 From San Francisco, it's theCUBE, covering VMware Radio 2019, brought to you by VMware. Hi, welcome to theCUBE. Lisa Martin with John Furrier. We are in the middle of the excitement and the action at VMware Radio 2019 in San Francisco. Please welcome back to theCUBE, David Tenenhouse, the chief research officer at VMware. David, welcome back. Thank you, it's always great to have theCUBE here at Radio. And it's, we had on a really exciting day, and then suddenly this whole space opens up, and you can imagine all the innovation and the collaboration that's going on in here. This is the 15th radio. This is just one of several big programs that VMware does that really inspires and fosters this really collaborative innovative culture. You've been here for five years, and you came from Microsoft. Tell us a little bit about what makes not just radio, but VMware's culture of innovation unique and really gives it some competitive advantage in the market. Yeah, well, so I think there's a number of different things there. People are super passionate about technology. I think there's also this shared thing at VMware, which is, we're a little understated, right? We're not a big consumer brand, and we almost pride ourselves in creating technology that goes under the covers, right? So whether it's inside the data center, can we make, with virtualization, can we make it so that you can run 10 times as many virtual machines as you had physical machines, and the applications never have to know, right? So that's kind of, for us it's perfect. Technically hard problem and a little understated. So that kind of fits with our culture. I think another thing that we found, having a research group, often a challenge is researchers will go to people in the product teams and they sort of want to start the discussion. I've got this new idea, and maybe it can really help you with your product. And meanwhile, of course, the product people are, they're working against deadlines, they want to get stuff out, they don't want people derailing their agenda and their work. So something we find at VMware, which is really, I find unique, is let's say we go to a product team in many other companies and environments, and I'm really not naming any one. What happens is, you go to have a discussion with somebody who sort of is the expert on it, whatever, name your technology, and you say, usually the starting point isn't, hey, I've got this whole new way of doing your stuff, right? The starting point is, can you tell me how your stuff works? And usually the response that other companies is, why do you want to know? It's a really quite a defensive, what we find at VMware is really, people are incredibly open. I don't know exactly how this got embedded in the culture, maybe because it was a spin-off from a university, but deeply embedded in this culture is, oh yeah, let me tell you how this stuff works, and maybe you'll have a better idea. We don't even have to start with, we have a better idea. And then from there, we can have ongoing discussions about, oh, that brings up, they can improve it. That proves why you have a community. Being transparent creates openness, that creates solidarity around open concepts. Exactly, and that's kind of what you see here at radio, I don't know if people can see in the background is, this is already day four in the Expo hall, and people don't want to leave. And they're walking around, they're looking at each other's posters, they're talking to each other, making connections, and then they're going to build on those connections in the coming weeks, months, and over the next year. And this is, as I said, this has been going on for four days. You'd think that by now, people have seen all the posters, they've talked about everything. They're still finding things that they want to talk to each other about. They're kidding, the candy store has a lot to taste here, and people are collaborating, engaged, traffic content's good, and congratulations. Thank you, and I just want to add something I love is getting here actually before people arrive on the first day each year, because when they come in, it's like greeting old friends, right? It's sort of like a reunion, except nobody's worried about school reunions. People are just playing happy to see each other, so that fits with that community thing. Because sometimes they're in their teams and they don't necessarily get to see the other people. Well, you're being humble. We've talked last year about some of the content that you put together and the team, so it's a hive mind. Well, you're the chief researcher, so you've got to figure out at least some canvas to start shaping framing sets of agendas to go after. So, at least we're just talking about this the other day, about how if you have a tech canvas, you don't want to create barriers of thinking. You want to open it up and not make it too restricted. That's your job. What can you tell us about the research agenda that's here and way out there, and how do you see that aperture range of topics? Well, I think I want to first, before even getting into that agenda, reaffirm a key point you made, right? Which is don't constrain people too much. So radio, by the way, is really very bottoms up. This is not about us saying, here's the four topics. People really submit. It's a very competitive process. People want to be, not every engineer in VMware gets to come to radio, right? It's 1,800 developers, which is an incredible commitment by the company. It's still a small fraction of our community. So they're actually submitting bottoms up to see, and then we have a program committee that reviews it. So that's a bottoms up part of the process. From where I sit, what I think happens is, whether it's our research team or filtering, we'll look at what comes bottoms up and say, well, what's the signal to noise? For example, we've had this year a tremendous amount of machine learning activity, and you see this in the posters here and in the presentations. However, it wasn't too hard to detect a rising signal a couple of years ago. So in that case, a couple of years ago, we said, okay, this is important. We see it in the external community. We see it in the developer community. We see it within our own teams and developers. Clearly important. So starting a few years ago, we pulled together some of the senior most technologists, the principal engineers, a subset of them and said, hey, we want you guys to do what we call a VITESTS study. VITESTS are going faster, but also VMware technology study. We want you to actually do a strategy, but not a business strategy, a technology strategy. Look at the landscape of this. Look at where we are. Look at where we need to be and start charting a course. So in that sense, what we've, coming out of that was, for example, formation of an internal machine learning program office. One of the key goals is build the ML community. You talked about that before, inside the company. It's not just a technical goal. It's an organizational and community goal. And that's just sort of kind of one example. That wasn't the only output of that, but it's one example. And what you see, hey, big surprise, kind of 10X the engagement in the space. So that would be one case. I think one of the key things is, we'll pick up on different topics. And then a key thing that we do that I think it's different from some other companies is stop and say, what are enterprises going to need to do? At the end of the day, we're an enterprise company. And our customers are enterprises and their needs are actually, although they're in different verticals. For example, let me just use machine learning, but it could be blockchain, it could be IoT. Actually, what they need is different from, say, the machine learning that the hyperscalers need. So we realized that actually, there's a very interesting niche for us to explore. The underserved parts of machine learning, because all of these companies, if you look at them, they have a larger number of machine learning problems to work on. The hyperscalers, Facebook, Google, love them, they're actually working on a very focused set of problems. It's ad serving, it's the social network graph, it's cat photo recognition. And I don't mean to knock those, and they've got great businesses built around them, but notice it's a small number of problems. They do it at an immense scale. A given enterprise probably wants to apply machine learning to a large number of problems. They're not going to run each of those problems on a million servers. They're actually probably running those problems on tens or hundreds of VMs. What's the technology they need to address those problems? And you can go through, we looked at machine learning that way, we looked at IoT that way, and we said, you know, look, we think the analytics and the ML, those are really cool things. We want to play in that space too, but you know what, everybody's trying to do that, and not a lot of attention being paid to how our enterprise is going to secure, right, and manage all these IoT devices and the gateways to the IoT devices. So we chartered a strategy for both research and business in that space. Blockchain, same thing. Really exciting technology. Now, for enterprises, it's not about Bitcoin, it's not about currency, right? It's about decentralized trust. It's an infrastructure for decentralized trust. And effectively, think of this as a database-like thing, except now it's going to be shared across many different organizations, and it's going to change how organizations work with each other, and how they work with their auditors, and how they work with their regulators. So this is great, but let's focus on what enterprises need. Well by the way, I just retweeted while you were talking. I just got the clip from last year. I asked you that question about blockchain. You nailed it, and we talked about how all the hype and fraud and ICOs, and just confusing it again, but the world kept moving along. A lot of progress on the supply chain side. As an infrastructure for trust. Decentralized infrastructure. This is not about the ICO. Yeah, so we view that as there's the hype curve, and there'll be a deflation after the hype curve passes, but there's real signal under there, and so again, we just chart our strategy, and we keep marching down that path, and we're building up more partners and more people to work with. So it's that sort of thing. Quantum computing, right? We're not developing our own quantum computers. I can tell you that right now. And we're not even doing quantum algorithms. I have some algorithms researchers, but they're not doing quantum algorithms. I kind of wish we were doing some of that stuff, but what we did do when we looked at this, as we said, okay, hold on. A key challenge is when the quantum computers do show up, we're going to need to transition to new cryptography, to quantum-resistant or post-quantum, if two terms are used, cryptography. Enterprise customers are going to need to do that. Well, one second. They can't wait till this shows up. It takes 10 years to change your crypto. And by the way, if you've encrypted data and other people got a copy of your encrypted data, if it's long-lived data like healthcare records, you don't have them decrypting that in five or 10 years. So you've got to sort of start now. And again, this goes back to what do enterprise customers need to do? Well, okay, the new crypto standards from NIST and others aren't quite ready. Okay, but by the time they're ready, it's going to be too late to get started. Okay, but we could start working with our customers to work on crypto agility to change how they handle their cryptography. First off, get a good inventory of it and then get set up so that they're using essentially pluggable libraries so that it's easy for them to change their cryptography as soon as the standard shows up. And by the way, even if quantum computing takes a lot longer than we all think, this is good hygiene anyway. In other words, it's just a no-regress move for our customers and can we sort of help them go down that path? And this is an example where we can actually also partner with our colleagues at RSA, other parts of Dell Technologies to help make that work. But we're working with others in the industry, you know, Intel, and we've kind of convened a form of players within the industry to start working in that direction. So again, what's the cool new technology? What do enterprises need? So you talked about this event being open in terms of like the agenda and the topics being driven from the bottom up, it gets really cool. So in the spirit of talking about customers and like you were saying, designing for what enterprises need and all of the variations that that encompasses, where is customer influence? Not just at radio, but within VMware's research and innovation, programs and strategies, what's that? I mean, I customer advisory words and sound like the right kind of term. It's a great question. So like many companies, we do have various advisory bodies, right? So we bring them in and we'll sort of half, like the CTAB is our customer technical advisory body. So the more technical people in some of our kind of more leading customers and we'll show them things that we're working on under any kind of NDA arrangement and get their feedback, sort of, okay, does this make sense? If not, why not? If it does, often it's not that binary, right? It's how would you use it? And we really sort of then give that feedback back to our teams. Now, many people do this kind of thing. But we have lots of other customer engagements. We bring customers into forums like radio to be on panels and breakouts and things like that to give presentations so that basically, let's face it, in one or two events, that's not going to convey much signal to our engineers. It's a meta message to our engineers. We want you to be out talking to customers, right? So getting our engineers to be at VMworld, but we have programs to actually allow engineers and encourage them to get out, make customer visits above and beyond. And by the way, if you look at, again, our principal engineers and our fellows, I think what you'd find is the vast bulk of them are distinguished because they love engaging with customers. They don't just do it because it's part of the job. They love getting that feedback. So it actually helps them in their career. And we try to sort of essentially teach that to folks. One of the programs we have that are in the CTO office that I love, it's not in my part. So this is the case of I love all the things that we have, not just my own, it's like loving your nieces and nephews, right, not just your own children. We have this program. We were going to ask you your favorite child, so. Right, that's coming. So we have like the CTO ambassadors program, which basically is coming from the field. So we have field engineers. They're not on the development side, but these are super technical people that are out in the field touching our customers all the time. In any company, there's always a subset of those folks that just have really good intuitions for where the customers are going and are good at raising their hands about that. So we actually have a program, the CTO ambassador program, CTOA, where literally we give them a pin, right? We give them a badge. And so we've tried to identify that subset of the field engineers. And we regularly bring them in to Palo Alto or bring them together, whether it's at VMworld or radio or whatever. Again, same thing. We're going to let them know what we've got cooking. We're going to get their feedback. We're going to hear from them on, and this is not just on research, by the way, this is on the product pipelines, you know, what's going on in the roadmap and everything else. Now, to me again, that's just actually a starting point because when I put my people in front of the CTOAs, it's telling my people, this is the group of folks. When you have a new idea, don't just talk to the product people. Go find CTOAs because, you know, one of the best ways, and I'm going to be a little selfish, one of the best ways for us to influence the customers, to influence the company, is to get customers excited about something new we're doing. Right? So I, you know, people often talk about technology push. And if you really want to be successful, get innovating in a large company, you need to create pull. And so the CTOAs are great. They help us find people to do POCs with because you have to find just the right, you have to find a customer that has a need for this new stuff. But they also have to be somebody that understands, this isn't yet a product, this is a journey, right? We're going to jointly try something out. You're going to learn about whether this new tech can help you and how it can help you. We're going to learn what the product ultimately means. But you know, you're not going to be able to actually take out your checkbook at the end and get it right away. So you have to, you know, be comfortable investing the time and energy. And then they help us find these people. And that's entrepreneurship, right? It is. That is really one of the core elements that's essential to drive innovation. Absolutely. And you need that, as I said, you need that customer partnership to help fine tune things. It's, you know, one of the things more broadly I try to do with the research team is, you know, on the one hand, give them the freedom to say, hey, I have a new idea and I want to explore that new idea. That's great. Now, if you think about it, right then they're running open loop. They're running based on, you know, kind of their guests as to what, educated guests, right? And their intuition, what people might want in the future. So that's good. What I then do is say, okay, that's great. You know, you did a little bit, you wrote a paper, build a prototype. Okay, so now they get a prototype built. Okay, that starts getting this idea a little more concrete. They're okay with that. The next step, it sort of is, okay, now you got to get somebody to use that prototype. Because I need you to get, and you need you to get feedback and create a feedback loop. Because otherwise what's going to happen is, they made that first intuitive guess. So let's say they had a, they're really phenomenal and they have a 75% chance of getting it right. Okay, but that gets, but if they now continue to make a series of educated guesses and they have a, you know, 70, 80% chance on each educated guess and they make a series of four or five of those, they have almost, you know, very quickly close to zero chance of being in the right spot. If you just multiply out the probabilities. But if they make that first big leap and they start getting customer feedback, that actually helps them, right, get more and more focused on where the bull's eye is, you have a really great chance of changing the world. So they don't build this great technology with no customers. The code was that good. You don't want to have a solution looking for a problem. Right, but if you want to, you know, kind of have some really big wall change, you got to be willing to make that first big step without the feedback because the customers don't know. It takes up awareness. Right, and if you just went to the customers and said, if you had this, what would you do? And they probably say, no, no, no, instead of that, I want another feature over here. So you got to go and build that first prototype and take the leap of faith. The issue is if you compound the leap of faith, your odds of being successful is low. If you quickly get it into the hands of the customers and get feedback and start focusing in on where the value is, your chance goes up dramatically. Awesome, I wish we had more time with you, David, but we're going to let you get back to all of the amazing innovation that I have no doubt is going on right behind us. Thank you so much for joining John and me on theCUBE today. Look forward to seeing you again soon. Absolutely, for John Furrier, I'm Lisa Martin. You're watching theCUBE's exclusive coverage of VMware Radio 2019. Thanks for watching.