NARRATOR: Luminaries, talking to the brightest minds in tech.
MALE SPEAKER: And my hope is that we come together to share more than technology, and expertise, and products, but that we share a vision of a future that is better than today, a vision of technology as the driver of human progress.
NARRATOR: Your hosts are Mark Schaefer and Douglas Carr.
MARK SCHAEFER: Welcome everyone, to another episode of Luminaries, where we talk to the brightest minds in tech. This is Mark Schaefer with my distinguished co-host, Douglas Karr, a.k.a. Dougy Baby. What’s going on, Doug?
DOUGLAS KARR: Well, the one thing that’s going on is the compliments just keep getting better episode after episode. So thank you.
MARK SCHAEFER: Well, you know, what can I say? You’re a distinguished gentleman. And we talk about how we talk to the brightest minds in tech, and this one of the biggest topics ever. It’s one of the biggest topics on the scene of technology today, internet of things, or as they say here in my home state of Tennessee, Internet of “thangs”.
It’s just one of the coolest things ever. And we are going to be talking today to a person who is at the absolute epicenter of the internet of things. It’s Jason Shepherd. Jason is the CTO for Dell Technologies leading market and technology strategy, solution planning, and the strategic ecosystem development for Dell and their internet of things initiatives.
So, Jason, first of all, welcome to the show. It’s such an honor to have you here.
JASON SHEPHERD: Thank you. Thanks so much for having me. My pleasure to be here. That was quite the intro, and I’m kind of wondering when I get my modifier, like a baby or something like that on my name.
MARK SCHAEFER: We can do that. How about Jason– we could do it Jason Baby, we can do Jason Dude. You know, I’m open.
JASON SHEPHERD: The Dude. Yeah. It’s all good. We’ll figure it our.
MARK SCHAEFER: So we’re here today with The Dude of Dell, Jason Shepherd. So Jason, first of all, this is like the coolest job imaginable. Tell us a little bit about– how did you get here? What’s sort of a summary of your career path that got you to this amazing position?
JASON SHEPHERD: Yeah. It’s actually kind of been a progression. I’m a mechanical engineer in terms of school. I started out, I think– well, I started out life annoying my parents because I had to touch every switch and button imaginable. And just kind of getting into things, and building stuff, and kind of that led me to being a mechanical engineer out of school.
But I started at Dell doing engineering, and just kind of kept progressing through various different roles. I had an opportunity to goes to a startup from Dell, kind of get into the startup world, and got into MP3 players at the worst time to be an MP3 player. It was right when Napster got shut down.
So that was a short-lived startup THAT I went to. But then i got on this track for getting into more of a full solution stack software, electrical stuff. And that led me, actually, ultimately, back to Dell. I was in engineering management for a while. And then got it got into R&D, led an R&D group doing a bunch of cool stuff around collaboration, remote employees, some of the early tablet stuff that we were doing, peer to peer sharing between devices.
I did that for awhile, got into CTO, and from there it kind of led to, hey, what do we want to do with this buzzword IoT? And a small group of us, across the company, we built up a strategy and shopped it around. And the way we really got it moving is we started building an ecosystem. And that kind of led to me learning a lot more about the market, and it just kind of grew from there.
It helped get this project started internally that turned into the open-source project EdgeX Foundry which also started pulling together more of a broad ecosystem of players. So it’s just– career path-wise it’s just been a series of, well, a, being naturally inquisitive, just kind of get your hands on things, surround yourself with good people, and kind of adapt, and fill gaps, and grow. So that’s kind of what’s led to this.
MARK SCHAEFER: That’s so cool, because in your LinkedIn profile you describe yourself as an entrepreneurial technology leader. And I was wondering, well, gosh, how does that help you or hurt you in a big company like Dell? But it sounded like to build this internal competency you really did take an entrepreneurial approach.
JASON SHEPHERD: Yeah. You’ve got to be– well, a, if it’s fuzzy, I’m on it. Ambiguous stuff, I definitely like to explore. You’ve got to be willing to take some risks, and stick your neck out a bit, and be willing to be wrong, and just sort of be creative. And again, it’s definitely about the people you surround yourself with. But yeah, you take some risks.
MARK SCHAEFER: See, that’s better than dude.
Jason, if it’s fuzzy I’m on it, Shepherd.
JASON SHEPHERD: There you go. I like it.
DOUGLAS KARR: Well, Jason– and I’m looking at it, nowadays it’s not a risk anymore. I feel as though the average business might be vastly underestimating the impact of IoT. I was speaking to a representative from a roofing company, and they chase hail and storm damage all over the country. And I asked if anybody was working on harnessing IoT to enable, possibly, network centers on roofs in regions that you know get heavy damage. And all I got was blank stares back from the table.
As someone who’s working deep within the industry, do you think IoT is a technology that will be disrupting virtually every business model?
JASON SHEPHERD: I would definitely say yes. We’re constantly seeing new use cases pop-up and just stuff that you wouldn’t really have thought about. And then you see a lot of solutions looking for problems out there, of course, as well. And there’s incredible gains that you can see in efficiency, and quality, and safety, and those core elements. But the real scale factor here is business model disruption.
And it’s interesting, though, when you look at that roofing example, in all these cases, it’s hard to instrument the physical world. This is what IoT is about. It’s putting more sensors out in the physical world, and ingesting data and running analytics on top of it, making things better, driving change. But when you look at any particular ecosystem, in your example, well, it’s roofing, but broader speaking it’s about insurance and that whole kind of ecosystem around mitigating risk and whatnot– making sure things are OK. It’s like, who has the most to gain from doing that instrumentation?
So while roofers could be the person to do it. They could go put a sensor out on people’s houses. Well, but they don’t necessarily have that constant connection with their customers. It could be there’s these services that do like hail zone monitoring, and they sell that data to insurance companies and roofers alike and. They could do it. But then who has the most to gain is actually the insurance company?
But then beyond the insurance company, it’s actually the telecos that serve the houses. Because then they could do it for insurance companies, and health care providers, and utility providers. And so sort of like this food chain of like who has that ultimate relationship, and who has the most benefit from doing it. But it’s an interesting equation of where people are going to play, and who’s going to be the one to really kind of come in and disrupt any given industry.
DOUGLAS KARR: Fascinating.
MARK SCHAEFER: You know, I saw this article recently in the New York Times, Jason. It was fascinating. It was talking about how the country of Finland is providing technology training to all of their citizens as an option so they can think about how to apply technologies like IoT, artificial intelligence, to their daily lives. So this whole roofing example made me think of that.
And I just think that’s brilliant, because these are not technology people– people in roofing, construction, dentists, let’s say. And I got to hear Kevin Kelly, the futurist, at Dell World a couple years ago. And I was so embarrassed, because up until then I had never heard of Kevin Kelly. And what a famous guy and an amazing author.
I love his book The Inevitable. And he says in this book that connecting these dots into our everyday lives is really going to be the future of innovation. That applying a layer of AI IoT to our everyday jobs, that is going to be the future of innovation.
And I’ve been reading some of your writing, and it sounds like that’s sort of where you are too.
JASON SHEPHERD: Yeah. I think that AI is obviously a big disruptive trend coming up. I mean, there’s various degrees of AI, of course. We’re still quite a ways off from the introduction of emotion, and then there’s a whole ethics thing, and all that. But we’re starting to see, of course, more and more common tasks, things that are predictable operations, that you can kind of automate through AI happen– chat bots, call centers, robotics in the sense of doing prescribed tasks– autonomous vehicles. It’s well-constrained with some sort of context and rules that you can kind of train and the models to adapt to.
So that kind of stuff is ripe for innovation. But another thing that he talks about often, which I think is a big part of it, is this notion of remixing. Where you break down components into a bunch of– these big systems into a bunch of smaller components, and then they’re kind of remixed and adapted. And then you kind of learn new things and then you build off of that. That’s a big portion of where we’re going to see innovation, increasingly, is just this kind of new normal that keeps developing over time.
We haven’t even scratched the surface on the use cases that we’re going to see. More screens everywhere, obviously new interfaces like AR and VR, and all of these things, I think, are where the innovation comes from. And as we start to see more interoperability across IoT devices, we’re also going to see these different AI solutions working with each other and building up new types of new ideas and solutions.
So it’s just this kind of cycle of innovation that gets faster and faster as things get more and more sophisticated. So foundationally I agree, but I think there’s just so many different elements that come into play.
MARK SCHAEFER: Well, you actually mentioned, in one of your blog posts, about this idea of predictive product development, which I think is fascinating.
JASON SHEPHERD: Yeah. Talk about this new product mindset– it’s used to be, when you ship a product, you would try to kind of retroactively go figure out from customers in one-on-one or in focus groups what they want. And in a sense, you’re predicting in that realm. But going forward, this is about constantly adapting products and services out in the field. The value is the total lifecycle of the product, not the day that you ship it.
And this is really fed by all that data that’s coming in and all the telemetry that you’re getting from IoT devices that tend to just kind of state the facts compared to the bias that people might have about products. You get a lot of new type of telemetry about how things are actually being used. And then you can supplement that, of course, with their perception. So yeah, it’s about software defining products, increasingly hardware defining products, and areas like 3D printing and whatnot, and kind of constantly developing new experiences, and predicting those needs.
And you’ll actually see AI solutions that start to– I mean, obviously, we’ve got the home speakers or home assistants that are leveraging AI to understand your needs. But actually changing the product itself will start to evolve over time. That’ll become a trend as well.
DOUGLAS KARR: Jason, you’ve described this as the AOL stage of IoT, with simple deployment of cheap computers and sensors attached to the cloud. With Edge computing, we can now harness a server-client architecture with virtually unlimited cloud resources. Is bandwidth the bottleneck right now? And will 5G networks open up that bottleneck?
JASON SHEPHERD: I think bandwidth– 5G is a clearly important trend. It’s a bit of the– I always joke, if you’ve ever seen the movie My Big Fat Greek Wedding, where the dad sprays Windex on everything. It’s like, oh, you’ve got a cut? Just put some Windex on it.
I think 5G is one of the Windexes of technology. Blockchain is the one.
To a certain degree, even AI– people talk about AI. There’s a good post floating along out there. It’s like, if it’s machine learning, it’s probably written in Python. If it’s AI, it’s probably written in PowerPoint.
But no– AOL stage of IoT, we’re just starting to see things getting online at scale. I do think that 5G and the overall bandwidth increase that we’re going to see, of course the throughput for any given connection helps, the latency with 5G– it’s all important. But I think the ultimate bottleneck, and why we need more Edge computing, is cost. Just because you have unlimited pipe doesn’t mean it’s free.
And so you really only want to be moving around data if it’s meaningful. I look at the analogy of traffic, and if you build more roads and freeways, more people come and more people move to the suburbs. In that case kind of the cloud, so to speak. And then it just gets more and more congested, and you’re right back where you started.
And people end up moving to the city. In this analogy, kind of like the Edge, because they can be closer to all the services. Or your computer– as your computer gets faster and faster, all of the apps get more and more rich. And then all of a sudden, you’re kind of back where you were. So I think it’s a bottleneck, but I think cost is one of the big ones as well.
MARK SCHAEFER: One of the areas I was reading about that Dell is working heavily in with these IoT applications is surveillance. That’s sort of a scary word to a lot of people. But what I was reading here is that Dell is really working on a lot of very interesting sort of positive applications and technologies for the world. Can you talk a little bit about some of those challenges and some of the opportunities you see to apply IoT in that space?
JASON SHEPHERD: Yeah. Surveillance does have sort of that big brother connotation to certain folks. The areas that we’re working in is more about public safety today. And we’re expanding on that. But things like, obviously, stadiums, and schools, and cities, and the like.
But cameras are one of the best sensors around. They obviously have that video component, but there’s just so much you can ascertain through a camera, whether it’s temperature, detecting gas emissions. You can detect vibration with the right cameras and software on top of that. So we’re looking at, how do we can you expand into these different areas?
This initial push has been, how do I simplify how surveillance solutions are deployed and make sure you have continuity in the data through all the storage? I mean, massive amounts of storage for select surveillance solutions. How do I secure and manage it? But there’s a lot of opportunities ahead in terms of applying more computer vision, more analytics, more AI to those data feeds. So that’s a big push of where we’re headed next.
Within that, some of the challenges, the privacy starts to spin up more and more as you’re getting past what most people are doing today in surveillance. It’s like Fancy DVR. When you start doing analytics, it’s one thing to be analyzing demographics data– so individual identity is abstracted– it’s another thing to be monitoring actual individuals. And so there’s a lot of discussion there.
And it all depends on context. If I was to put up cameras and I gave people the impression that I’m watching whether they’re working within their jobs, kind of flat out working, that’s one thing. But if I put up cameras that we’re making sure that you’re safe, that’s quite another– you’re wearing your hard hat and gloves, you’re not going into areas that are hazardous, and all that. So there’s a lot of stuff to kind of work through on that front.
A big thing on– even the quality of the algorithms are getting better and better. I don’t know if you guys have seen, if you google muffin test for AI, there’s a–
MARK SCHAEFER: I did see something about that.
JASON SHEPHERD: Yeah. There’s a set of images. So it’s easy enough to recognize some of the kind of binary things– very obvious differences between images. But when you get into the muffin test for AI, it’s basically like a series of blueberry muffins and the faces of cute little chihuahuas all kind of blurred to together. And if you just kind of look at it, they look very, very similar. So how do you start getting to some of those advanced things?
So getting to that level of detection is a challenge. But yeah– and then you get into this notion of advanced reasoning. It’s one thing to call out absolutes, it’s another thing to start bringing in judgment and reasoning. And is that person really going to steal something?
MARK SCHAEFER: Oh my.
JASON SHEPHERD: Yeah. So people talk about AI as it’s going to take over the world and do some of these things, but it just comes down to fundamental judgment. There’s this notion of the ethics around it and this potential of profiling. And so these are some of these key challenges that start spinning up fast when you start looking at computer vision on top of that foundation.
MARK SCHAEFER: Jason, it’s such an important topic, such a fascinating topic. We really could do a whole show just on that discussion someday. If you look at some of the things that are going on in China, for example. And it’s really– I’m conflicted about some of these things, because you could make a case and look at how they’re applying IoT and all the surveillance in China. And you think, well, in some ways they’re positioning themselves to be one of the most secure, safest countries in the world. But one of the things you know about life in our world in general is that where corruption can occur, corruption, eventually, will occur.
JASON SHEPHERD: Of course, yeah.
MARK SCHAEFER: Yeah. It’s just a fascinating, fascinating topic.
JASON SHEPHERD: Well, and there’s rumors– it’s kind of– usually, when there’s rumors, there’s some truth you. It’s not clear yet, but there’s this notion of the social score. Have you seen some of the stuff around that?
DOUGLAS KARR: Mm-hmm. That’s
MARK SCHAEFER: Well, absolutely. Yeah, I actually wrote a book about that.
JASON SHEPHERD: Oh, yeah?
MARK SCHAEFER: I did.
JASON SHEPHERD: Well, hey. There you go.
MARK SCHAEFER: Yeah.
JASON SHEPHERD: Yeah. So I don’t know a lot about it. But you start seeing– it’s, how do we use our powers for good, not evil?
DOUGLAS KARR: There’s a great Black Mirror episode on that.
MARK SCHAEFER: Yeah.
JASON SHEPHERD: Yeah.
MARK SCHAEFER: The famous Nosedive episode.
DOUGLAS KARR: Yeah.
MARK SCHAEFER: Yeah.
DOUGLAS KARR: All right. Well, you guys have scared the heck out of everybody now.
JASON SHEPHERD: All right.
DOUGLAS KARR: Let’s take a different gear. We’ve talked a lot about IoT and its applications in a business setting, but you’re actually doing some amazing things with IoT right now with one of the programmers tracking endangered species. It’s called Data For Good. How does that work? And tell us about the results.
JASON SHEPHERD: Yeah. So Data For Good, it’s actually our partner, SAS analytics, SAS Institute, they’ve been working on that. They’re using Dell gear. They actually work with a nonprofit called WildTrack in this case. So the Data For Good is a broader program.
And what they’re doing there is, they’re looking at how do we use AI and deep learning to kind of train models to think like indigenous trackers. It’s actually more of a core analytics use case than IoT per se, because today they’re using photos of tracks out in the wild. But based on analyzing and training models of footprints, they’re able to start learning about migration patterns and demographics of the different animals– black rhinos, and cheetahs, and the like. And they haven’t published results yet, but they think that it’s going to help them really figure out ways to mitigate a lot of the challenges they’re seeing with endangered species. You look at cheetahs, I think over the past 100 years the population has dropped 90%.
MARK SCHAEFER: Oh, wow.
JASON SHEPHERD: Yeah. So it’s really cool stuff. It’s just an example of the different things that you can do with technology for good.
MARK SCHAEFER: Well, one of the things I’ve been studying about– learning a lot about– I’ve just come out with a new book that really looks at how technology has been used in the marketing area in ways, frankly, that sort of annoy people. And it’s frustrating, because I love marketing. I love our field of marketing. And you look at medicine, or you look at your example, how we’re using technology to save these animals, in medicine we’re using technology to extend lives, and in marketing we’re using technology to annoy people.
And I love my field of marking, but so frustrating that we’re abusing technology in that way. And a lot of the people in my field, Jason, are holding out this hope that with all these data points and all this data collection we’re going to be getting into this wonderful new era of personalization. But personalization is different than personal. How do you see how IoT can help companies be more human, to be more authentic, and connect to people in a more human way? Are you seeing examples that you can point to that could give me hope?
JASON SHEPHERD: Well, it’s tricky. There’s a difference between personalized and personal, as you said. The other portion of it is people are inherently fickle. Their schedules change. You have different moods from day to day or time to time. And so getting to be truly personalized without being overbearing or annoying is hard.
And that’s why a lot of the personalization you see today is really more about absolutes. I walk in the room, I turn on the lights to a kind of a baseline setting. Obviously the home assistants that are proliferating widely are using analytics to kind of figure out your general tastes, but not so much today being able to figure out exactly what you want right now based on current mood and current context.
So the easy examples are kind of the if this, then that programming within your environment, setting to some baseline tastes. You’re driving home to kind of turning on your house so to speak, and all of that. So we’re seeing, obviously, a lot of that kind of stuff. But getting to where it’s really personalized to your current tastes, we’ve got quite a ways to go on that front.
DOUGLAS KARR: Well, and we’ve covered a lot of bases here basically on the trends, the problems, the issues, and everything else. Jason, I’m curious, the scope of what you’re working on right now, and what are the projects that are really exciting that you’re working on?
JASON SHEPHERD: Yeah. For us it’s really, how do we– across Dell technologies we’ve been spending a lot of time on, how do we build more of a consistent infrastructure foundation for scaling out these types of solutions? We’re getting into all kinds of different use cases within manufacturing, and energy, and in transportation, and retail, and stuff like that. Actually, a big push lately in retail, which actually ties back to that personalization element of how do we improve the in-store experience– clearly there’s a lot of challenges that stores are having in their brick and mortar operations with all the online shopping. How do you kind of bridge that experience from online into the stores?
We’ve got some big things that we’re Using And that’s actually an area where you can take the surveillance appliances that we’re building today, which is just kind of about overall loss prevention and safety within stores, and switch it over and start using that for analytics around demographics and triggering new actions within the store. So there’s some cool stuff happening there.
Just a lot of different markets that we’ve been exploring. And those interconnections between the different solutions are where I think it’s really interesting.
MARK SCHAEFER: Jason, it’s interesting. I’m curious, from a leadership perspective, the things– the competencies that you’re building in IoT certainly are going to be relevant for every sort of function at Dell, for every business unit at Dell. So how do you, as a leader of this core competency, organize to collaborate with all these different sub-companies of Dell to make sure everybody’s sort of rowing the same way?
JASON SHEPHERD: It’s a lot of different interaction with the different teams, and kind of working together, and figuring out. Because our whole solution stack is all about, how do I enable any one of the businesses to be valuable independently? Whether you’re using some of the tools that VMWare is building for IoT. Like Pulse IoT Center– how do I manage, monitor, and scale out these solutions in heterogeneous environments? Pivotal, of course, being able to help people modernize how they build new software stacks and the core Dell, Dell EMC infrastructure.
It’s just meeting with the various teams, and making those connections, and kind of figuring out how we’re better together, but then also valuable independently. It just takes time to kind of work through it. And then the big part of this too is our partner ecosystem of different folks that are building value around our infrastructure offerings. It’s just a lot of discussions and building out plans, and collateral, and all that.
And then we’ve got, of course, engineering teams that are kind of stitching it all together.
MARK SCHAEFER: Well, yeah. That must just be an amazing challenge. And, Jason, thank you so much for this just amazing discussion today and really an inspiring discussion. And thank you for your leadership and all that you’re doing right there at the epicenter of IoT, such a critical, critical technology for the world for business and, as we talked about, for social good.
And thanks everyone for listening to Luminaries, where we talked to the brightest minds in tech. This is Mark Schaefer, and on behalf of Doug Karr, we thank you. We ask you to listen in next time. We’d love to hear from you.
Leave us a review on iTunes or your platform of choice. And thanks for listening. We’ll see you next time on Luminaries .
NARRATOR: Luminaries, talking to the brightest minds in tech, a podcast series from Dell Technologies.