 What do you make of that? Well, I think it's great. For me, I always observe technology from the cheap seats. I'm just a trend spotter, and I kind of noticed things. And I noticed a long time ago that data was going to be increasingly the focus of competitive advantage. It was really through a process of analogy looking at what happened with the IBM PC. There was this idea that hardware was the source of lock-in in the IBM era. And when they introduced the PC, they thought hardware was still the source of lock-in. But yet they introduced this commodity PC, which completely destabilized the industry. They signed away their future to Microsoft because they didn't realize that software was going to become valuable as hardware became a commodity. So I started asking myself, what happens when software becomes a commodity? And I saw that happening through open source software and the open protocols of the internet. So I started asking myself, what's the next source of competitive advantage? And I came to the conclusion that it was large databases generated through collective action over the internet. And of course, you see that. And that was really the heart of my thinking about Web 2.0. But what's so exciting about this conference is you now have real practitioners coming forward and saying this stuff and talking about how they're doing it. It's real world mainstream. Yeah, real world mainstream. We were here yesterday and it was palpable. And we're starting to take it away from just the consumer internet and areas like finance and starting to look at the potential impact in areas, for example, like health care, where there's a massive problem that could be handled very differently if we built the kind of feedback loops that we see in today's applications. So with all the trend type and the megatrends of conversion networking, you talk about some of the trends over the years, but conversion networking, cloud, and mobility. And those are all inside industry kind of topics. And we all kind of talk about it and know that. But is this data revolution, this data phenomenon with strata and what you guys are doing here, is this just a touch point for real people to kind of get it, like see examples? Absolutely. I think what's really made all of this concrete is the smartphone. Years ago, I started talking about this idea that we were building an internet operating system and that the subsystems of it would be databases. And it was very abstract. But now, when you pick up your smartphone and you ask for directions or you ask for the nearest Chinese restaurant, it becomes pretty obvious. Oh, yeah. That information is not on my phone. Effectively, it's reaching out to some big database over the internet. And so people don't necessarily think about it. But when you explain it, they go, oh, yeah, yeah. In the old days, think about your GPS device. You had a CD-ROM. Maybe they had the maps and you had to get a new one. Now it's like, hey, this thing is downloading on my phone in real time based on where I am in real time. And oh, and by the way, it's got real time traffic conditions. And all those things start to just be part of people's expectations. And the change, I think, is a couple things. I mean, consumers are becoming used to it. So the question then becomes, why is this not working the way I expect? But there are big industries that are lagging where there are massive amounts of data, but it's not being used correctly, often because of? Structural issues. Structural issues. Yeah, I mean, health care being a great example. We have data that could tell us what procedures work and which ones don't. And yet we still pay for the volume of procedures. Our whole reimbursement system does not differentiate between doing more of the things that work or just doing more of anything. And so there are changes in the works, for example, in the Accountable Care Act, where I think we may start to see some changes coming down eventually in Medicare reimbursement that will start to trigger marketplace changes. And I think what's really exciting to me is that we have there in health care a really big problem, something that's going to literally bankrupt the country if it's not solved. We have outmoded business models, and yet we have in this data revolution the potential to really make a difference. So we hear a lot about privacy. We've talked about it a lot, but I'd like to hear your take on it. So for instance, who owns the data? I mean, do we have to see a change in whether it's public policy? What's your angle on that? I think it's very, very hard to frame this in the way that most civil libertarians would like. And the reason is that so many of these databases become valuable only when they have all the data. So this idea that we can go back to a world in which the individual user kind of keeps his data or her data private, I think is it would be wonderful. But I don't know how we do it, because the utility comes from giving up privacy. When you want to use your mobile phone, you are telling the carrier and you're telling your location. Yeah, sure, you can say, oh, yeah, don't release my location, but then all of a sudden, you have all these applications that don't work. And so people will go, oh, well, of course, I'll release my location. So then it really becomes a matter of. Indifferent, really, to how indifferent they are. Yeah, well. To benefit, indifference to benefit, right? No, well, so I think people will give up a lot of formally private information to all kinds of people. And then, of course, it's very, very easy to, you can say, oh, well, we can anonymize certain things and so on. But you really, it's very, very easy to re-identify people. When you have enough data sources, you start to be able to pull patterns out and recognize people regardless of how much anonymization. So that makes sense. So what I think that we need to shift is to defining certain classes of harm and then penalizing those harms. Whereas what we have right now is we're trying to prevent the possibility of harm. So if you look at, for example, going back to healthcare, you look at HIPAA, the Health Regulatory Privacy. A lot of hypotheticals, yeah. Yeah, it's sort of like we must keep some kind of data privacy moat around all this information. Of course, that means that we can't do a lot of useful things with it. Meanwhile, what are we really trying to prevent? We're trying to prevent adverse selection by insurance companies. That's right, right. So wouldn't it be better to say, okay, of course, which is what the healthcare bill was trying to do, say, hey, insurance companies can't deny coverage based on pre-existing conditions. All of a sudden then it becomes much less critical to hide medical data. And of course, even then, the privacy issues, I've always thought of it a little bit overblown. Oh my God, we can't let anybody know that Aunt Jane has cancer. Meanwhile, she's talking to her church group about it. Well, it's gossip all around her neighborhood. So where I was going with that is, and I agree with that by the way, my thinking is more, I want access to my data. I feel as though I don't have a right to it. I don't have access. So should there be some kind of public policies? Great, okay, we're going to get over that whole libertarian thing, but you must give some kind of access and it's incumbent upon you as the data provider to give access to me as the consumer. Well, is that reasonable? No, I think it's reasonable, but a little bit wrongheaded. And the reason why it's wrongheaded is if you require it of people, they'll do a bad job of it. This is like the disclosures on credit card forms. And they'll send you some big complicated form that you're never going to read and you don't really, you know, or you can update your credit report, but it's such a nightmare that you're never going to do it. So if it's productized and I pay for it, maybe? Yeah, so what we have to do is to find actual applications and tools that are really useful for people when they manage their own data. And if we can do that, then we will start to change the perception. So there's a startup that we're showing, a startup showcase here about Singly, where Jeremy Miller, this guy who built Java originally, is trying to build sort of a tool set where you will download all your data from all of the social sites and so on into a local store, but it's not from this idea that now it's mine and I can take it away from these other guys because it's only useful when it's out there on the other sites. It's more like I want to build a platform in which people can build innovative interfaces and innovative tools so they can do more useful stuff with your data. And if we actually can start that kind of virtual cycle going, then the market will start to address some of these issues. It's an incentive system. So Tim, let's talk about something that's kind of a little bit different than no one's really talking about. It's been some hallway chatter between some bigger companies and startups. Obviously, the startup community is robust right now with data and we're seeing the charity with Y Combinator and all these things, giving out a lot of money and so on. But big companies are also retraining and a lot of the new technology, even for the big companies, hasn't been around five years ago. So there's a real emphasis on a workforce, whether it's entrepreneurial workforce and or IT as a service or cloud, how that's all mobile. What is your angle on that? And obviously, training is super important. But what are you seeing in terms of the trends for that? I mean, honestly, you sell books to help people get knowledgeable. I mean, it's a major issue right now. Unprecedented at the scale. I mean, the massive tsunami of data, the world's shifting pretty fast. Classic training kind of doesn't work. So is it a crowdsource model? Is it a group on for training? Is it live streamings? Things that we're doing? What's your angle on this? I think to get good at the kind of skills, it's a professional skill and people need real training. That being said, it's been happening. I remember, I think it was Mike Franklin as a UC Berkeley professor told me, must have been three years ago. He said, you know, they're teaching Hadoop to all computer science freshmen now, versus, you know, that scowl years ago and I went to school. And some other stuff in between then, Jim. And that was one of the things that put it on my radar. I was like, oh, you know, this is really coming up and coming. And I think we will, you know, the data set and the skill sets are becoming pretty obvious to people. And I think, you know, there'll be people, obviously people who are really good at it are going to be very high demand. They're already in high demand in, you know, financial services and they have been for years. I mean, is it a data jock, quant jock kind of guy? You're seeing computer science, obviously intersecting with data sets as a developer environment. We've wrote about that at Silicon Angle. And so you got computer science guys actually coding, you know, data, kind of, you know, the old model was, you know, quant jocks out in the basement, working with huge data warehouses. That's changing. What are the skills? I think it's great. I mean, there are going to be a lot of people coming from hard sciences, you know, physicists and like as they've been going into financial services. Now there's more opportunity for them. I think it's good. You know, I think, you know, you can have a shortage for a while, but eventually, you know, the market correct. So I'm not too worried about it. Meanwhile, there are plenty of opportunities for companies like mine to try to accelerate the learning curve. What do you think the cycle is on this data? It's like, we were, Dave and I were speculating earlier when we kicked off the show, you know, on the storage side it's pretty slow. It's, you know, five to 10 year horizon before, you know, platforms can be built. And there's a comment on stage by one of the presenters saying, oh, the Microsoft guy, oh, it's 90 days and we're cloud, that's like an eternity. But I mean, it's just, this is going to be a cycle that's going to have a duration. What is your view on that? Is it going to be like the classic internet dog years, kind of seven years, or it's going to be a more longer built out, or how long will this data cycle? I think it's, it's, Five years, 10 years. We're going to ramp up slowly. It's going to be a massive shift up, or is it more complex? Well, you know, I guess I would say, first off, you know, it's a continuation of, you know, the evolution of the computer industry. You know, just think back, you know, 25 years ago, maybe a little longer. It could be 30, certainly 30. Well, 30 years ago, I think about, you know, you could say a computer on every desk and in every home was an aggressive goal for Microsoft. And now we take for granted that computers everywhere, software is ubiquitous. Data-driven applications are going to become ubiquitous. But the transformation in industries is a long way to go, and some are further along than others. Do you see this as powerful as the PC client service, not a flash in the pan? Oh, well, God, no. Clearly, no. I mean, way more than clients. I mean, it's, it's, it's, it's, it's certainly as, it's bigger than the, than, I mean, it's hard to say it's bigger. It's a continuation. It's a continuation. If you look at the continuum from the PC through the internet, this is the next stage. So, and things like, you know, client server were actually evolutionary dead ends. You know, it was really the, you know, the real issue was getting ubiquitous computing, you know, where computers were everywhere. So, you know, the computer, and then the, I guess the smartphone is another big, big piece of the edge. You now have the edge. You have the edge. And then you have the network, you know, becoming the ubiquitous platform. And now it's like, okay, what happens on the network? And what happens on the network are these data-driven applications? And I think we're going to see them penetrating into more and more aspects of our lives. You know, we already see, you know, demos of self-driving cars. I see a lot more of robotics. These things, you know, these are fundamentally driven by data applications. You know, I mean, you know, we're going to see it going into new industries. There's already big industries that have been data-driven for years, but increasingly, you know, there's going to be more advantage in data. A good example is Alex Rampell, who's a really smart guy on this subject, runs a company called Trial Pay. Said to me recently, we were talking about, we'd just gone to lunch. The place was packed when we went in. And it was empty when we came out. And he said, you know what's wrong with Groupon? Is that they don't have enough of a sense of time. You know, people, that restaurant didn't want any more people at 1230. They wanted more people at 130. And so Groupon, to continue to succeed, will have to become a real time. We'll hear from the Groupon CTOs coming on from theCUBE. They'll have to become more of a real-time inventory management system for perishable time slots. You know, and you think about that, and you go, oh yeah, and so that's more real-time, more data. You know, already you have this wonderful aspect of say, group purchasing, but think about how much better that will be when that system is smarter. We're here with Tim O'Reilly inside the Cube at his conference, our rally media going on strata, making data work, making things happen. Just a final question from my standpoint, and then Davey give one final question for, I know he's got to go, Maureen's kind of hovering. You've been involved in collective intelligence and data for a long time. What's different now? I mean, what's, you know, that, because a lot of work's been done in this area, but what's the flash point right now? What makes today, this time, and going forward unique? Just quickly explain to the group. Well, I think it's really just been a, it's a tipping point where everybody is aware of that this is the frontier where value is being created. You know, when I first started talking about this, say with our Web2 events in 2005, one VC came up to me and said, well, you stop talking about this. You know, he was like, we're investing on this trend, and you know. Don't tell anyone. Don't tell anyone. He was joking, but, you know, and now everybody knows it. And so that's one part of the tipping point. Now, the other tipping point that's still, I think, in our future is, something that's already unfolding, is that there's a dark side to all this, because of course, you know, you see all the stories now about Google fighting the spammers or cybersecurity, you know, fraud and payments. There's an underbelly. There's an underbelly. An unregulated underbelly. We're going to be seeing, I think a lot of struggles over the next few years in dealing with fraud and abuse and gaming, the gaming of systems and so on. And you know, I would predict that actually some of the most valuable people in the next wave are going to be people who have experience specifically in the big data of risk. Big data risk analysis. Yeah, so my last comment, and it's really not a question, it's just an observation. It's the second time we've been on theCUBE, and you're a great storyteller. You're known as a great storyteller, but your ability to take sort of this information from Alpha Geeks and turn it into a business narrative is something that our audience loves. I've been monitoring it. We had 5,300 people on watching you, so thank you very much for coming on. We really appreciate the support. All right, thank you. Tim O'Reilly inside theCUBE. Thank you. Great to see you again. All right, take care. All right, it was a pleasure. Congratulations on a great show. Bye.