 The Cube at IBM Impact 2014 is brought to you by headline sponsor IBM. Here are your hosts, John Furrier and Paul Gillan. Okay, welcome back everyone. We're here live in Las Vegas for IBM Impact. This is Silicon Angles theCUBE, our flagship program where we go out to the events, extract the signal from the noise. I'm John Furrier, the host with my co-host Paul Gillan and our special guest is Judith Horowitz, an associate's welcome back to the Cube alumni. Again, thanks for coming back on. Thank you. This is a special addition because Paul Gillan and yourself have a lot of history to withdraw from. And so we will draw from that history in the computer business. It's 2014 and it feels like we're back in the 80s again, back to the 80s, a time machine, hot tub time machine. Your name frames are cool again, I think. So, it is like the hot tub time machine. It feels like we're back in the data processing era. I mean, how many repurchase reports have you done back then about getting the data, decision support, you know? I mean, how far off is it? I mean, it seems to be similar. Well, I think it's similar because we're dealing with the same issues and the same aspirations that we had 20 or 30 years ago. People always wanted to do this stuff, but we didn't have the infrastructure. We didn't have the... The cost of it, right? Data technology was not cost-effective. People wanted to do this, but you would need so much computing power that people gave up, that you just couldn't do it. So I think if you look back in history to what people were talking about in the 50s and 1960s about what they wanted to do with artificial intelligence and machine learning and how there would be all these data rules and you would have machines that could think and to anticipate what people wanted to do, people were talking about that in the 50s and 60s, but the technology infrastructure just wasn't there. So you guys, you've been documenting, and Paul, you've been covering with computer world back in the day. I mean, all these major inflection points. And a lot of times you get shoveled a lot of vendor, marketing and like, here's all we're going to do this. And a lot of the time it's just to keep up. Someone's just trying to keep up. Some are holding on with their nails, rocket ship have changed. Some are really ahead of the curve. How would you guys peg IBM right now in that category? I mean, obviously they're not going anyway. 50th anniversary of the mainframe was just here. Are they holding on with their nails? Do they actually have a good position? What's your take on that? So my take is that IBM is actually at a really important inflection point. If you look back in history to when Gershner came into IBM, IBM was at another inflection point where they had a lot of good technology, but it was not packaged and reflective of where the marketplace was going. And I actually think that we're back to that, that era with IBM right now, where IBM always had a lot of critical technology, but it wasn't packaged and it wasn't designed for where the market was going. So I think what you see with Ginny Rometti coming in is IBM beginning to refocus and to sort of get focused on really the key technologies that are the future. So I think IBM is doing the right thing. I think that this is a very complicated transition for IBM because they are really having to get very structured and very oriented around bringing elements together that used to be their own little fiefdoms and now they're really coming together. That's a hard transition for any company. But it's a good point you make going back to the future. If you refer to that last inflection point, if you remember in the early 90s after IBM nearly went out of business, a lot of people were talking about break up the company. They had 40 different brands and a lot of great technology with no coherence to it. And then a few years ago, they come out with this vision, smarter computing and smarter planet. And that really has held together pretty well over time. And the whole theme of this conference about bringing together cloud analytics, mobility, social, still plays well to that message. Yeah, I absolutely agree with you. And I think what you saw with smarter planet was sort of the beginning of framing this message around data, around collaboration, around context, around applying this technology to the real world of business and environment and just the way we live as human beings. So I think it set the stage and then what happened when that message first came out was people said, I like this idea. What do I do first? So being able to go from that grand vision, which was great to say, okay, so I always like to talk about what do I do when I wake up tomorrow morning? And that's what they've been filling in. What do I do? How do I get this data to be smarter? How do I bring it together? How do I create context? How do I have the type of environment that scales up and scales down based on what I'm trying to do? Of course, they say the devil is in the details, of course. And here you have a company, IBM, that has done, I believe, over 130 acquisitions over the last five years. And historically, not a company that's done a lot of acquisitions. They've done some very big ones. How important or how good a job do you think at this point IBM is doing at digesting those acquisitions and really taking advantage of the technology and the people they acquire? I think that they've done a pretty good job. I mean, obviously it is difficult acquiring companies, but I think because what IBM has really done well is they have a roadmap for the technologies that are part of this uber strategy. And what they do is they look to see where they have holds or where they have a technology that isn't delivering for the future and that's where they do acquisition. So if you look, for example, at soft layers, I mean, they had cloud technology and they went to market with it and then they discovered that in order to get to the level they wanted, they needed the type of infrastructure and technology the soft layer had. So I think that that was a pretty gutsy move because a lot of companies that had already developed a lot of technology, went to market, were offering this technology. They had somebody like Danny Saba leading this charge in the cloud from a technology perspective, took a step back and said, okay, this works, this doesn't, this is what we need. And I think that that's the sign that IBM is really looking at acquisitions that really make a difference. I mean, you could talk about hundreds of them and where they fit and why they were, if you look at the technology, for example, that IBM has purchased in security, phenomenal set of technology, Q-Radar, a fabulous acquisition. If you look at a lot of these, they are really foundational technologies that make the difference between customers being happy and unhappy. Judith, I want to just get your take on the big picture. So here's the picture of the whole map solutions, Watson foundations, big data analytics, you know it's strange to what they've got going on there. And Watson certainly has got some great sex appeal for the geeks out there and there's a lot of innovation around that. It's the brains behind the future kind of world of data connected, all that stuff, internet things. So I got to get your take. Integration is a huge issue on these acquisitions. So like, if you look at web 1.0, Cisco did a ton of acquisitions and they were criticized for not being really building that platform. Is IBM good at acquisitions? And where are they on the integration? Is it a should you item, do you see a big to do item? Do they have a good discipline? So I think what people don't understand about acquisitions, the integration, it takes probably 20 times as long as anyone expects. It's really hard because not only do you have some great technologies, but they have to work with the legacy that's already there. And you have to create APIs and everything has to be modular and designed so that you can connect things that have different design bases. It's really hard to do. I think what's good about IBM is they've been very diligent about it, but I think it's taken them much longer than what they anticipated, but slowly but surely they're doing it. And it's different in different areas. I think they probably moved faster, for example, in security than they have in some other areas. Yeah, I mean also too, this gets complicated when you overlay open source and this new openness framework because it's not just the internal stuff that they're used to, they got to actually keep track of the fast moving software world. Well, IBM has actually been working with open source for quite a while, going back to Eclipse. It's use of Unix and Linux. So this is not something new to IBM has probably been going on, I'd say probably for the past 20 years, they've donated technology to the Apache Foundation. They've moved from some of their proprietary technology to some of the, you know. And they have a heritage in open source, they will document it. The question is how relevant is it? Are they modern? Are they up to date? Are developers in the IBM ecosystem the new hot DevOps guys that you're seeing Amazon require? I think that that's what Bluemix is about. I think that's why they're doing Bluemix. I like the approach a lot. It's taking them a while to figure out what they need to have in terms of technology to really attract the next generation developers. I think they've got a shot at it with Bluemix, but it's hard to really get those developers to say, you know, IBM is the hot, cool platform. Yeah, I mean they're scratching their head looking for value, but we'll see. IBM has yet to punch that through, so we'll see how they do. We'll keep track of that. I got to ask you about what's going on with cognitive, because at all the events, we love talking about Watson because that's like, you win at Jeopardy, that means you can do a bunch of other things. We're hearing about streaming a lot of cool stuff going on that Watson could be a big part of, although it's one element of a role portfolio, but what's your take on the cognitive vision? Could you share what you've learned? Yeah, so we're actually in the middle of writing a book on cognitive computing. Oh, how ironic. So spending way too much time on that topic. It's a really fascinating area and I do think that it really will change everything because it's not just about a game about Jeopardy. Jeopardy, in fact, was a proof of concept. It was a grand challenge very much like the chess game was a grand challenge, but what you got out of the grand challenge was the idea that you can have a learning system, that you can feed in massive amounts of information, have that information be used to learn in collaboration with subject matter experts, creating ontologies and basically views of specific areas of whether it's a city, whether it's treatment of a specific disease. One of the issues if you just look at oncology and the type of work that's going on with a number of the initiatives that IBM is collaborating with in healthcare is looking at the vast amount of information out there around treatments of different cancers. So one oncologist can't possibly understand and have access to all of the new research and look at all the results and know that there's a new clinical trial that may have relevance to somebody they're treating. Or analyzing DNA and looking for patterns in DNA structures that may indicate the likelihood of cancer. I was fascinated by Watson after the Jeopardy experience and I thought there's so much potential here and it seemed like for a year, Watson almost went underground. There was almost no talk of it and now IBM is sort of bringing it out again. But there's not been a lot, it seems to me that IBM has not done all it could with evangelizing the potential of that technology. And the reason I think is because, if you look at what you have to deal with in terms of the underlying technology for this, you have to understand natural language processing. You have to understand the context and it's not just regular old natural language processing but it's looking at how words and context is related across the vast field and then how much can you trust data? So it's advanced analytics, it's big data, it's machine learning, it's natural language processing. You can go on and on. Do you have the right ontologies and taxonomies? So can you feed in the right data at the right time? So it's really complicated stuff. So if they had come out at that point and done a major marketing of it, it would have been too early. It's probably- There would have been anything there to sell. Yeah, it's probably too early right now because this is something I think 10 years from now, this will be really the foundation of how we do computing. It is that transformation. Do you think it is that far ahead of the market or do you see as someone who's studying this area now, do you see other technologies from other companies that are Watson-like in their capacity? Yeah, I do see other emerging companies that are really leveraging both natural language processing and machine learning to begin to, and what's interesting is a lot of them are very solutions-focused and solutions-oriented. So I think that this, because one of the problems with the way we've traditionally developed technology, it's always fit for purpose. Okay, I've got this specific problem. Here are my rules. Here's my data. Okay, I solve for that. But you're always looking in the rear view mirror. You're always solving what the biggest problem for the market was three years ago. The difference with a cognitive learning system is that it morphs and changes because problems don't stay the same. So it's actually not programmed, and that's the real difference. So that's why I think it's so fundamental. We're going to get to something that you don't have to say, okay, this is my story. I'm going to write a beginning, middle, and an end, and I'm done. Well, that's why these systems today are so complicated to build because you have to anticipate everything, and we can't do that. And you're really talking about predictive analytics. You're talking about the ability to forecast the future based upon what we've seen in the past. Predictive analytics and advanced analytics is a core part of a cognitive system. Is that the real big opportunity in analytics is predictive rather than historical? Well, you actually need both. So you need predictive, and you need some of this streaming data and looking at things at real time, but you also have to look at historically. So when I have a good working system, what did it look like? How did it act? So I then have to compare the historical data to what's happening right now. So if it looks like it's meeting the norms and nothing else has changed, I know I'm in the right ballpark. If on the other hand, if what was normal before over the last six months, this looks totally different. Either I've really improved performance, so I'm doing something right, or something is desperately wrong. Looking out at the industry landscape, we see HP clearly in a lot of trouble right now. We see Dell taking itself private. We see Oracle thrashing in the hardware market. I mean, the big players who have dominated this industry for many years, pretty much all having trouble. IBM seems to be an exception. Is this industry devolving into much smaller specialty players, or is there a role for the big companies like IBM? I think there is definitely a role for the big companies, but they have to, the big companies have to take more risks. I think one reason that Dell went private is so it could take more risk. Would a public market have allowed Dell to move away from commodity PCs? I don't think so, because you're- It would become compact. Yeah, you've got revenue, but not profit. And people get scared of giving up revenue, even though it really is draining the company. So- What about VestaBread? I mean, that's always been kind of like, in those kind of like trough between inflection points. People get VestaBread, okay, I'm the VestaBread server. VestaBread is now, if you believe the cognitive thing will be transformative, which I believe, I agree, that levels the field in terms of now, or does it? I mean, that's my question, is that- Does cognitive level? If things like cognitive continue to arrive on the scene, it's going to disrupt the status quo. So, will the VestaBread mindset continue to be that Dell's taking more chances? I can see them saying, hey, let's build a mobile app. Maybe, I don't know, they mobile phone, who knows? Obviously, Michael's got some plans. So, how do people like IBM HP, do they maintain that we're VestaBread, or does it go away more- Well, I actually don't think the large companies do necessarily VestaBread. They do VestaSolution. So, they invest in VestaBread in order to create things that link elements together. So, this idea around creating composites is really the future, and I think that's what we started with, with service-oriented architectures. You like the composite apps vision. I do, that's where we started with service-oriented architectures. You had business services, you linked them together to create value. That's what we've been, none of these new technologies happen in a year and then go away. All right, so I got to ask you the question. You might have hit it, but what's around the corner? What are people missing in the landscape? Obviously, we know with big data, cloud social, it's also happening. But what, from your perspective, seeing, drawing on some of the historical views we were just talking about, 80s, 90s, 2000s, what's your perspective? What's the blind spot for the vendors, the industry? What's around the corner from your perspective? Well, you know, I think a lot of it is still, I think we are only sort of at the 1% mark with things like cognitive, big data, predictive and prescriptive analytics, is how do you really use data? If, for example, you have the whole area of probes and been able to collect data in real time and then use that in an anticipatory way to figure out what's going to happen next, I don't think that right now we're looking at things that are totally different, but I think that we're about to get into an absorption phase. We've got, you know, how do you do prescriptive things that tell you not just, okay, this is what's going to happen, but if that's happening, what do I actually do about it? That's where a prescriptive comes in. What do I do about all that dark data, you know, log data that I've been collecting for 30 years, what do I do with that? Dark side of the moon kind of stuff, it's a little bit of a Pink Floyd kind of metaphor. How do I really create cognitive systems and how much is it human intervention versus how much is it, you know, do I let a machine just tell me what drug to prescribe or is it a collaboration? I think there's going to be some new stuff that's going to pop out of the woodwork that's going to catch people off guard and we'll try and track that. But I think that new stuff will come out of what we learn from all this. Yeah, it's hard to predict. So even at best predictive analytics, you can't predict the next black swan. This is theCUBE. We're trying our best to predict the future. We'll be right back here at IBM Impact right after this short break.