's been 24 hours. That's amazing. So not even just 30 days. That's, yeah. But 24 hours, it's pretty incredible. And the stuff that we're seeing, not just nationally, but internationally, it's coming together for people that are using this in traditional methods. So it's just, it's so cool. I really had a great time here. Yeah, I did too. I think the things that they're talking about is exactly what we were hoping to see and we got to see it. Yeah, yeah, no, that's great. And we do have one more session. So we've got the closing general session coming up. That'll be interesting. I know one of the gentlemen I spoke to today at the beginning of the day will be on that. He was the country's first chief technology officer. He actually worked for the president and he's involved. All right, let's head over to the session now. We're gonna go live now. So we'll see you in a little bit. Remember when you were 10 years old? What did you think the future would look like? Something fantastic and unbelievable? Maybe something closer to home. Predicting the future is more of an art than a science. Life today has turned out a little different than the movies would have us believe. Connected cars, wearables, and touch screens are mass market. We might not have our jet packs or hoverboards yet, but we're making progress. What's next is anyone's guess. But one thing is for sure, it's a connected future. Today, almost five billion devices make up the Internet of Things ecosystem, and 20 billion more are expected in just five years. More than half of this connectivity is currently focused in a few industries, but very soon it will reach everywhere, raising new public policy and security issues. The Internet of Things will be an $11 trillion per year industry by 2025. 70% of that is expected to be in business to business applications. So let's dare the dream of a future more connected than ever before. Think of the opportunity. Ladies and gentlemen, please welcome tech reporter on NPR's business desk, Arthi Shahani. I'm assuming extremely well rested. You went to bed early. You weren't out on 6th Street. Is that right? Good, good, good. So thanks for coming to the closing plenary for Dell World. My name is Arthi Shahani. I'm a reporter for NPR. I report on technology. And as I've observed it, we are basically wiping the internet on everything. In the last year, I've had the chance to visit 10 different countries. And I recall in one country I went to, Tunisia, for example, meeting a young man who was 24 years old. He'd never owned a computer until he was about 18 or 19. And he figured out, having grown up on a farm, that if you put sensors on the refrigerators, when there is a blackout, you can actually remotely figure out if the temperature is too low and your crop is going to go bad. So that's like a very simple use of IoT that he developed really hacking on his farm in Tunisia. I've also interviewed people, for example, in competitive sports. With the America's Cup, I've gotten to interview sailors from the Emirati team New Zealand who, after they finished playing and sailing and the blood, sweat, and tears that go into that, when they're back processing how the game went, they're not just getting feedback from their coach. They're getting feedback from engineers on how they performed, given how the sensors give feedback on their performance and their work. And then just last week, I actually got to drive a Tesla Model S that was semi-automated. I actually wasn't sure how automated it was. So either me or the car, I'm not sure which combination of me and the car. We almost hit someone, but didn't. And we'll talk about that later, candidly. And so the point of this panel is basically to talk about the internet of things, the promises, the applications, the way it's being used in enterprise and for the consumer, and then talk about things about it that are inspiring and maybe also scary because it's the mix of things, right? And it's clearly one area where arguably technology is outpacing our ability to process it, manage it, regulate it, use it. And so we're going to have six people on the panel today. My job is to facilitate that and their job is to be interesting. And hopefully we'll have a lively conversation and I'll ask you to tweet your questions as long we'll be able to take that from you. So let's invite on the six people who we have. We're seated. I'll go ahead and introduce each of you briefly. Anish Chopra is co-founder of Hunch Analytics and also former CTO of the United States. Jeff McGrath is with McLaren Applied Technologies. We have Joyce Mullen with Dell OEM Solutions, Alexis Ohanian with Reddit, Michael Rayner with Deloitte Services LLP and Paul Rogers with GE WorldTech. Thanks for joining us. Great. Thank you. So I want to just start with... It's a question that really any of you can answer and I'm assuming all of you can answer very well. In terms of investments in IoT or use cases that are particularly exciting to you, I mean, what's caught your eye quickly? When you think in terms of GE, when we think in IoT, we look at it from an asset perspective, so the industrial internet. And so we've made a billion dollar investment on the West Coast to go after that market. When you look at the numbers, which are extremely exciting, you have 50 billion assets being connected by 2020 with $15 trillion worth of GDP value by 2030 and that's with a T. When you start to really look at how that's possible, in the industrial world, a 1% increase in efficiency on the industries that we serve can range from $2 to $6 billion worth of value. So just from a numbers perspective, it's a darn exciting field. And use cases applications that excite you? Well, I guess something that I find fascinating is not a particular use case, but the way in which different technologies are being commercialized. So I spend a lot of time thinking about different pathways for commercializing different innovations. Arthur, you mentioned the Tunisian refrigerator example. That's what I would call a potentially disruptive innovation, a way to start out at the fringes of a hard problem and sneak up on those hard problems from the side. There are other applications that are much more almost moonshot in their audacity, some of what you've described with GE. You put sensors on a jet engine or a locomotive, that stuff's got to work first time. And then there are other problems where we're taking both approaches simultaneously. You mentioned the autonomous car. Sometimes it's going after the mainstream, right out of the blocks, and sometimes it's driving combines in empty fields. Very different kind of problem, very different approach to solving a hard problem. Alexis? I think about this mostly as an early stage tech investor and I'm just excited to finally have the future arrive, the one that I was promised in movies. I think for me in particular, I was really excited to see the Model S demo, in part because a company that I invested in called Cruise was doing a very similar road test about five months ago, and that's an upstart in the valley. It's so cool to see so many smart, well-funded people trying to solve this problem that I really do believe is gonna make our roads a lot safer and make our lives a lot better. I love it when we find new applications and we're able to instrument environments that we would never have thought possible. So my favorite example is the connected cow. We have worked with a dairy. Which you've all heard of. Connected cow. We've worked with a dairy in India to try to figure out the behavior and the genealogy and the eating and drinking habits of the perfect dairy cow. And when farmers know what that cow looks like, they can farm to that outcome and increase their milk production. It's a great example. Well, at McLaren, we've been racing connected cars for over a decade. And what we learned from that is in a completely controlled environment, with interoperability, we can use data to drive intelligence that drives decision support for strategy in race. And ultimately allows us to inform models that help us accelerate the development of the system through immersive simulation. In both cases, keeping a human in the loop. Now that approach inspired us to go to far more complicated application in oil and gas, offshore drilling, where basically we treat the driller like the driver. We get the data from downhole on the rig where the garage crew, you could say, is operating. Then we integrate that in real time to the onshore operations control unit. It's inspired by racing, but it's ordered as a magnitude more complex. But it only took us two years to go from concept to production, actually now going global out of Houston. And if I may just to take this in context, yesterday the president unveiled the strategy for American innovation and once again reiterated that infrastructure in the modern economy includes digital assets. And so you're gonna see a lot more continued investment in what I believe to be smarter cities and other foundational infrastructure that can be part of applications. But this is also about the problems that we're looking to solve. In healthcare, we believe that 30% of the spend is not valuable. And the opportunity for the IOT to help keep people in the home longer without having to go to the hospital is a huge opportunity just as the president's goal to get doubling of our energy productivity levels so we can extract more value for the energy we do produce in the home or in the office. A huge opportunity, but also some hype around it, right? So when I think about, for example, in the health and fitness space, I can track my steps and then I can gamify the tracking of my steps. So it's not just a pedometer, but it's suddenly a competition with all of my loved ones or my mild timer, whatever. But beyond gamification, what exactly are we getting? Like, what kind of interesting data analytics are we getting in health and fitness in particular, maybe, to start with? Maybe I'll just start to get down on this piece because this is a critical opportunity. The idea that we're gonna be quantifying ourselves is interesting, but in the context of how people are engaging the healthcare system, we're gonna see a different model. A healthcare organization now that's responsible for your entirety of your care, maybe actually prescribing that you censor up the home or censor up your loved one because they wanna be more aware of an alert to keep you from falling sicker. So it may be a prescription for a censor or monitor that the doctor and their care team is engaging with you so that they can keep you safer at home as opposed to just we all individually have put up our health. And the idea being that you're not always at the doctor's office and so a way to extend the ability to see what's happening with you when you're not there. And the point of the punchline here is that there's a business model. The government's now giving you a check if you're a doctor that keeps people out of the hospital, you make more money keeping people healthier. And so with a business model, a problem to solve and data, you're gonna see those combinations emerge into. And we have a bunch of customers working on creating products and solutions to meet those needs right now. So examples include things like HealthNet Connect where you can understand, after you've been discharged from the hospital, whether the patient is taking their medication on time. The way they do that is by instrumenting a pill box. And if the patient doesn't take the pill, they get a call from their provider or their nurse and they say, hey, don't forget to take your medicine. And the readmission rates to hospitals with that solution are down by 30%. So it actually works. We do that for flossing. Yeah, you can do it with flossing too. There's such a stratification in the maturities of industries. So when you look at taking data analytics from a jet engine that's in flight and how important and technologically advanced that is, and then compare it to an industry like oil and gas when you think in terms of oil fields, we had a solution where we went to an oil field customer and we were super excited. It was very advanced. It was gonna optimize the oil field. We gave our presentation, talked about timing and cost. We all high five that we had nailed it. And the question back to us was that is great. We see that as the future, but could you tell us if the equipment is on? And so we did this because how long would it take to tell us if it's on or not? Well, then you get back to the value of on off. It took them three weeks to understand whether a drill rig was down or not. That's three weeks of lost production. So a week's worth of technology working with this customer saved millions of dollars for something as simple as on off. But I think that's a great example of the challenge that we've not risen to yet is the design challenge. It's no in the questions you answer that have the most impact. If you're not giving me meaningful insights about me, that gadget on my wrist is useless. And usually the algorithms in the cloud are aggregating across a population. They're not profiling me. To do that, you need machine learning and that's a massive data problem at every individual in a population. We'll talk about machine learning moment. One thing I do wanna drill down into that I think I interview a lot of startup entrepreneurs. And I think that there's this idea there's this inherent excitement around, look at all the sensors I've built into this thing. There's so much enthusiasm in it. Not as much thought on the data design as you call it on, what are you doing with this? Are there ever instances where actually it's just better to collect less data, like focus, do less, and here's what you get with it. Do you have any examples of that? Absolutely. We did some work with the English rugby squad. We started out trying to analyze how training translates into poor performance in a game, not pure science. We set out with body area networks, satellite tracking for motion and building up very complex algorithms to assess the work rates from these players. We now use $3 sensors that we can extract very meaningful insight from the minimum amount of data, process it locally so we don't need any bandwidth for transmission and we get very, very good correlation. Down to an individual about how that individual is likely to perform in the future. If you could apply that into our everyday lives, then I think you will have cracked something, but it's complex. We'll see founders who get very seduced by this idea of, oh, look what I can do and not actually solving a meaningful problem with it. And I think that's where we see a lot of these kind of like stocking stuffer ideas that may make for a really interesting like, hey, here's what I got you this year, but never get much use beyond that. What we find in, we have about 130 or so proofs of concept in process right now and what we're finding an awful lot of the time is a customer is trying to solve a specific problem they think and then as they start to look at the data, the problem set actually expands or contracts depending on the use case, but they don't always know going into the project exactly what they're trying to get done. And I think part of our job is to help sort of ferret that out early enough because you can waste a ton of time and a ton of money if you do that wrong. There's a lot of data out there and there's a lot of time spent thinking about how do I ingest all of this data? But in the industrial world, only or less than 1% of all the data collected from all of the hundreds of millions of sensors is actually used in analytics. One percent? One percent, less than one percent. That even sounds high to me. Yeah, that sounds high to me because there's just so much out there. Which then, when you think in terms of using that data, it actually gives credence to some of the GDP numbers that we're talking about. But that is why though it's really important as you described to do some of that analytics at the edge because you don't want to move all that useless data all around. And you need very little to have a big impact. Exactly. Yeah. You want to be careful it seems to me collecting enormous amounts of data using a very small amount of it. There's a tendency to think that well it's cheap to collect and cheap to store. It's got enormous option value, why not just do it? Except for the fact that having all of those data actually creates a risk because it sits there waiting to be hacked, waiting to be stolen, creating all different points of entry. I started with a downer, but everybody's so excited I had to do something about it. Well I think it's a very good point. It's a real issue. I mean it's the Office of Personnel Management learned that. I mean it was a real question in the federal government about why did they digitize and store the most classified sensitive things that you actually didn't need on the cloud. It could have been in paper. That's one arguably one lesson from that hack. We're going to take questions from the audience via Twitter. If you want to tweet in your questions, hashtag Dell World, we'll take them and I'll start to take a look. And one other question I have on this in terms of analytics and the way it works, as a consumer, as a citizen, does analytics ever work against me or does it always work for me? Like I'm curious, are there perverse outcomes or should I just love it because it's more info? Yeah, it can go either way of course, right? These technologies can be deployed. Sometimes consumers make out like bandits. Look at what Craigslist did to the newspaper business, took billions of dollars in classified ad revenue, turned it into tens of millions of dollars in revenue, massive consumer surplus, but the industry wasn't so happy. It can go the other way potentially. If you look at the insurance business, for example, I think insurance companies would be well advised to think about where the center of gravity lies and who has the power in that relationship. It starts out with put this widget in your car, give me data on your driving behavior, and as a result of now being able to measure your actual riskiness as opposed to using your credit score as a proxy, I'll reduce your insurance rates. Well, that's terrific, but I got bad news for everyone here. 50% of you are below the median, and so it doesn't always work out for everybody, and that's a problem. I don't have a glib answer, especially given the context, but it's something I think worth taking very seriously. One of the opportunities is to revisit the idea of privacy and essentially the rights that we have when we interact with these technologies, and one of the questions that we tried to grapple with in the last several years is, can we build a framework that allows individuals control over the use of this information and a compact that would say, you're only gonna collect what I'm authorizing you to collect. So there's a little bit more implicit understanding around what it is that you're doing and how and why. By compact, do you mean industry standard or do you mean law? Well, our model was that there'd be a baseline fair information practice standard, and that part of that standard would be that the individual has the control over the use of the data that has been collected on their behalf, and that the organization is gonna collect the minimum necessary to perform its function. And so the idea is we obviously don't have a universal privacy law in this country. We do in healthcare with HIPAA and educational records and the like, and so could we have at least a baseline protection that would give us more tools to control that information in the digital age. That's not law, we don't have that law, we did advocate for that to be law, and we hope the industry will continue to move. If even involuntary industry enforceable codes of conduct, which gets you further down the road. Could I bring back to the question on analytics? I think there's no doubt that using real time analytics that are predictive, you can make the human more likely make the right decision, especially if it's a mission critical operation and there's time pressure. But the driver for this is in some industries we can't hire enough people to actually manage the complex operations. The threat is if we use machine learning early and then we start building up and codifying this knowledge and recognizing the patterns that we think of typical of an experienced operator, but now codifying that reaction, suggesting interventions and knowing the outcome, at what point is the human's intervention less important than the suggestion from the code? So we're not at that crossover point yet, but it's a world that's moving extremely fast and the IOT is just fueling that initiative. And I do wanna talk about that in a bit because I need to process what happened last week as well. Tesla. In terms of, I mean, one thing I think about in terms of lawn policy is to smart devices raise unique kinds of oversight issues. So for example, over the summer Fiat Chrysler had to recall 1.4 million jeeps. And the way that the software holes got put on the radar was not by a regulatory agency combing through all of the software that lives in cars. It was by two independent hackers who figure this out and then frankly dramatized it in a way that you had to pay attention because it was dramatized. You read it, yeah. Yeah, and so. Well, let me be precise. We've grappled with capability across the regulatory agencies. Some have more talent than others when it comes to these emerging technologies. In fact, we put together a bit of an emerging technologies kind of a baseline approach. One of the models that we're seeing is that this is a responsible for many of these agencies. So the Food and Drug Administration is thinking about its role in the digital health ecosystem and it's trying to strike the balance of not overly regulating but allowing for there to be some more transparency and disclosure. So you'll see, for example, the rise of open government data for regulatory purposes. The Consumer Protection Bureau has invited every American to submit their complaints about financial products that may not be designed to do the right thing. And rather than just having the regulators read all of that material, we've opened up all of that so that others could join in the cause of mining it for patterns. So that maybe an ill-timed recommendations engine that was steering you to higher fees or to a worse economic outcome for your family now might be caught earlier in the regulatory process because of that openness. And I think that that's key because when you look at the enterprise side of the house or IT and you look at the operational technology side of the house, which is OT, which is the assets, there's a tremendous amount of focus on the enterprise side of the house and that's been going on with layer upon layer of security to protect your data. But there hasn't been the same level of focus on the OT side of the house, which is your actual asset. And so one side's protecting data, but the other side has to protect the asset itself with regards to control of that asset or whether that asset disrupts the process that it's running, like an oil refinery to your car. So when you look at the OT side of the house, it has the characteristics of the IT side of the house in the early 90s facing 2015 threats and that's what WorldTech focuses on primarily. A question we're getting on Twitter has to do actually with how laws figure into this conversation and the incentive structure around it. I mean, right now, for example, as pointed out by someone on Twitter, that when there's a data breach, there are HIPAA requirements, some HIPAA requirements around disclosure, but they're just, I mean, we're sparse on laws around disclosure, right? Should there be a universal data disclosure law when it comes to breaches? What should that look like? Yes, 70% of all industrial companies were successfully hacked last year. 80% of industrials over the next two years expect to be successfully hacked. These are the things that you don't normally hear about because of the lack of regulation. And I think security professionals should be allowed to collaborate around those breaches to try to figure out who the bad guys are and put a stop to them. Explain the barriers a little bit. I think they're legal barriers, right? That's the biggest issue. So you're not supposed to be talking about those because you basically violate a bunch of other laws. The current laws typically center around the consumer. So if your credit card information is stolen from a data hack, that typically is reported. But if an industrial asset is taken offline and you lose your power, that's a different situation altogether. Yeah, and this is a very important thrust of policy right now because the struggle that we're grappling with is a patchwork of regulations even across the states which is various levels of responsibility to report. There are industry consortia working with the government to come up with codes of conduct and models for information sharing. I mean, the core of the idea is common threat vectors. Each of us might be hit in our own way, but the sharing of that information would allow us to be collectively stronger. And so that is happening, but the legal framework for this, frankly, Congress has to institutionalize the mechanisms for liability protection and others that have been central to getting a cybersecurity framework in place. We're not there. And I do think that this IoT revolution is really bringing that to the fore. And this isn't just good and the right thing for consumers. I would argue it's also the right thing for businesses to the points of making all of our ecosystem sort of healthier and stronger. Businesses that choose to deal with problems through being opaque are not going to have a good time in the future because as we're seeing more and more, the information wants to be free. The sunlight is going to get shined upon just about all of us. And so it's so important to take that initiative and transparency and trust are really, really critical. And in terms of shining sunlight, I mean, do you think that we're talking about the problem of these breaches and information sharing? Do you think it's that some sectors are really lagging behind? Like as I understand it, the financial sector has been making a lot of headway in sharing information around breaches and actually, as they get smart, also collaborate. Your thoughts on other sectors and sort of where do you see glaring problems like, oh my God, come on, let's move on here already. Well, one way to think about this is that the regulated sectors tend to have a safer place to have that conversation about what is or is not required. Part of the reason the financial services industry is grappling with this is that their regulator is demanding that they take better control of their data. In healthcare, that's a similar framework. In education, that's a similar framework. So what that means is what's not regulated tends to be where you don't quite appreciate. It may not be that the particular device that you're smart enabling is itself risky, but the mashup of that data with others could lead to unintended consequences. And so the mashup is where the anxiety mostly sits. I think that's a great point. I mean, the creating value using IoT technology is a bit of a buzzword, but I don't have a better one is very often turns on the ability to create a relatively and generally much more complex ecosystem of partners than companies have had to deal with in the past. And so all of a sudden, all the other cliches come to the fore, every chain is only as strong as its weakest link and so on. And very often you may not even know who all is involved in allowing you to create value from the new sources of data that you brought on stream and hard to fix problems you don't know you have. We are seeing, we just did a review with our SecureWorks guys the other day, we're seeing a lot more operating technology focus on the breaches and that's, I think, one of the biggest concerns we have about IoT, right? And I was gonna touch on that, that I think it's just a lack of awareness. I don't believe there's a resistance in the industrial world. There's a belief currently that the titanium doors on the front of the house are protecting the open door in the back of the house. In the back of the house is where those assets reside. So as we meet with industrials, there's a beginning awareness that their assets are in peril. Their data is relatively protected and so this conversation has just started but I do think it's going to take some regulatory requirements as we go through very much like the enterprise side of the house. I think this is particularly obvious in the health sector because I think what's holding us back is not the art of the possible with technology. We can do some quite extraordinary things but people are very low to adopt the technology until you can legislate for privacy, until you can control how people are gonna use the data and the purpose for which they'll apply the data. To your point with the mash-up, it will go in many different directions and the incumbent profession is naturally going to be defensive but it will feel under attack from innovators on the outside. And we can't release the floodgates of good ideas until we can reassure the public. One tweet to Alexis but I think it's open to anyone who wants to respond is how do we get millennials, young people to embrace wearable devices? And so do you have the golden nugget solution answer that? Or do you also have hesitation around that? Go. I would have invested in it if I had it. I think it's gonna make it, it's gonna be figuring out how to make the experience seamless and valuable. And I think to points that have voted and raised on the panel, there's also this, I think I'm misconception about millennials speaking on behalf of all of them that we don't care about privacy just because we're willing to tweet our location everywhere or what have you. But it's a misconception because we actually do really care about it. We use these things to overshare because we are opting into it. We are deciding to say here is where I am. We really do care. A big part of why Snapchat's successful is because of the ephemerality of it. And so I think it's really, really important that for all these tools that we're building, yes, I want there to be a seamless way to track all those things about my life but I want as a consumer to feel in control of it, you know exactly what I'm giving and have the ability at any time to flip that switch. I also think there's a value element. So I have four millennials, my four kids, all millennials. That's a lot of millennials in one house, by the way. Sorry. I know. And they just are trying to figure out what's really valuable. So everybody's inventing these wearables but is it really worth tying it on your shoe or is it really worth wearing around your wrist all the time? And it's only when the answer is absolutely yes, do they do it? So we've been through lots of wearables in our house and only a few stick and not for very long so far. Yeah, I think a revolution in the health space is probably not going to come from that industry. I think we might be surprised that people from the fashion industry, the literally the wearables, they might actually crack the code because I have seen and we are working with some people who you'd be really surprised what they're trying to put into everyday clothing but they're trying to crack the design challenge of what the people think is meaningful insight when you really just want to wear a shirt. Well, just again, to stick in healthcare for context, 5% of the population accounts for like half the expenditure. Many of them are older, chronically ill. A prescription for a vest that will monitor whether you're going to have a heart attack will be used and it will be valuable. It'll help them live longer. It'll lower costs in the long run for the system. So there are plenty of IoT applications that are not targeted to the healthy millennials at least on the health side. And I do think that that is going to be part of the excitement here is to say, what's the problem we're trying to solve? And now there's a new set of capabilities to tackle it. Not just sort of the fitness junkies. Absolutely, because at the end of the day, we are all about the condition that we're in is changing. And if we can identify that change in condition sooner, we might be able to intervene before I end up in the emergency room with a few seconds to live with a knife on my chest for an open heart surgery. So the millennials have put the wearables on their parents. That in fact, it is an area of great interest and opportunity. All right. I might want to just if I get on this, on this general topic about the constraints, the privacy and security and the OT risks. There's a new chapter of this, which is that as society begins to open up new infrastructure, it becomes data feeds for us to build even more valuable products and services where by design, it's not a privacy risk. The mayor of Louisville, Kentucky was able to cut the concentration of particulate matter, 2.5, reducing asthmatic attacks by 27% in a region. Why? He's now got GPS chips on asthma inhalers opt-in by the community to record the time and location of asthma attacks so that their limited environmental staff can focus where needed. They're now, they've worked with the EPA to tweet changes in air quality for every EPA air monitor in the city. So right now, you can subscribe on Twitter to air quality baits and you can figure out that this little neighborhood, today it's an unhealthy pollen counter, whatever. And so the internet of things may become a new open data asset, which is not about risk and privacy but an opportunity for folks to build off of. Anisha brings up a question also on Twitter is you've just pointed out an example of local and state governments beginning to embrace the power of data and how it can be used to solve problems. There is a very deep cultural challenge in organizations, public and private sector, to using data. So one of the questions from our audience is how do you help an organization work through that? When the capacity to collect data is here but the willingness to actually let it lead in any decision making is here. Is that an issue? I don't want to dive in on this too soon but this is essentially what the president had done in creating the chief technology officer. This is not about tech leading the conversation at all. It was the president saying I am declaring from this day forward we're moving the default position from closed to open. That was the reuse of data was far more important than its primary use in the sense of how we could leverage this national asset. And the cultural change was a lot more important than whether or not we found a new data set that we could play with that hadn't been seen before. It was the cultural point and frankly that's a leadership judgment. So the president calling out cabinet members who had done innovative applications with technology and giving them their shout out was a really important leadership moment that really built the culture change necessary for the techies to have a shot to play. And I also think we're seeing really good developments in the entrepreneurial community. New companies that are starting up that are building analytics capability using all this publicly available data and they're drawing correlations and they're figuring out how to use it for purposes far different from the original intent I'm sure. But those are demonstrating the art of the possible again to companies who use them. I think you have to be able to show value. Value is the driver of a lot of things even including wearables is a giving value. So from a corporate perspective you know we had this awakening that said we are creating all this data and are we using it? And when the answer was no we started to look at our business processes and said well we could continue to invest hundreds of millions of dollars developing the new metal for a gas turbine to run a half percent efficient or more efficient or we could start looking at all this data that allows that gas turbine to work within the system that it's in and generate 20 percent efficiency. Oh wow. And those are the epiphanies that corporations like GE are having. I don't know if it's a lack of willingness to act as and tell me if you agree with it or not but an opportunity to see how to act. So we've been lots of talks about lots of talking about sensors not so much talking about actuators right how you actually do something with it. So to take the two extremes you've got operational tech where you've got pipelines and drill rigs that you can shut down like that and then you've got wearables where you can't get people to change their behavior. Right so in one application the actuation is almost too automatic because it creates all these risks and in the other application it's not nearly automatic enough because people don't change for nothing and so you can put sensors all over them but if they're not actually going to change how they behave I guess nothing's gonna change. But I think we get frustrated going after the data first because it's overwhelming the amount of data that's out there that's not used. When you start with the problem you're trying to solve and then understand whether the data is available I think that's when real value is dropped. I think that's really key because should some things just not be connected? I wonder about that. Like you know I interview a lot of cybersecurity experts and whenever a breach happens they always talk about the attack surface we're increasing the attack surface oh my god the attack surface and I feel like there's a whole lot of attack surface. Absolutely I mean there are definitely some things that should not be connected. Like what? So there are plenty of industrial environments where the production line is going and it's working perfectly fine and you don't need to connect every single sensor to the internet that's operating that production line. You do not need to do that. I think there are also plenty of- I would disagree but we'll yeah. Okay well let's argue about it. Yes. Yeah exactly. I think you want that in a closed system I don't think you need to connect that to the internet by any stretch of the imagination. You might want it in an intranet so you can work it in and try to drive insights from the data but I don't think you want that necessarily connected outside. I would think it would be a part of what it's attached to so the best discussion with the elderly perfect but you wouldn't want a pacemaker online. You hack into a pacemaker that's a bad day. So things implantables and things that are derived with human life I think are key. But you know the- You hack into a production line it's a bad day too. But it is and that's where security comes in and that's why I keep talking about the OT space and the existence of a company like World Tech is those have to be protected for the industrial internet to see the light of day. So it's an advancement that needs to occur. Are we ready? Probably not particularly from us. Can we agree that diapers should not have sensors in them? That is actually the perfect transition into my next question. I actually you really hit the nail on the head. One area and not joking one area that I feel is under reported certainly I have not reported on is parenting at IOT. Because when you think about again the incentives who is motivated to monitor? I mean it's often parents who care about their young children more than anything provides a huge motivation to monitor. And I don't know that we've really talked about you know we talked a little bit about law and policy there's also just culture and what people will adapt and what people will do in their day to day what are acceptable norms. Have you guys as parents or people who know people raising kids have you thought about like okay you know should there be a monitor in the teddy bear that your daughter hugs to make sure her heart rates okay when she hugs the teddy bear what cause she might be sick should there be something in the diaper because maybe there's some loose motions going on. I don't know you know like I'm thinking crying works. I think when they cry that means you should change the diaper I don't think it needs to be in the diaper. I do think there are certain again health care applications that can make a lot of sense if a kid's you know got some kind of sleep issues and apnea that kind of thing absolutely I'd be all over that. Every day monitoring I guess it's probably a choice and as I said I have four kids probably the first one I would have wanted to monitor a lot but the fourth one you don't want to monitor at all. That's in hindsight. And there's some applause. Other thoughts on that because there are lots of kids and babies give you a great example because there's no pattern to kids across the populations. We do quite a bit of work and intensive care and every single kid has its own pattern of recognition that there is no normal. So that's a great example of where ultra sensing very complex analytics can give you foresight to when an intervention is required. My own personal thought raising kids is I wonder whether they've been educated properly to be able to perform in the future world that we're creating because they're being trained to do predictable things predictive rule based mathematics and engineering. I don't think that is the really the future. I think we'll be surrounded by systems that will be at least as clever as us. So I worry about how they create value and survive in that future where we've codified a lot of those predictable professions. We're already doing it in the financial sector the legal sector. I think medicals next. I wonder what the future will be like and are we educating our kids in an appropriate manner to adapt to that future. And your thought I'm just chilling I'm a little bit lost drill in a bit. Your thoughts on remote surveillance for children. Personally I don't want to do remote surveillance for privacy on my kids. I'd rather not know what they're doing. I can usually tell I know what they're buying on Amazon because I get the angry birds confirmation. Yeah I'll let them be free. I think kids should be free. It's a personal view. The challenge in that again is there are plenty of problems to solve in the U.S. alone infant mortality is abysmal. And monitoring the health of a pregnant mother in the home might be a tremendously important way of identifying a potential at risk birth and to intervene with a high risk OBGYN earlier in the cause. We ran a small experiment of this in a rural part of Virginia and we were able to cut preterm birth rates 25% by doing a much more active job monitoring populations that were deemed high risk and they never had access. You know they can't see the doctor all the time and especially in the rural areas there are no high risk OBGYN. So it's a tough one to say you should or shouldn't build the technology because it's kind of creepy or what have you. There are applications that would see if those capabilities were affordable and available would be widely deployed lives would be saved in a whole number of applications. Yeah I agree. I see from Twitter people saying the idea of IOT with parenting is a nightmare idea. And so just to close off on that any defense of it or echoing of that sentiment. Well yeah you could use a drone to helicopter parent it'd be perfect. Yeah. One very dramatic way that we're seeing IOT is it's hitting the roads right. I mean it is actually remarkable and maybe people who have been involved in the industry for a while. It's remarkable how quickly vehicles have become smart. It's just it's extraordinary. One issue though is they can't necessarily talk to each other. So as IOT develops particularly in cars do I have to worry that the Ford and the Google car and the Honda are not going to be able to speak to each other. Is that a real issue right now. I would say yes. I would say generally I think in the world of IOT standards are not yet widely understood. That's a little bit the Wild Wild West out there. And I think that's prohibiting adoption for sure. And I think it's certainly prohibiting efficient adoption. So I would say that's true absolutely in the industrial space as well. And we would like to see those those and we have an obligation I think as Dell to help promote those and help set those and help establish those. But we're probably several years away I would say from from real standards generally in IOT and common standards by definition mean better security. So it's by having everyone with different standards difference connectivity patterns and different maturities in their security is what allows hackers to be successful because it gives them a multitude of attack factors. Exactly. I'll make one observation. This is an argument in favor of a lower case G government playing the convening role because in many of these circumstances there is an angle where there is a either a missing role for a convener or there's a regulator that may get this wrong. And so a mechanism by which industry can collaborate maybe with some nudging and some convening we could actually get to a better place. Interoperability as much as privacy and security or concerns and constraints interoperability will be fuel for a lot of these applications and that will come about in part with the convening and many folks can play the convening role but I do believe especially in regulated sectors you're going to say a lower case G addendum to the regulator who can maybe help get folks to agree on some common common standards. I think there's a good parallel here when the mobile internet started going back in 1999 very quickly the industry which was dominated by telecoms agreed on a global standard. I don't see the same in the IOT and I'd like to see that level of interoperability accelerated because it will free up all of these opportunities that we're describing here. And you know one last question I have then to close it up is is machine learning a big part of the value proposition of internet of things. So it seems like at least within companies like in driving the Tesla autonomous vehicle last week something I was surprised by was Elon Musk describing that the Tesla gets better at driving not just through over the air software updates but by learning from the network of cars that are part of this data set that are continually mining for lessons and I hadn't thought about that before. I think that's crucial. I mean I think that's how information creates value right in an industrial setting when you're pushing atoms around it's basically a linear value chain. You start with raw materials you finish with a funny finished product job done when it's information it's all about going around a loop right the actuators change your behaviors as a result of that you create new and different data and on and on you go. So I think it's a defining characteristic this entire space. And I would say the action that has to happen after you learn something is really important because if we have to intervene as humans every single time to get something done then the efficiencies that we're counting on just disappear. Machine learning is the only way that these pieces of equipment know that they're in a bigger system. And so most equipment is optimized to run alone or run at its highest optimal relative to itself. But cars need to be able to self learn that there's other cars. I think that's crucial to get the real value out to the IOT you have to optimize the system as a whole not optimize locally. There's no point taking a measurement just because you can it should lead to actionable intelligence. And the machine learning goes from actionable to predictive to prescriptive. So actually the machine tells you what the winning moves will be in a race or in operation. And so that's got to be the goal. But for equipment to learn it has to be exposed. Yes. And so security becomes the big factor again. It is it is ironic though that Elon who is so it is very ironic. But I but I do agree this is the way to do it. And and you know ultimately we do risk our cars also developing road rage. But at least at least we don't have to worry about that. I actually have to say it was interesting driving the Model S because we hit the highway. It took over and I actually wanted to change lanes and so I hit my my my signal to change lanes and there was a car accelerating behind. And so I totally would have cut that guy off. Like I would have done it if it were me driving but the car wouldn't. And so the car actually did this little like hover this way hover back weight. And so it's like a very polite kind of thing. And I know that can be calibrated. It could be more aggressive. It could be more of like a New York cabbie whatever. But angry mode. But it didn't start that way. You know and that impressed me about it. They are a little annoying to drive behind. They they come to a full and complete stop at every stop sign as do you. Only behind those cars. And it's machine learning and different. I mean we're talking quite a bit about about Tesla's car. But are you seeing this happen in IOT. I guess you know on the one hand there's this problem of in you know the company walls being very strong and corporations not wanting to or not being able to share their data sets across entities. But within it seems like there's some very vigorous mining going on. Oh yeah. I mean there's loads of applications where this applies. If you look at the air traffic control anywhere in the world actually the poor operators react as the data unfolds as the events unfold because they work with printed schedules. Why can't we have systems as we now do that can suggest what interventions to make in order to get more passengers out on time or better through ports through an airport. This is novel. It's only just coming online. The military have it. The civilian people don't have it. Anyone who flies into London half of you will do two laps over London before you land. That's because there's no code advising how to do the scheduling. It's totally predictable. We know when the planes have left we could slow a plane down and avoid it having to do the two laps. But it's that type of thinking that's missing. It is going to create a whole new class of concern around the regulation of the algorithm in the sense of what is it attempting to accomplish. And so you'll start to see this in judgments because what is it optimizing and how disclosure affect that particular outcome. And recommendations engines are going to be critical in most of these sectors but recommendations to what end. And that I think is going to be another frontier we're going to be grappling with as this evolves. Personally I think so long as we keep the human in the loop I'm a big advocate of keep the human in the loop we'll get the benefits of their judgment, experience plus the benefit of the collective intelligence of the recommendation engine. I'm not rushing to full autonomy in any sense. And do you think with the airport example that kind of algorithm could help us to cut down a very annoying and environmentally wasteful experience but are there many barriers to that actually being implemented? We actually did the proof of concept four years ago it's still not gone live even though we can show 20% increase in efficiency in the airport. So yeah culturally there's a lot of resistance people are proud of their ability to react and get it right most of the time we can demonstrate that the code can do it better but that's a threat to people so we call it the second screen we're not replacing the dashboards people grew up with and got trained on we're introducing it subtly non-threateningly and we're keeping the human in the loop and it's only when they start to trust it because it resonates with their intuition that's when people start embracing this tagline. That's an amazing tagline keep the human in the loop. Yeah. But you know on that one we are I mean the FAA is definitely working on we had a pilot literally in Houston a pilot program to basically accelerate this optimized profile descent specifically because you're taking the human judgment out you're like cutting waste 30, 40% if you can standardize through algorithms how to land the planes and it's funny the regulation is how do you get the one approved so that you could actually have that be available to all it is a lot of opportunity but it's gonna see this evolution. So it's interesting this point about how so much of what technology gets adopted is our emotional comfort with it as we go along so with that that's the end of our closing Plenary at Dell World on Internet of Things thank you to all of you for joining us and to you for listening and tweeting and some great questions. Thank you. Please welcome Senior Vice President and Chief Marketing Officer Karen Kintos. Great panel, great discussion. We know that your most valuable resource is your time so we absolutely appreciate you being here at Dell World and letting us share our story with you and what I hope you came away with loud and clear is that our story is really the combined stories of all of you and what you accomplish each and every day. You stole the show and our solution showcase and our breakouts and in our general session keynotes and we would not have had it any other way. So it's entirely fitting that we end Dell World this year with just one more customer story. This year we launched for the first time ever our Customer Impact Awards. We are recognizing customers who are innovating in the areas of cloud, the Internet of Things, big data, security and mobility. And we had some terrific nominees and our category award winners this year were Owens and Miner in cloud, ELM Energy in the Internet of Things, Clear Creek ISD and Societal General and Mobility, Insecurity Robo Bank and in big data the University of Florida. Some of these customers are actually here with us in the audience and I'd ask them to stand up and let's give them a round of applause. So hopefully you have seen their stories and read their stories. Each are an incredible examples of innovation and problem solving. But we also asked each of you to get involved in the action side of this also and voting for the number one award winner around our People's Choice Award. And you did just that with more than 7,000 votes casted. So thank you for participating in this. So without further ado, I'd like to announce our People's Choice Award winner, which is Clear Creek Independent School District. Serves more than 41,000 students in prides itself and being the leader in visionary education. It's a district wide one-on-one student device initiative. It's diverse range of innovative technology solutions are truly making our students in classrooms and teachers and the whole community future ready. We have a number of folks here from the Clear Creek School. We'd like to invite you to come up. Dr. Greg Smith, the superintendent, the chief technology officer, Kevin Schwartz are here. Michael, if you join me on stage, we'll do another big, huge appreciation for all of the great work. Thank you. Congratulations. Well done. Congratulations. I think they want you over here.