>> Narrator: Live from Las Vegas, it's The Cube. Covering IBM Think 2018. Brought to you by IBM. >> Hello everyone and welcome to The Cube here at IBM Think in Las Vegas, the Mandalay Bay. I'm John Furrier, the host of The Cube. We're here in this Cube studio as a set for IBM Think. My next guest is Jennifer Shiin who's the founder of 8 Path Solutions. Twitter handle Jenn, J-S-H-I-N. Great to see you. Thanks for joining me. >> Yeah, happy to be here. >> I'm glad you stopped by. I wanted to get your thoughts. You're thought leader in the industry. You've been on multiple Cube panels. Thank you very much. And also Cube alumni. You know, IBM with the data center of the value proposition. The CEO's up on the stage today saying you got data, you got blockchain and you got AI, which is such the infrastructure of the future. And AI is the software of the future, data's at the middle. Dave and I were talking about that as the innovation sandwich. The data is being sandwiched between blockchain and AI, two super important things. And she also mentioned Moore's law. Faster, smaller, cheaper. Every 6 months doubling in speed and performance. And then Metcalfe's law, which is more of a network effect. Kind of teasing out token economics. You see kind of where the world's going. This is an interesting position from IBM. I like it. Is it real? >> Well it sounds very data sciency, right? You have the economics part, you have the networking. You have all these things in your plane. So I think it's very much in line with what you would expect if data science actually sustains (mumbles), which thankfully it has. >> Yeah. >> And I think the reality is you know, we like to boil things down into nice, simple concepts but in the real world when you're actually figuring it all out its going to be multiple effects. It's going to be, you know a lot of different things that interact. >> And they kind of really tease out their cloud strategy in a very elegant way. I mean they essentially said, 'Look we're into the cloud and we're not going to try to.' They didn't say it directly, but they basically said it. We're not going to compete with Amazon head-to-head. We're going to let our offerings to do the talking. We're going to use data and give customers choice with multi cloud. How does that jive for you? How does that work because at the end of the day I got to have business logics. I need applications. >> Yes. >> You know whether its blockchains, cryptocurrency or apps. The killer app's now money. >> Yep. >> If no one's making any money. >> Sure. >> No commerce is being done. >> Right. I mean I think it makes sense. You know, Amazon has such a strong hold in the infrastructure part, right? Being able to store your data elsewhere and have it be cloud. I don't think that was really IBM's core business. You know, a lot of I think their business model was built around business and business relationships and these days, one of the great things about all these data technologies is that one company doesn't have to do all of it, right? You have partnerships and actually partners so that you know, one company does AI. You partner with another company that has data. And that way you can actually both make money, right? There's more than enough work to go around and that much you can say having worked in data science teams right? If I can offload some of my work to different divisions, fantastic. That'd be great. Saves us time. You get to market faster. You can build things quicker. So I think that's one of the great things about what's happening with data these days, right? There's enough work to get around. >> And it's beautiful too because if you think about the concept that made cloud great is DevOps. Blockchain is an opportunity to use desensualization to take away a lot of inefficiencies. AI is also an automation opportunity to create value. So you got inefficiencies on block chains side and AI to create value, your thoughts and reaction to where that's going to go. You know, in light of the first death on a Uber self-driving car. Again, historic yesterday right? And so you know, the reality is right there. We're not perfect. >> Yeah. >> But there's a path. >> Well so most of its inefficiency out there. It's not the technology. It's all the people using technology, right? You broke the logic by putting in something you shouldn't have put in that data set, you know? The data's now dirty because you put in things that you know, the developer didn't think you'd put in there. So the reality is we're going to keep making mistakes and there will be more and more opportunities for new technologies to help you know, cheer that up. >> So I was talking to Rob Thomas, GM of the analytics team. You know Rob, great guy. He's smart. He's also an executive but he knows the tech. He and I were talking about this notion of data containers. So with Kubernetes now front and center as an orchestration layer for cloud and application workloads, IBM has an interesting announcement with this cloud private approach. Where data is the central thing in this. Because you've got things like GDPR out there and the regulatory environment not going to get any easier. You got blockchain crypto. That's a regulatory nightmare. We know a GDBR. That's a total nightmare. So this is happening, right? So what should customers be doing, in your experience? Customers are scratching their head. They don't want to make a wrong bet, but they need good data, good strategy. They need to do things differently. How do they get the best out of their data architecture knowing that there's hurdles and potential blockers in front of them? >> Well so I think you want to be careful of what you select. and how much are you going to be indebted to that one service that you selected, right? So if you're not sure yet maybe you don't want to invest all of your budget into this one thing you're not sure is going to be what you really want to be paying for a year or two, right? So I think being really open to how you're going to plan for things long term and thinking about where you can have some flexibility, whereas certain things you can't. For instance, if you're going to be in an industry that is going to be you know, strict on regulatory requirements right? Then you have less wiggle room than let's say an industry where that's not going to be an absolute necessary part of your technology. >> Let me ask you a question and being kind of a historian you know, what say one year is seven dog years or whatever the expression is in the data space. It just seems like yesterday that Hadoop was going to save the world. So that as kind of context, what is some technologies that just didn't pan out? Is the data link working? You know, what didn't work and what replaced it if you can make an observation? >> Well, so I think that's hard because I think the way I understood technology is probably not the way everyone else did right? I mean, you know at the end of the day it just is being a way to store data right? And just being able to use you know, more information store faster, but I'll tell you what I think is hilarious. I've seen people using Hadoop and then writing sequel queries the same way we did like ten plus years ago, same inefficiencies and they're not leveling the fact that it's Hadoop. Right? They're treating it like I want to create eight million tables and then use joins. So they're not really using the technology. I think that's probably the biggest disappointment is that without that knowledge sharing, without education you have people making the same mistakes you made when technology wasn't as efficient. >> I mean if you're a hammer, everything else is like a nail I guess if that's the expression. >> Right. >> On the exciting side, what are you excited about in technology right now? What are you looking at that's a you know, next 20 mile stare of potential goodness that could be coming out of the industry? >> So I think anytime you have better science, better measurements. So measurement's huge, right? If you think about media industry, right? Everyone's trying to measure. I think there was an article that came out about some of YouTube's failure about measurement, right? And I think in general like Facebook is you know, very well known for measurement. That's going to be really interesting to see, right? What methodologies come out in terms of how well can we measure? I think another one will be say, target advertising right? That's another huge market that you know, a lot of companies are going after. I think what's really going to be cool in the next few years is to see what people come up with, right? It's really the human ingenuity of it, right? We have the technology now. We have data engineers. What can we actually build? And how are we going to be able to partner to be able to do that? >> And there's new stacks that are developing. You think about the ecommerce stack. It's a 30 year old stack. AdTech and DNS and cookiing, now you've got social and network effects going on. You mentioned you know, the Metcalfe's law. So with all that, I want to get just your personal thoughts on blockchain. Beyond blockchain, token economics because there are a lot people who are doing stuff with crypto. But what's really kind of pointing as a mega trands standpoint is a new class of desensualized application developers are coming in. >> Right. >> Okay. They're dealing with data now on a desensualized basis. At the heart of that is the token economics, which is changing some of the business model dynamics. Have you seen anything? Your thoughts on token economics? >> So I haven't seen it from the economics standpoint. I've seen it from more of the algorithms and that standpoint. I actually have a good friend of mine, she's at Yale. And she actually runs the, she's executive director of their corporate law center. So I hear some from her on the legal side. I think what's really interesting is there's all these different arenas. Legal being a very important component in blockchain. As well as, from the mathematical standpoint. You know when I was in school way back when, we studied things like hash keys and you know, RSA keys and so from a math standpoint that's also a really cool aspect of it. So I think it's probably too early to say for sure what the economics part is going to actually look like. I think that's going to be a little more longterm. But what is exciting about this, is you actually see different parts of businesses, right? Not just the financial sector but also the legal sector and then you know say, the math and algorithms and you know. Having that integration of being able to build cooler things for that reason. >> Yeah the math's certainly exciting. Machine learning, obviously that's well documented. The growth and success of what, and certainly the interests are there. You seeing Amazon celebrating all the time. I just saw Werner Vogels, the CTO. Talking about another SageMaker, a success. They're looking at machine learning that way. You got Google with TensorFlow. You've got this goodness in these libraries now that are in the community. It's kind of a perfect storm of innovation. What's new in the ML world that developers are getting excited about that companies are harnessing for value? You seeing anything there? Can you share some commentary on the current machine learning trends? >> So I think a lot of companies have gotten a little more adjusted to the idea of ML. At the beginning everyone was like, 'Oh this is all new.' They loved the idea of it but they didn't really know what they were doing, right? Right now they know a little bit more. I think in general everyone thinks deep learning is really cool, neural networks. I think what's interesting though is everyone's trying to figure out where's the line. What's the different between AI versus machine learning versus deep learning versus neural networks. I think it's a little bit fun for me just to see everyone kind of struggle a little bit and actually even know the terminology so we can have a conversation. So I think all of that, right? Just anything related to that you know, when do you TensorFlow? What do you use it for? And then also say, from Google right? Which parts do you actually send through an API? I mean that's some of the conversations I've been having with people in the business industry, like which parts do you send through an API. Which parts do you actually have in house versus you know, having to outsource out? >> And that's really kind of your thinking there is what, around core competencies where people need to kind of own it and really build a core competency and then outsource where its more a femoral invalue. Is there a formula, I guess to know when to bring it in house and build around? >> Right. >> What's your thoughts there? >> Well part of it, I think is scalability. If you don't have the resources or the time, right? Sometimes time. If you don't have the time to build it in house, it does make sense actually to outsource it out. Also if you don't think that's part of your core business, developing that within house do you're spending all that money and resources to hire the best data scientists, may not be worth it because in fact the majority of your actual sales is with the sale department. I mean they're the ones that actually bring in that revenue. So I think it's finding a balance of what investment's actually worth it. >> And sometimes personnel could leave and you could be a big problem, you know. Someone walks about the door, gets another job because its a hot commodity to be. >> That's actually one of the big complaints I've heard is that we spend all this time investing in certain young people and then they leave. I think part of this is actually that human factor. How do you encourage them to stay? >> Let's talk about you. How did you get here? School? Interests? Did you go off the path? Did you come in from another vector? How did you get into what you're doing now and share a little bit about who you are? >> Yeah so I studied economics, mathematics, creative writing as an undergrad and statistics as a grad student. So you know, kind of perfect storm. >> Natural math, bring it all together. >> Yeah but you know its funny because I actually wrote about and talked about how data is going to be this big thing. This is like 2009, 2010 and people didn't think it was that important, you know? I was like next three to five years mathematicians are going to be a hot hire. No one believed me. So I ended up going, 'Okay well, the economy crashed.' I was in management consulting in finance, private equity hedge funds. Everyone swore like, if you do this you're going to be set for life, right? You're on the path. You'll make money and then the economy crashed. All the jobs went away. And I went, 'Maybe not the best career choice for me.' So I did what I did at companies. I looked at the market and I went, 'Where's their growth?' I saw tech had growth and decided I'm going to pick up some skills I've never had before, learn to develop more. I mean in the beginning I had no idea what an application development process was, right? I'm like, 'What does that mean to actually develop an application?' So the last few years I've really just been spending, just learning these things. What's really cool though is last year when my patents went through and I was able to actually able to launch something with Box at their keynote. That was really awesome. >> Awesome. >> So I became a long way from I think, have the academic knowledge to being able to apply it and then learn the technologies and then developing the technologies, which is a cool thing. >> Yeah and that's a good path because you came in with a clean sheet of paper. You didn't have any dogma of waterfall and all the technologies. So you kind of jumped in. Did you use like a cloud to build on? Was it Amazon? Was it? >> Oh that's funny too. Actually I do know Legacy's technology quite well because I was in corporate America before. Yeah, so like Sequel. For instance like when I started working data science, funny enough we didn't call it data science. We just called it like whatever you call it, you know. There was no data science term at that point. You know we didn't have that idea of whether to use R or Python. I mean I've used R over ten years, but it was for statistics. It was never for like actual data science work. And then we used Sequel in corporate America. When I was taking data it was like in 2012. Around then, everyone swore that no, no. They're going to programmers. Got to know programming. To which, I'm like really? In corporate America, we're going to have programmers? I mean think about how long it's going to take to get someone to learn any language and of course, now everyone's learning. It's on Sequel again right? So. >> Isn't it fun to like, when you see someone on Facebook or Linkdin, 'Oh man data's a new oil.' And then you say, 'Yeah here's a blog post I wrote in 2009.' >> Right. Yeah, exactly. Well so funny enough Ginni Rometty today was saying about exponential versus linear and that's one of the things I've been saying over the last year about because you know, you want exponential growth. Because linear anyone can do. That's a tweet. That's not really growth. >> Well we value your opinion. You've been great on The Cube. Great to help us out on those panels, got a great view. What's going on with your company? What are you working on now? What's exciting you these days? >> Yeah so one of the cool things we worked on, it's very much in line with what the IBM announcement was, so being smarter, right? So I developed some technology in the photo industry, digital assent management as well as being able to automate the renaming of files, right? So you think you probably a picture on your digital camera you never moved over because you, I remember the process. You open it, you rename it, you saved it. You open the next one. Takes forever. >> Sometimes its the same number. I got same version files. It's a nightmare. >> Exactly. So I basically automated that process of having all of that automatically renamed. So the demo that I did I had 120 photos renamed in less than two minutes, right? Just making it faster and smarter. So really developing technologies that you can actually use every day and leverage for things like photography and some cooler stuff with OCR, which is the long term goal. To be able to allow photographers to never touch the computer and have all of their clients photos automatically uploaded, renamed and sent to the right locations instantly. >> How did you get to start that app? Are you into photography or? >> No >> More of, I got a picture problem and I got to fix it? >> Well actually its funny. I had a photographer taking my picture and she showed me what she does, the process. And I went, 'This is not okay. You can do better than this.' So I can code so I basically went to Python and went, 'Alright I think this could work,' built a proof of concept and then decided to patent it. >> Awesome. Well congratulations on the patent. Final thoughts here about IBM Think? Overall sentiment of the show? Ginni's keynote. Did you get a chance to check anything out? What's the hallway conversations like? What are some of the things that you're hearing? >> So I think there's a general excitement about what might be coming, right? So a lot of the people who are here are actually here to, I think share notes. They want to know what everyone else is doing, so that's actually great. You get to see more people here who are actually interested in this technology. I think there's probably some questions about alignment, about where does everything fit. That seems to be a lot of the conversation here. It's much bigger this year as I'm sure you've noticed, right? It's a lot bigger so that's probably the biggest thing I've heard like there's so many more people than we expected there to be so. >> I like the big tent events. I'm a big fan of it. I think if I was going to be critical I would say, they should do a business event and do a technical one under the same kind of theme and bring more alpha geeks to the technical one and make this much more of a business conversation because the business transformation seems to be the hottest thing here but I want to get down in the weeds, you know? Get down and dirty so I would like to see two. That's my take. >> I think its really hard to cater to both. Like whenever I give a talk, I don't give a really nerdy talk to say a business crowd. I don't give a really business talk to a nerdy crowd, you know? >> It's hard. >> You just have to know, right? I think they both have a very different sensibility, so really if you want to have a successful talk. Generally you want both. >> Jennifer thanks so much for coming by and spending some time with The Cube. Great to see you. Thanks for sharing your insights. Jennifer Shin here inside The Cube at IBM Think 2018. I'm John Furrier, host of The Cube. We'll be back with more coverage after this short break.