(uptempo music) >> Narrator: Live from Anaheim, California it's The Cube! Covering Nutanix dot Next 2019 brought to you by Nutanix >> Welcome back everyone to The Cube's live coverage of Nutanix dot Next here in Anaheim California, I'm your host, Rebecca Knight along with my cohost John Furrier. We are joined by Dr. Ayanna Howard, she's the professor and chair of the School of Interactive Computing in the College of Computing at Georgia Institute of Technology. Welcome Dr. Howard to The Cube! Thank you, thank you. I'm excited about this conversation. >> Yeah so you, you're a fascinating person, when you were a little girl, watching Bionic Woman you said "I wanna be a scientist," you started your career at NASA. >> Ayanna: Correct. >> You are an entrepreneur, a researcher. Tell us what you're doing today. >> So what I'm doing today, and what I'm really excited about is bringing robots into the home of children with special needs. So one of the things about kids and those that may have a developmental disability is that there's not enough contact hours with human clinicians. And so, how do you augment that in the home environment? How do you bring technology into the home to do therapy with them, to do even education? And so that's what I focus on. >> So, we want to hear so much more about that, but what are you gonna be talking about at this conference? It's the future of AI, and robots. >> Yes, I'm gonna talk about the things that make my robots work. And so, the future of AI and robotics and where it leads, it's a combination of things like wearables. So if you think about all the data around us, we have wearables with our phones, and our smart watches, all that data that's being collected about us, allows our machines to do very interesting personalized things with us and for us. The other thing is that if you think about collaborative AI, collaborative machines, we're going to the place where the workforce and how you do your work, you're going to have an AI as a companion, a robot as an assistant, so you might not be sitting next to a human, you might be sitting next to a robot. And so, what does that look like? And then, of course, emotional AI, and so, yes, machines do have emotions, which is, counts kind of weird, but in order for us to work with others, we typically have a bond, so why not have a bond with our machines? >> What's the software look like? I'm rifting in my mind here, I'm just thinking about, I'm gonna write some software that might be dynamic, a neural network, these kinds of words have been kicked around in the industry. How do you make software have emotion in AI? Because it has to be random, but yet not, it has to be programmable. >> It does. But think about it. Emotions are not necessarily random. Emotions are pretty repetitive, i.e., if you're hurt, what do you do? If you're young you cry, if you're older you hide the cry, right? I mean, it's very repetitive, if you're happy there's a certain emotion, what makes you happy? There are certain things that we can all say if I suddenly woke up and I won a prize, I'd be happy. Emotions are actually very predictable, they're not that hard to model. >> And the data sources could be coming off my Fit Bit, facial recognition, you know the morning... >> Well facial recognition, you can see it in the face, in fact your pulse, and you sweat a little bit when your emotions change. Remember the mood rings back in the day? (laughter) >> Sure! >> OK those were fake, but still, their concept about them was that your body gives a response based on the emotions inside. >> Yeah, that's so cool. So what's the state of the art, you look at bleeding edge and state of the art kind of mainstream, where are people with software, machine learning, AI, what's some of the things that are notable to you that are important to highlight? >> Yes, so I think that the two areas that are the furthest ahead, one is facial recognition and emotion detection, and it's because the application are out there. As an example, airports are putting in these systems, and so imagine, I mean, the positive is, is that you don't have to book or print out your ticket, right? You just walk into the airport, you walk though security, you don't get padded down, and you walk to your gate and get on the plane. I mean, just imagine that. You're like How would you do that? Well, if I know who you look like, and I can model you, and I grab your wearable, and your data, I know who you are! So, I don't have to make sure that you are who you are, I know. I mean, so that's kind of a benefit. Of course, there's some negatives, which we won't talk about, but that's one area, this facial recognition aspect. The other I think it's in healthcare, I think it's in the fact that our data, and about us, about our health, it's so much there, and as we mine it we just get better. There's, for example, some research that shows stress can be detected and I can then have a, think about it, I can have an AI that if I know you're stressed like, I'm not going to send you that email, I'm going to halt a little bit, until I realize that your stress level is a little bit better, and then I will give you the bad news. Right? Like, because we don't want to be stressed. >> I need that, I need that app. >> Rebeca: But that's a manager with really good intentions, I mean, you can really see the perils of this going... >> No, that's, that's the negative. That's the aspect of, all these things are, really have good return on investment, good quality, but the negatives are is that if you have nefarious manager or an organization like I just wanna make money, money, money, you can sway that, and I think, though, that most organizations are thinking about this. I think there's this push now to do things like regulations, to basically protect us, but still insure that we have a positive relationship with AI and robotics. >> What's the coolest thing you've seen or built recently that could tie into the robotics? >> So, I will personally say it's one of our machines, that has, it emotionally responds to you based on what you're doing, and so what does this mean? It means I have robots that are just looking so cute, right? You look at them, and anyone looks at them, and it's just , it's like, it's real, it's intelligent, it like understands me. Of course, it's programmed based on modeling but it's just as fascinating, and I watch people interact with robots, and it's like oh, my gosh, this person, this individual, is really engaged with my robotic creation. >> And you mean, in conversation or just in feeling the comradery? >> In conversation, in interaction, and the robots, they have a limited script, but people will adapt to that, right? And they will, it's just like when you talk to your phone, have you noticed that when your phone doesn't understand you, what do you do? You speak a little slower. You might choose different words, right? I see that with the robots, you change your behavior based on the limitations. >> Speaking with someone who doesn't speak lour language natively. >> Correct. Same thing with robots. >> So describe what you see... Returning to the beginning of our conversation talking in particularly with kids with special needs. >> Ayanna: Yes. >> Describe what you see, the changes in the child, who is developing a relationship, a bond with a robot. >> Yes, so what we've actually shown, not just seen and observed, is that when we have a child interacting with a robot their, and what we call, whatever milestone we're doing, so maybe it's movement therapy, which means I want them to say, move a little faster than their normal space of moving, what I see is with the robot there is a partner encouraging, guiding, providing them input on how well they're doing, or in terms of correcting, the child improves their behavior, and so between day zero and day n, the child has gotten better. We see that. We have the data that shows that. >> Incredible. I wanna also ask about women in technology, and this is, this is really a theme at every single tech conference you go to because it's such a problem, it's such an issue that is finally getting the attention it deserves. We know about the dearth of women leaders, the dearth of underrepresented minorities, particularly in management leadership positions, what do you see as you role in tackling this problem, as the head of an important department in technology and also as a woman of color? >> Yeah, so I think there's always been kind of two dilemmas, one is what they call the pipeline, which is now the pathway, like how do you get women to come into stem? And the data has shown that is not that girls are not interested in stem, it's that they lose interest because of their society, right? So that's one thing. It's like make sure that where they are in the society is encouraging. The other is that when you get older, you look up, you're like, okay there's no one there. Obviously, I'm not supposed to be here, or when things get tough, it's like, okay, I need to move out. And so the other is, how do you do mentorship and sponsorship, so that women are pushed forward as managers and supervisors. So those are kind of the two things. And so, as a, and I consider myself a leader in this space, I actually feel it's my duty to be up front, and be a mentor, and be a lead, and actually be vocal, and make others realize like, if I'm in a room, and we're deciding on, you know, a student or a candidate, and there's no representation, you know, I'm comfortable enough to say, hey, I should not be the one that says this, right? And eventually what you see is that people start looking and thinking about this, at every instance of time. >> Do you feel like it's getting better? >> I do. It's getting better. And it's not perfect but it's getting better. Like, if I look in the classrooms, I look in the computer science curriculums, I see more female students coming in, and lasting, and then going into corporate America and continue on to grad school. I see it being better, of course it's not on parity, but is is better. >> That's awesome. And the technology has shifted the definition. It's not programming, or electrical engineering, the surface area for tech is gaming to analytics, data science, it's huge. >> Human-centered interaction. >> There's new artistry around us, so I think it's a great surface area. >> It is, and I think one of the reasons why it's so important is that the world is diverse, I mean, in terms of all the different aspects. An so, if you're gonna create products for a diverse world, you should have individuals that are also diverse, creating them for everyone so that there's some equality in the process. >> As the analog world connects with the digital world, fascinating we talked before we came on camera around the technology in digital. So the human experience for me, whether having robots, detecting emotions, and having some sort of new notifications, like hey, you know, cheer up, or do something clever... >> Right >> Is that you can now immerse, so augmented reality has been the first killer app before virtual reality, but gaming is an indicator of what's happening onscreen, so, the onscreen digital realm is intersecting with our lives. >> Ayanna: It is. >> What's your view on this? Because this is an area that's new, it's cutting edge, it's a first generation problem, an opportunity. >> Opportunity. I think this, this blending of the, I would say, even, I would say the blending of the digital and physical and the gamefication aspects, is really gonna enhance two areas. One is education, and the retraining, and so what does that mean? It means that, instead of me having to, not to say go to college for four years, but instead of me trying to study everything in this one-semester course, it's like, I just need some basic knowledge and I can then work in the field, and I have my augment reality and so I see things and there's some scaffolding, there's some indication of here's step one, here's step two, ahh, you did that step two a little bit wrong, let's revise it. So you learn with real-time training and that's with doctors, well except for live patients, but you know, with doctors or residents, factory workers, or even teachers, teachers who are teaching say, calculus, that may have an English background. That's where it is. >> The progressions are not linear like they used to be. >> Ayanna: No! >> They are different, and now you have dated instrumentation with on-demand digital robots... >> Robots, agents... >> John: Agents, assistants... >> Adaptation, taking things from other places, so if I, for example, learn the best way to provide information to this human and this factory, well guess what, I can take that information, connect to the cloud, connect to the data centers, and apply that information to another worker in a different factory, but very similar characteristics, and so you have this transfer of knowledge as well. >> So education was one. What's the other one? Healthcare? >> Of course it's healthcare! (laughter) Of course. >> As someone who is immersed in it and a believer in technology, what do you do to disconnect? Well, first of all, do you disconnect? Do you worry about our over reliance on these little devices in our pockets, and what do you do to sort of leave the digital world behind for a while? >> Yeah, so I do worry about our over reliance because we've shown, and other researchers have shown, that there's actually an over trust factor. We will use devices, and of course these devices they have errors, right, even if it's you know 1% of the time, and that 1% of the time when they have errors we find that a lot of individuals will trust those errors, because they're over relying, they kind of go in zone mode, they're like, it worked all this time, so that 1%, they just don't question it. >> It must be real news! (laughter) >> But it's scary! >> Yeah, it is. >> It's scary. I do worry about that. And we're thinking about ways to try to mitigate that, 'cause that does worry me. How do I disconnect? I think that with anything mind, body and soul, so I love listening to music, although that's not disconnecting from technology 'cause I'm using technology to listen, but it's this zone period. Exercise, I think most of us think about exercise I'm fairly religious, even when I'm traveling, like okay, I'm going to find the gym and at least walk on the treadmill because we do have to have that combination, in order to be healthy ourselves. >> Finally, for that little girl, the little girl you, who's watching Bionic Woman I think that's the thing, we need more shows like that, to get, to get >> Click us interested >> Well exactly, what would be your advice to the smaller you, who says I want to be a scientist someday? >> So I would, and this is like some advice that people told me as I was growing up, and I didn't realize I had really good mentors, is one is, don't listen to the naysayers, i.e., believe in yourself, right? And I think that's the one thing we sometimes forget to do, like believe in that dream, even if others say that it's not possible, and it's like, no, everything is possible if you believe in yourself. >> Words to live by. Thank you so much for coming on the show. >> Thank you. >> Rebeca: This was great conversation. >> Awesome! >> I'm Rebeca Knight for John Furrier. We will be back here tomorrow with more from Nutanix dot Next. We hope to see you then. (electronic music)

>> Announcer: Live from Nice, France, it's The Cube covering Dot Next Conference 2017, Europe. Brought to you by Nutanix. (techno music) >> Welcome back, I'm Stu Miniman happy to welcome back to our program, from the IBM Cognitive Systems Group, we have Bob Picciano and Stefanie Chiras. Bob, fresh off the keynote, uh speech. Went a little bit long but glad we could get you in. Um, I think when the, when the IBM Power announcement with Nutanix got out there, a lot of people were trying to put the pieces together and understand. You know, we with The Cube we've, we've been tracking, you know, Power for quite a while, Open Power, all the things but, but I have to admit that even myself, it was like, okay, I understand cognitive systems. We got all this AI things and everything but on the stage this morning, you kind of talked a little bit about the chipset and the bandwidth. You know, things like GPUs and utilization, you know, explain to us, you know, what is resonating with customers and, you know, where, you know, what's different about this because a lot of the other ones it's like, oh well, you know, software runs a lot of places and it doesn't matter that much. What's important about cognitive systems for Nutanix? >> Yeah, so, first off, thanks Stu. And, as always, thanks for, you know, you for following us and understanding what we're doing. You mentioned not just Power but you mentioned Open Power, and I think that's important. It shows, actually, the deeper understanding. You know, we've come a long way in a very short amount of time with what we've done with Open Power. Open Power was very much at it's core about really making Power a natural choice for industry standard Linux, right? The Linuxes that used to run on Power a couple of generations ago were more proprietary Linuxes. They were Big Endian Linux but Open Power was about making all that industry standard software run on top of Power where we knew our value proposition would shine based on how much optimization we put into our cores and how much optimization we put into IO bandwidth and memory bandwidth. And boy, you know, have we been right. In fact, when we take an industry standard workload like a no sequel database or Enterprise DB, or a Mongoloid DB, Hadoop, and put it on top of Linux, an industry standard Linux, on top of Power, we typically see that run about 2X to 3X better price performance on Linux on Power than it would on Linux on Intel. This is a repeating pattern. And so, what we're trying to do here is uh, really enable that same efficiency and economics to the Nutanix Hyper Converged Space. And remember, all these things about insight based applications, artificial intelligence, are all about data intensive workloads. Data intensive workloads and that's what we do best. So we're bringing the best of what we do and the optionality now for these AI workloads and cognitive systems right into the heart of what Nutanix is pivoting to as well. Which is really at the, at the core of the enterprise for data intensive workloads. Not just, you know, edge related VDI based workloads. Stefanie will you, you want to comment on that a little bit as well. >> Yeah, we are so focused on being prioritized and what space we go after in the Linux market around these data centric and AI workloads. And at the end of the day, you know, Nutanix has Nutanix states. It's about invisible infrastructure, but the infrastructure underneath matters. And now with the simplicity of what Nutanix brings you can choose the best infrastructure for the workloads that you decide to run, all with single pane of glass management. So it allows us to bring our capabilities at the infrastructure levels for those workloads, into a very simplest, simple deployment model under a Nutanix private cloud. >> Yeah, I, I think back when, you know, we had things like, when Hadoop came out, you know, we got all these new modern databases, >> Right. >> You know, I wanted to change the infrastructure but simplicity sure wasn't there. >> Yep. >> Uh-huh. >> It was a couple of servers sitting under the desk, okay, but when you needed to scale, when you needed to manage the environment, um, it was challenging. We, we saw, when, you know, Wikibon for years was doing, you know, research on big data and it was like, ah, you know, half the deployments are failing because, you know, it wasn't what they expected. >> Right. >> The performance wasn't there, the cost was challenging. So it feels like we're kind of, you know, turn the corner on, you know, making, putting the pieces together to make these solutions workable. >> I think we are. I think Dheeraj and his team, Sunil, they've done a wonderful job on making the one click simplicity, ease of deployment, ease of manageability. We saw today, creation of availability zones. High availability infrastructure. Very very simplistic. So, you know, as, you know, I've had other segments with Dave and John in the past, we've always talked about, it's not about big data, it's about really creating the ability to get fast actionable insights. So it's a confluence of that date environment, the processed based workflow environment, and then making that all simple. And this feels like a very natural way to accomplish that. >> I want to understand, if I caught right, it's not Power or x86 but it's really putting the right workloads in the, in the right place. >> That's right. >> Did I get that right? >> That's right. >> What, what are the customer deployments, you know? >> Heterogeneity is key. >> How do I then manage those environments because, you know, I, I want kind of homogeneity of, of management, even if I have heterogeneity, you know, in, in my environment, you know. What, what are you hearing from your customers? >> I think how we've looked at Linux evolved. The set of workloads that are being run on Linux have evolved so dramatically from where they started to running companies and being much more aggressive on compute intensive. So it's about when you bring total cost of ownership which requires the ability to simply manage your operations in a data center. Now the best of Prism capabilities along with the Acropolis stack allows simplicity of single pane of glass management for you to run your Power node, set of nodes, side by side with your x86 set of nodes. So what you want to run on x86 or Windows can now be run seamlessly and compatible with your data centric workloads and data driven workloads, or AI workloads on your Power nodes. It really is about bringing total cost of ownership down. And that really requires accessibility and it requires simplicity of management. And that's what this partnership really brings. It's a new age for hyper converged. >> Yeah. >> What should we be looking for, for the partnership, kind of over the next 12 years, 12, 12 months. (laughs) >> 12 years? (laughs) (laughter) >> 12 years might be a little tough to predict, but over the next year, what, what should we be looking for the partnership? You know, I think back you talked about, Open Powered Google is, you know, a big partner there. Is there a connection? Am I drawing lines between, you know, Nutanix and Google and what you're doing? >> I won't comment on that yet but, you know, but, as you know we have a big rollout coming up as we're getting ready to launch Power Nine. So there'll be more news on some of those fronts as we go through the coming weeks. And I hope to see you down in Dallas at our Cloud or Cognitive event. Or at one of the other events we'll be jointly at where we do some of these announcements. But if you think about where this naturally takes us, Sunil talked about mode one and mode two applications. So what we want to see is increasing that catalog for mode one applications. So things that I'd like to see is an expanded set of relationships around what we both do in the SAP space. I'd like to see that catalog of support enriched for what's out there on top of the Linux on Power space, where we know our value proposition will continue to be demonstrated both in total cost of acquisition as well as total cost of ownership. >> Yeah. >> I mean, we're really, you know, seeing some great results on our Linux base. As you know, it's now about 20 percent of the power revenue base is from Linux. >> Uh-huh. >> And that's grown from a very small amount just a few years ago. So, I look to see that and then I would look at more heterogeneity in terms of the support of what we do, both in Linux and maybe, in the future, also what we do to support the AIX workloads, uh, with Nutanix as well. Because I do think our clients are asking about that optionality. They have big investments, mission critical workloads around AIX and the want to start to bring those worlds together. >> Alright and Stefanie, want to give you the final word, you know, anything kind of learnings that you've had, of the relationships as you've been getting out and getting into those customer environments. >> I have to say the excitement coming in from the sales team, from our clients, and from the business partners have been incredible. It really is about the coming together of, not only two spaces of simple, and absolutely the best infrastructure and being able to optimize from bottom to top, but it's about taking hyper converge to a new set of workloads. A new space. Um, so the excitement is just incredible. I am thrilled to be here at Dot Next and be able to talk to our clients and partners about it. >> Alright well Stefanie and Bob thank you so much for joining us. >> Thanks Stu. >> Thank you Stu. >> Sorry we had to do a short segment but we'll be catching ya up at many more. Alright so we'll be back with lots more coverage here from Nutanix Dot Next in Nice, France. I'm Stu Miniman, you're watching The Cube. (techno music)