Lori MacVittie, Distinguished Engineer at F5, joins Jeff Dance to delve into the future of cloud computing, specifically, the benefits of the technology, the current state of the cloud, and how businesses can create more value from the cloud in the future.
Lori (Guest) – 00:00:00: Nothing is going to scale quite as fast as the cloud, and that’s because they have all those resources at their fingertips already. So if you have to scale very quickly and across the globe, you can’t beat the cloud for that. You really need it for that.
Jeff (Host) – 00:00:19: Welcome to The Future Of, a podcast by Fresh Consulting, where we discuss and learn about the future of different industries, markets and technology verticals. Together, we’ll chat with leaders and experts in the field and discuss how we can shape the future human experience. I’m your host, Jeff Dance. In this episode of The Future Of, we’re joined by Lori MacVittie, an F5 distinguished engineer and cloud technologist to explore the future of the cloud. Lori, for those who don’t know you, can you just spend a minute sharing a bit more about your background and the relevancy to the topic?
Lori (Guest) – 00:01:00: Sure. Hi, thanks for having me. I have been in technology, let’s see, probably by the time this airs, 30 years. I know, it’s exhausting. I started as a software developer. I worked my way up to architect. I went into technology publishing, doing hands-on testing of interesting products, and eventually landed at F5. This was before cloud came on the scene. But by this point I’m well versed in analyzing markets, looking at technology, understanding how they apply to both what F5 does, but the market and customers. So I’ve just kind of applied everything I’ve learned along the way to understanding what is this cloud thing, what can it do for us?
Jeff (Host) – 00:01:52: I noticed that you’ve written several books, or at least helped contribute to several. Can you tell us more about those?
Lori (Guest) – 00:01:58: Sure. Some of them are very developer focused. XAML in a Nutshell. It’s about the XAML language and it’s a nutshell book, so it’s pretty technical. Web Application Security is a Stack. I was very passionate about application security for a long time, decided to write something about that. Later contributed to The Cloud Security Rules, which is really a book about all of the different rules about security in the cloud specifically. We did that very early on because it was a very contentious topic and well, I mean, you could argue it still is today. It’s still a problem. Most recently, I co-authored the Enterprise Architecture for Digital Business. It’s hanging on my wall. You’d think I could just say it, but it’s a new one, so it’s not in my little spiel. And that is really about all of the different domains. Touches heavily on cloud because cloud is a part of that evolution of enterprises as they move toward becoming a digital business.
Jeff (Host) – 00:02:59: It was noted in your profile also that you helped with the CAD profile for NCITS 320-1998. Tell us more about that. Why are you famous for that?
Lori (Guest) – 00:03:10: Famous for that? Maybe infamous. There was early in the days of GIS, there was an effort by the US government, NOAA. To develop an open standard for transmission of geospatial information. So all of the maps that you see today, Google Maps on your car, on your phone, were using all of that data. And SDPs was one of the ways to do that. What we did was then extend that so you could do CAD, right? 3D, CAD, buildings, architectures, all sorts of interesting things. We developed that profile that eventually became that very long and complicated standard that was then adopted by governments and businesses as they were actually building out their architectures and building layouts, all sorts of that for sharing all the information that way.
Jeff (Host) – 00:04:04: Got it. It’s clear you’re an evangelist and a longtime veteran of watching how cloud technology has emerged. And so we’re really excited about your wisdom and your thoughts about the future. But what do you do for fun?
Lori (Guest) – 00:04:19: I play video games.
Jeff (Host) – 00:04:20: Love it.
Lori (Guest) – 00:04:20: Yeah. Tabletop RPGs. I go fishing. I love to go snowmobiling in the winter. Like to mess with my teenage son because lots of fun. Not really a fan of memes, although I see a lot of them. Just a lot of different things, right? I’ll go in like, waves of different activities because there’s just never enough time to do all the things you want to do.
Jeff (Host) – 00:04:47: Awesome. I’m sure there’s not many teenage sons that probably have their moms geek out on video games, so that sounds pretty fun. But that mix of activities also sounds healthy. So let’s talk a little bit about the cloud. Just kind of a little bit of the 101, maybe a few definitions. I know it’s debated, but then let’s talk about the present state and kind of transition to the future state. In your opinion, what is the cloud?
Lori (Guest) – 00:05:09: Somebody else’s computer. Its that. It really is, right? I mean, you are renting compute and power from someone else on a regular basis.
Jeff (Host) – 00:05:22: What about some of these other terms that we hear? There’s hybrid cloud, multi-cloud, serverless, edge. Can you walk us through some of the common terminology that we might experience in our conversation today?
Lori (Guest) – 00:05:35: Sure. All debated names. I love that. Let’s start with multi-cloud because there’s all right, so I would define multi-cloud as using multiple public clouds today. So if you’re using Amazon and Google, you’re multi-cloud. Easy. Hybrid cloud, or hybrid IT would also then incorporate your on-premises private cloud. And those exist. I know some people would say, oh, those are unicorns, but they do exist. And that’s hybrid because you’re basically joining two different, very different environments together. Hybrid. Serverless, which is a complete misnomer because there is a server, you just don’t have to care about it. It’s more appropriate, I’d say, function as a service. I just want to deploy this function somewhere. It’s only going to run once in a while. It’s very cheap. Great. Usually associated with a cloud, public cloud as well. The edge is yet another type of cloud, but it’s closer to the user, and it’s really more appropriate for different types of workloads, like monitoring, supporting IoT, security services, and then some applications that maybe you need to distribute real close to users. Right. There’s only so many public cloud data centers. You can only build so many data centers, but edge ends up being very small data centers all over the world. So you could reach people that you’ve never been able to reach before, who may benefit from actually being able to connect and have access to the same services as everyone else.
Jeff (Host) – 00:07:19: What about infrastructure as a service, kind of platform as a service, software as a service? Those are often associated with the cloud. Can you kind of break down just some examples or how that’s different?
Lori (Guest) – 00:07:30: Sure, let’s start with SaaS, because that’s the easy one. I’m still not convinced it is cloud. So SaaS, software as a service, so that would be like your Salesforce. Concur. Netflix, technically.
Jeff (Host) – 00:07:44: Microsoft 365.
Lori (Guest) – 00:07:47: Exactly. That’s a perfect example. Right? It’s software that’s delivered as a service, so you access it over the Internet. We used to call those application service providers, but then Marc Benioff decided they were SaaS and cloud, and now we have SaaS. Platform as a service is really about providing an entire development platform. I can actually go out and just build an application and use different components from that. So a Tanzu. There’s not a lot that are actually platform as a service.
Jeff (Host) – 00:08:21: Maybe Heroku or Google App Engine, like some of the infrastructure that sits on the cloud to kind of try to help you use it.
Lori (Guest) – 00:08:29: Yes, Google is a good example. They really do target developers and provide a lot of services that you can go out and just build an application from all the middleware, databases, things like that. Amazon provides them, but it’s not quite as seamless or as easy because their focus is really IaaS, which is about the compute, the network, the storage, and then as it grew, the actual application delivery layer and the security layer that you need. So all those services, and those are more architectural than they are developer. It’s very unlikely that a developer is going to go, oh, I need that service. No. Ops or security might do that. As opposed to, like you said, like a Heroku or even Google Cloud, where, oh, I need middleware, I need a messaging framework. Well, I’ll go use Google.
Jeff (Host) – 00:09:19: What about I guess I’ve seen some terminology of, like, cloud 1.0, 2.0, 3.0. I guess they’re referring to kind of the evolution of the cloud. Do you have any examples or help us understand how things have evolved from a cloud perspective?
Lori (Guest) – 00:09:33: So when I hear those terms, I look at it through a lens of just looking at IaaS, because that’s really where my focus would be. It’s more hands on. Right? You start with the bottom at 1.0. It was about, here’s a server and here’s some networking and maybe some storage. That was the first iteration, just go rent that. Very cool. Later on they started adding that application delivery layer. Oh, if people are going to put apps out here, they need things like load balancing and they might need a WAF, right. Some security, they need DNS, they need other services. So Cloud 2.0 was really when the IaaS vendors started delivering differentiated services. Like, here’s all these cool things. That built up marketplaces. So I say really, you get a marketplace, you’re probably Cloud 2.0. Cloud 3.0 is really more about turning that into a platform. So we see that happening today, especially around now. You saw AI enter the scene and just bust everything up. And the cloud vendors are on this very quickly because they can, and they are starting to provide all of these different services for the developers to hook into, whether it’s at the infrastructure layer. So hey, here’s some good CPUs for doing AI because you need that more memory. And here’s these services you can start calling like these different models. So they’re starting to build into a platform. And I see that as really being that evolution of Cloud 3.0 is now we’re turning into an actual platform that has a marketplace and all the core capabilities. And we’re going to allow you to build these really cool solutions on top of us.
Jeff (Host) – 00:11:14: Not just the infrastructure, but a lot of different services they can call on demand.
Lori (Guest) – 00:11:18: Yes.
Jeff (Host) – 00:11:19: Great. In addition to the conversation we had with our guests on today’s episode, we asked another expert to provide their insights on the future.
Duncan (Soundbite) – 00:11:30: Hi, my name is Duncan Epping and I’m a chief technologist working for VMware in the Cloud Infrastructure Business Group. I’ve been with the company for about 15 years in various types of roles, but over the past couple of years, my primary focus has been multi-cloud, and on top of that, storage and availability. I think one of the key reasons that we’ve seen the adoption of cloud growing extremely fast is because of the different expectations that a lot of our customers have in terms of time to market. In the past, when people would develop a product or a service, the expectation was that that new service or that new product will probably come to market six months from now, maybe twelve months from now, depending on what kind of service and products they are selling. But what we are starting to see today is that in a lot of cases when someone is developing a new service, they actually expect that service to be available for their customers literally within weeks or maybe months, of course, depending on the size of the service or the product that they’re offering. Now, if you have a whole procurement process when it comes to hardware for that particular platform where it needs to land on, maybe that procurement process is potentially, let’s say twelve weeks, depending on what, of course, you need to acquire. And in some case, that may even be six months. You can imagine that takes way too long for these particular solutions to be deployed on. So in that particular case, a cloud solution is of course great because you could just simply grab a credit card, swipe it, and just grab the resource that you need to have for that particular platform to be available and then start developing and of course, put it into production. Now what we do see in this particular case, when we have customers developing in-cloud, they don’t necessarily stay within the cloud. So what we also see customers doing is they move their workloads back from cloud to on-premises as soon as they have that platform deployed locally and they have all of the different resources in place to manage those new production workloads and protect those production workloads on top of that. So I think that is one of the reasons we’ve seen this massive explosion in terms of the cloud adoption. It’s mainly the speed that people expect in terms of time to market.
Jeff (Host) – 00:15:24: I went to this conference in San Francisco and they talked about the cloud. This was 2007, but I know, I think was it Eric Schmidt kind of made it more popular in 2006 when he talked about the cloud and then it probably dates back a lot farther. I know it’s debated about when it was first coined, but it almost seems like people use the cloud ubiquitously for the Internet today. But as we kind of get into the present state of where we are, who would you say kind of the big players are right now in the cloud? We just talked about kind of the infrastructure platform and SaaS, but from an infrastructure perspective, the IaaS who are the big players?
Lori (Guest) – 00:16:02: Well, Amazon, Google, Azure, Alibaba and say those are like your big four, right? Yeah, you can’t forget Alibaba, right? We tend to be like, Where are they? No, they’re big. They are really big.
Jeff (Host) – 00:16:16: Interesting. So similar to Amazon, who has an ecosystem for buying things as a store, the world store, Alibaba is not too dissimilar to that. But they’ve also gotten big into the infrastructure as a service play.
Lori (Guest) – 00:16:31: Yes, they do, right, just about everything. So very interesting company to watch.
Jeff (Host) – 00:16:38: Interesting. Okay. As we think about the transition, I remember hearing maybe ten years ago when they were talking about I was talking to other technology leaders and they’re talking about like, oh, the cloud can be now cheaper to rent than it is to kind of buy. Instead of buying the car, I’m going to go rent the car. And there’s sort of this notion that, okay, it can be cheaper to operate in the cloud. And a lot of people started transitioning that way. And now it seems like what I think is like what 90% of computers workloads are done in the cloud something or higher. Where do you see the cloud shining? I know we also see some people still using on-prem, but where do you see like, hey, this is where it shines, this is why it’s why people have switched over?
Lori (Guest) – 00:17:20: So I think there are a few cases where you just can’t compete with cloud. It’s not on cost. We could argue about just the cost of cloud for quite some time, but it’s not on the cost and it’s not on agility, but it’s on scale and reach cloud. You cannot beat the reach of cloud. Cloud can reach across the globe. How can you compete with that? If you want to expand into a global market or even just into another market within the US, you almost need to use the cloud in order to get there, to deliver the experience that people want. Being a gamer, right. The ping cross-country is bad, okay? You do not want that. So you’re going to want to use the cloud. So global reach and just reach is one scale. Nothing is going to scale quite as fast as the cloud. And that’s because they have all those resources at their fingertips already. So if you have to scale very quickly and across the globe, you can’t beat the cloud for that. You really need it for that. The third reason people go to cloud and this was actually something we did research on this last year and found out is the number one reason is still business continuity. In the olden days when my mother was still writing code, they would purchase land somewhere and build a second data center exactly the same. So if something happened to the first one, they could just go like that and still be in business. Why would you do that? You’ve got the cloud. You can effectively build a second data center somewhere, anywhere in the world and have it available in the event you need it. So those three things, it just seems silly not to adopt cloud if those are your use cases.
Jeff (Host) – 00:19:09: Right. It seems like speed is the last great competitive advantage in technology. It’s like we see this with generative AI and kind of the AI race right now. Right? If we talk about scale and the ability to get set up quickly, it seems like cloud is enabling speed. It may not always compute from a cost perspective, but from a speed perspective, it seems like that’s another—if I were to synthesize some of the things you said, is that fair to say? Like, it really helps with speed?
Lori (Guest) – 00:19:42: I think it depends on who you are. Me as an individual developer wanting to get something to market fast, cloud all the way, I’m not going to even try to build out my own thing, right? In a well established enterprise with lots of processes and people and procedures to follow? Okay, those don’t go away just because you’re using cloud. It’s not necessarily technology holding us back. It’s really people in process and how we go through those systems that can still hold you back. It probably does in most cases.
Jeff (Host) – 00:20:17: Yeah, certainly. It’s interesting, there seems to be some trend of people also moving to kind of on-prem and realizing, okay, I don’t need the cloud for everything. And maybe that’s hybrid cloud, multi-cloud sort of trend as well, right. Where it’s like, okay, what do we need the cloud for? What do we not? We need it to scale. We need it for having a redundant data center, maybe, but maybe we don’t need it for the stuff that’s not really changing. I read an article from the founder of 37signals, who’s, who’s famous for Basecamp and software that really hasn’t changed much for the last 13 years, but he seems to buck the trends. And he had a famous article about like, hey, I’m leaving the cloud, and he’s creating his own on-prem infrastructure. And that was in 2022, I think. Like, I don’t know if that was ten months ago or something like that, but then he talked about it, and then there was an article ten months later, hey, our stuff just arrived for building an on-prem. I’m like, almost ten months later, he gets this stuff to be able to build his own on prem, like infrastructure and I was like, that’s not fast. But I could see how this is how it’s debated and how there’s probably a lot of internal planning to think about, well, what do you put on the cloud? What could you put on-prem? But it still seems to be the major trend is keep moving to the cloud because it just keeps getting better and better and better, and we’re adding more and more and more. And how do you compete with all those services that are offered on demand?
Lori (Guest) – 00:21:39: It is, and we’ve seen we call it repatriation. We’ve been in the cloud and now we’re coming home. There’s a variety of reasons we can go back to that cost discussion we wanted to avoid earlier because, yeah, as they always say, compute is cheap, but storage is not. And neither is the transfer of data out of the cloud. Right. That’s a huge cost that’s biting people. They don’t like that. One of the other things we see is that corollary to people in processes. Can we operate as efficiently on premises as we can in the cloud? Well, in 2016, I think it was Google helpfully put out a book that says, here’s how to operate like a cloud provider called SRE Engineering. Woo. Right. So organizations that are adopting that approach and learning from Google tend to be the ones who go, yeah, we’re going to go back on prem because we know how to do it now, because Google gave us all their secrets. Thank you. And they tend to pull back, but not everything. And that goes back to that more strategic use of cloud. We know we want to scale, or we know we want reach, or we know we need business continuity, but what are the business reasons we need to use cloud? And they’re thinking about it in a more mature fashion now.
Jeff (Host) – 00:22:57: Versus a marketing sort of trend fashion. Yeah, okay, that’s helpful. What are some of the recent developments that you’re watching in cloud computing?
Lori (Guest) – 00:23:08: Well, there’s two. So one is of course, the rush to AI and how are the cloud providers supporting it? Because generally speaking, anything that requires that amount of data and compute is going to the cloud because the compute is unbelievable. The memory you need just to run one of these models, it’s incredible. I don’t have that kind of capacity. So you need to be in the cloud. So I’m watching that in terms of analytics and then the AI, but I’m also watching at the infrastructure layer because there’s an infrastructure renaissance, I call it, going on, where suddenly we’ve realized, oh wait, we need hardware to run all this stuff. We still need that. And there’s been a refocus on, well, how can we make all this stuff run faster and more efficiently? So they’re going to all of the specialized chips, we’re going back to ASICs and FPGAs, we’re going to all of the stuff that Nvidia, Broadcom, Intel are developing around running AI workloads—because it’s really math—really fast. And the cloud providers are starting to incorporate that and provide that infrastructure so that you’re getting even more efficiency out of their infrastructure when you’re running those workloads. So I’m watching both of those because it’s very very interesting to see, right? It’s way up here, but it’s also way down here. It’s like the two ends of the stack we’re looking at.
Jeff (Host) – 00:24:44: Interesting. Yeah. I think, as you think about AI, if you’re running an LLM, large language model, I would imagine that spinning up all that compute to do that. And if everyone wants their own LLM to kind of specialize to them, that’s going to create more rise for cloud. Right? Because it has spikes. It has spikes. And how do you handle spikes? That’s where cloud really shines, right?
Lori (Guest) – 00:25:11: Yes, absolutely.
Jeff (Host) – 00:25:13: Okay, that’s helpful. What about this seems to be a growing concern around data privacy and security. You’ve probably been watching it for the last 30 years, in a sense, and helping with it. Thank you. But how are cloud providers kind of addressing these issues right now? Any thoughts on how you see cloud providers sort of addressing security from a privacy perspective?
Lori (Guest) – 00:25:36: I don’t, and I see that personally and I see that professionally, right? Data is gold right now, and that means all data. If you’ve read the EULA and the Terms of Service for Office 365, you know that you are abdicating your right to your data. They can look at it. Same thing with using Gmail. It’s only free because you’re giving them access to your email and your data. And it’s the same is true right now. There are tons of legal questions around AI, if I’m going to train it on my code, if I’m going to train it on my intellectual property that I own in any form, who has access to that? And even if it is only the cloud provider, well, some of us do compete with Amazon in different ways, right. Not at the infrastructure layer, but we provide some of the same services. Who owns that? What kind of a relationship do you have and where’s the data going? These are questions that aren’t answered yet. We don’t know. We’re still deciding who owns the copyright on AI generated art. We’re in new territory here. And it’s very scary because we don’t know the answers yet. And you could get easily locked into something and then find out I don’t own anything anymore. So I think people are being cautious. They’re playing, everyone is exploring, everyone is trying to figure it out, but they’re being very careful about what data they share until we get those answers from someone who has legal authority to give them.
Jeff (Host) – 00:27:21: Behind these companies is also countries, right. And government bodies which also compete and have security concerns. So I can see how they get magnified at multiple layers.
Lori (Guest) – 00:27:34: Yes. And we see that with data sovereignty as well. Cloud is having to answer that question of data sovereignty. I need my data, my customer data needs to be stored in the country where my customer is. So even if I’m headquartered in the US, if my customer is in Norway, maybe their data needs to be stored there. Again, it’s driving people to the cloud because the cloud is coming up with answers to, okay, this is how we’re going to help you comply with those regulations.
Duncan (Soundbite) – 00:28:08: If you would have asked anyone what happened to cloud computing 15 years ago, I would say that the majority of people would not have guessed that we would be where we are today. Now, what I do suspect will happen is that the majority of our customers will actually start adopting this multi-cloud strategic direction. And based on that particular strategy, they will probably be selecting their services depending on cost, performance, efficiency, et cetera. That is something that I think will grow immensely in the upcoming years. Now, of course, it also means that the environment that people will be managing will be a lot more complex. So if you look at it from an IT perspective, you won’t be managing just on premises workloads. On top of that, you’ll also be managing different workloads and services across multiple clouds. And of course, that also means you need to have a monitoring platform available that will allow you to monitor all of these services and the different capabilities in the SLAs that you’ve agreed to with all of these different cloud providers.
Randall (Soundbite) – 00:29:08: What are the top ways AI advancements are affecting cloud usage? There are a couple of ways I see current AI advancements affecting the cloud, mainly with the current functionality of AI. It requires vast amount of data to properly train models, so you need a place to store that, and storing it locally is just a hassle and not reliable to keep around. On top of that, it takes a lot of processing power. Hardware can be expensive, and it can take a long time to run these models. So when you host this in the cloud, you can distribute those processes, distribute the storage of the data, and basically get yourself running much faster. I really foresee this kind of taking off as we get more OpenAI models available to the public, that people can just build their own AI. Essentially they can host it themselves. We’re already seeing some companies and some services for this, such as Steamship or Brancher, that are providing low-code options for this, where you can just basically make a site in the matter of minutes and be able to plug in a few kind of options and ways that people can interact with AI. ChatGPT itself is an Azure-based cloud architecture, and they’re using the cloud heavily for their project. I think when we see larger companies begin to adopt their own AIs for specialized needs, they’ll probably want to host it themselves, most likely in the cloud. They’ll want it to specialize and train it for what they’re doing and what they need it for, specifically within their company or within their areas of expertise. And they’re going to want to have proprietary data on there. They’re not going to want that in ChatGPT or in some of these other services that exists out there. So they’ll really leverage the cloud to build these systems for themselves.
Jeff (Host) – 00:31:02: Gartner is predicting a nice uptick in spend. They’re suggesting 600 billion in spend, so more than a half a trillion by the end of 2023. Google also says that, and they’re biased, probably, but 41% of their respondents said that they’re planning to increase their spend in cloud based services and products. So even though we have multi-cloud and some people thinking about coming on-prem, it seems like the trend is, hey, we’re going to be spending a lot more. Why do you think the cloud is still growing so fast? It’s been growing so fast for like the last ten years. But what are some of the key drivers behind the increase?
Lori (Guest) – 00:31:37: I’m going to be controversial and ask, cloud spend does not mean necessarily more workloads, more apps, more services. If you’re scaling well, and I hope you are right, if you’re growing, you need more of everything. So if you have a workload in the cloud and suddenly you need three workloads in the cloud, you’re spending more. You’re paying more. So one of the questions I have about all of these predictions and these reports about, oh, we’re making more money is but where is it coming from? Is it new customers? Is it new workloads? Or is it the scale and the consumption of existing workloads that are growing? We really don’t know. Probably a combination of both. People are still going to the cloud. They’re building new things for the cloud. But I think there’s some mix-in. Just as there were in early days of cloud, they used to lump SaaS into IaaS in order to make it look bigger.
Jeff (Host) – 00:32:38: Got it.
Lori (Guest) – 00:32:40: Everyone does their accounting tricks. So is it growing? Yes. Is it spectacular? I’m not convinced yet.
Jeff (Host) – 00:32:48: Okay. I guess if we have the majority of our workloads in the cloud and the economy is growing, then we would expect that to grow too, right?
Lori (Guest) – 00:32:57: Yeah.
Jeff (Host) – 00:32:58: McKinsey has said that the cloud has immense potential, but most companies are still only scratching the surface. Again, maybe that’s maybe they’re in bed with some of the cloud providers in that report as consultants. But one of the things they cited was recently with the pandemic, how Moderna had used the cloud to quickly sequence their kind of virus and kind of position them in the fly. And so they said, well, without the cloud, they wouldn’t have been able to use all that compute power to do what they needed to do. And so they talked about how other companies could do things faster and had this prediction of picking up another trillion in value from the cloud by the year 2030. Do you buy into this, that, hey, we can create a lot more value from the cloud in the future, or do you think this is more marketing?
Lori (Guest) – 00:33:47: No, I think that goes back to that analytics and somewhat AI use case. Right? Because those kind of analytics that are looking for patterns and sequencing are all based on it doing ML and they’re doing a lot of math and they’re doing a lot of processing. You do kind of need the cloud. Most people have already decided that they’re going to be doing all of that kind of analytics in the cloud and they’re definitely going to do AI in the cloud because otherwise you can’t scale it that fast. You can’t run those kind of workloads. It’s expensive and prohibitive. Right. So I think there is more value and I see there may be more focus on that by certain cloud providers, or maybe new ones will crop up to say, hey, this is what we do. We just do analytics, we do AI, and that’s all we’re going to do. And I can see that market cropping up as well.
Jeff (Host) – 00:34:42: Centered around AI. Yeah, it seems in the last three to four months, the whole generative AI AML trend has just exploded. It’s like ChatGPT now has like what, 2 billion users between them and the ecosystem and they went to zero to 100 million and faster than any other technology has. Obviously that has impact on the cloud. But any other thoughts related to what we’re seeing right now with generative AI?
Lori (Guest) – 00:35:10: Wow, it’s just going to keep going. I think just like the cloud, early on, people were trying to find use cases for it and right now people are exploring. We’re trying to figure out what is it good at, what can I trust it at? Right? It’s generative AI. It’s not smart, it doesn’t know the answers. There’s a whole lot of probability and statistics that back up what it gives you. So that’s why it’s sometimes wrong. So what can I trust it to do? What can it do predictably and reliably? How can I leverage it? All those kinds of things. It’s fun, right? I mean, the kids are playing with it. You don’t want to know what they’re doing with it, but I mean, they’re doing stuff with it, that’s for sure. But they’re learning it. Right. So we’re all exploring where it’s going to go. We know it’s going somewhere. We also know that we’ve only scratched the surface of what can it do, how can I apply it? Right? We started to see generative user interfaces where it’s building a user interface. An entire app. Well, not an entire app, but the user interface dynamically. Right. So wow, that’s cool. You don’t even need to do that anymore. As a developer, I’m happy. I don’t want to do that right. So those kind of things are still coming up. They just released AutoGPT, so doing agents, right? Doing RPA type work automation. Now we’re digging into the automation side. It will autonomously go do a task for me. So now when does that start playing into infrastructure and IT? They have a lot of tasks that they do to operate and deploy all of the applications that they already deal with. How might AI impact that? That’s exciting to me. We haven’t even started playing with that.
Jeff (Host) – 00:37:05: Yeah, it is really interesting how I think the typical person’s experience with generative AI is sort of ChatGPT. But the proliferation of the business applications like things like Unreal Engine or MetaHuman or what Adobe just released with Firefly, like all these other applications that are using those engines to then create. That’s where I think it’s sort of like we’re seeing this wave of new applications. Like just in March we tracked like 50 different news events around significant advances that other companies are releasing. And it seems like from a business perspective we all know that it’s going to impact business. Kind of like how maybe we saw the static web go to the more dynamic web where we’re now using all the software that’s collaborative or connected but we haven’t quite seen the business applications yet. Right? We’re seeing the software but we know it’s going to change us but we don’t know exactly how yet. And it feels like we are just scratching the surface. So I’m just kind of repeating what you said. But all of this has an impact from a cloud perspective, right? All of these generative applications that we’re seeing, aren’t they just tying more and more compute?
Lori (Guest) – 00:38:17: Oh yeah, a lot more compute a lot more in order to get up that fast right now. Right? All these little because it’s all independent kind of developers, startups, little groups, right? They are moving fast and yes, they’re using the cloud because they need that kind of power but they also need that kind of and you alluded to that, right? Speed. A small I don’t do anything else but this, bam, in the cloud, it’s up and running and now it’s got all the power it needs so it’s going to keep fueling that. And you can see cloud is going to move more and more toward supporting those kind of workloads already because they’re already announcing all of the ways that you can use it and they’re providing services. They’re moving fast because they know they have to get ahead of this because this ecosystem is like already moving, right? And they got to get ahead of it. So they’re going to continue to build that and I think we’ll see that explode in terms of what you can do in the cloud very quickly just because of the natural synergy between the two.
Jeff (Host) – 00:39:20: Yeah, that maybe ties into industry clouds. I’ve been reading more about industry clouds where it’s like special clouds dedicated to kind of industries. Is that a good example where like maybe you’re—because I don’t really understand the term industry cloud, but is it sort of the amalgamation of services that are kind of specialized to an industry?
Lori (Guest) – 00:39:40: Yes, and also to the regulations and requirements around security. You see it for things like the federal government FedRAMP, you see it for healthcare. They really want that constrained because they have very special restrictions on access and security and all of those kind of things. The data, sovereignty, rules, HIPAA, you name it. And those clouds are tailored to we support all of your special regulations and all of your things so you can be compliant and you can be secure, but you can still use the cloud.
Jeff (Host) – 00:40:16: Okay, those are good examples. Government kind of, healthcare in particular. And then I could see other especially as we get into AI, if you have more and more services dedicated around a certain space, maybe like there’s a gaming cloud.
Lori (Guest) – 00:40:34: Yeah, it’s Nvidia, right. You can do that now. You can run it off your mobile phone now, which is some games not so good for, but for a lot of them they are so they’re already gaming is like my favorite use case for edge. Right. Reduce my pain, please. Right. No lag. No lag. So we’ll see those kind of things crop up.
Jeff (Host) – 00:40:56: Nice. You mentioned edge and we’re in the robotics space, so we’re working more extensively on the edge as well. It seems like the cloud providers has also been trying to cater to that industry, IoT, robotics, kind of edge. And any thoughts on the trends we’re seeing there?
Lori (Guest) – 00:41:14: Yes, I see them trying. I see them not expanding far enough. Their definition of edge is not—it’s kind of like sort of edge. Close. But you talk about things like robotics, but manufacturing any kind of monitoring things that need immediate response, they can’t wait. You think that subsecond response is fast enough, but it’s not. It’s got to be subsubsecond, right? For some of these things, especially around safety of people. We have a lot of paper mills in Wisconsin, and a lot of them are very automated, very automated. There’s everything running. They can’t even go out to an edge node that’s in Chicago. It’s too far. It will not respond fast enough. If you need something to shut down, you need to shut it down now because the safety of a human being is involved. So the cloud providers are trying and I think for some edge cases it suffices, but for other ones, especially around OT, I think that’s where you need it even further out, you need a service provider cloud at the edge. Right? Service providers at the edge, probably.
Jeff (Host) – 00:42:28: It seems like that’s probably the case for a lot of robotics applications as well, where it’s like safety. The proliferation of robotics is sort of more integration with humans. But if you’re going to be integrated with humans, you need to be responding on a dime because a little movement could mean you hit someone. Right. And that seems to be the case also for AR/VR, right. Where all the latency issues that kind of go along with that being successful. So I would assume your distance to the cloud, your cloud, your servers, that time matters as well, from that perspective.
Lori (Guest) – 00:43:04: Yes, it really does. If you’re right there with near one of the cloud data centers, you’re gold because that is like nothing. If you’re on the fiber ring and you’re right, you’ve got all the speeds you need. But me, no, I don’t have that. I don’t have fiber, I’m not on a fiber ring. I think the closest one is an hour and a half away. You’re kind of like you’re just not going to get that kind of speed, so it’s not going to work as well.
Duncan (Soundbite) – 00:43:37: One of the most recent pivotal developments in terms of cloud computing, in my opinion, isn’t necessarily technology, but I think the biggest change has been the change in mindset that customers have towards cloud. Over the past decade or so, a lot of customers were talking about public cloud as a strategic direction and cloud only and cloud first. But what I’ve noticed over the last year or last couple of years is that a lot of customers are now talking about hybrid cloud or in this particular scenario, they’re probably mostly talking about multi-cloud solutions. And the key reason for it being is that now what they’re actually doing is they’re trying to figure out what it is that their platform, their application, their end users, their line of business actually require. And based on those requirements, they select a platform where it should land on. This not only provides them the most flexibility, but hopefully in a lot of cases, also provides them the most cost-effective solution and the best solution as well from a time-to-market perspective. And I think that is probably one of the biggest changes that we’ve seen over the last couple of years, which is the mindset.
Jeff (Host) – 00:44:37: As we think about the future. Any other predictions, any other thoughts on kind of where things are going?
Lori (Guest) – 00:44:43: I mean, I think the cloud stays relevant for a long time, a very long time. I think what it focuses and where it focuses will shift. I don’t think it continued to grow the way it has in terms of just becoming that big and being everything to everyone. I think at some point in history and the cycles of technology and innovation and standardization, modification, innovation, that same cycle happens over and over and at some point you can’t expand anymore. It’s just too big. You almost have to it starts all over again, like, I’m going to solve this problem and we see this in multi-cloud. One of the big challenges people have is I want to operate in two, three, clouds. I also want to be on-prem, but all of you are different. You use different languages, different tools, different APIs. This is a mess. I can’t do anything quickly anymore, and it costs a lot. So how do I solve that? Well, that’s where we’re at. It has grown to the point where, okay, here we are. How do we fix it now? So I think there’s going to be a lot of movement toward how do we deal with that so that people can use the cloud more in order to get that growth again, except in those specific areas. So I think we’re at that point. It’s a tipping point. It’s time to tipping point.
Jeff (Host) – 00:46:13: We’ve grown so fast. We got to fix some things now.
Lori (Guest) – 00:46:16: Yeah, pretty much.
Jeff (Host) – 00:46:18: Okay, I have a few more questions to kind of wrap up. I’ve really enjoyed your insight so far. What’s been rewarding for you, having been so deep in this space? What have been some of the rewarding kind of professional experiences you’ve had just in your experience?
Lori (Guest) – 00:46:32: I have really enjoyed being able to go out to conferences or just interact with people and share knowledge and information. Not even necessarily talking a product or something, but just being able to share that knowledge so that they understand a topic or learn something new. That’s very rewarding. It’s very rewarding to share that and see people understand technology suddenly or get an idea and be able to go execute on it. So that’s very rewarding. It’s kind of fun to watch the evolution of an emergence of new technologies and then be able to say, okay, what does this mean? What is it going to do in five years? And since I’ve been around long enough, I’ve been able to watch and go, was I right? Play armchair quarterback X years later and go, is that right? Yeah. All right. Oh, got that one wrong, but I got that one right. That’s kind of rewarding, especially when you get it right because it helps you the next cycle be able to go, no, this is what’s going to happen based on X and Y and Z. So that’s very rewarding.
Jeff (Host) – 00:47:40: What about you’re a woman leader in tech, and you’re kind of more rare. Even though we see the advancements, we see some of the growth, who are some other women in tech that inspire you, that you interact with frequently?
Lori (Guest) – 00:47:54: So Lydia Leong, who is the queen of cloud. Anytime I end up on a list with her on the same list, I’m like, I’ve made it. There’s nothing better than this. She is just so insightful, so knowledgeable, and she’s just great. Wendy Nather, who is the CISO over at Duo and Cisco, I believe. Great sense of humor, great leader, really informed, really knowledgeable, and really fun. I love interacting with her and just watching what she has to say.
Jeff (Host) – 00:48:32: What about in history? Can you think of some women in history that have inspired you as you look back, that have been in tech?
Lori (Guest) – 00:48:40: In tech. So I’ll list too, because I think it’s requisite that you mentioned Grace Hopper, right, of course, the good Admiral. But then I guess I would say my mother. I think that’s one of the reasons I went into tech is because she was in that early wave of tech and got me involved. I mean, you’re playing with punch cards, you’re like, what is this? And she gave us our first computer and encouraged us to explore it. I think that’s very influential in actually following this path and going into technology.
Jeff (Host) – 00:49:17: You mentioned you had a teenage son. As you think about those that are coming out of college or going into college or just specializing in what they’re going to do in their career, any advice for kind of young professionals that are emerging on the scene?
Lori (Guest) – 00:49:33: So what I told him, and the fun part is he’s got an older brother who’s already in data science and doing AI, so I get to hear both sides of it. But this one is less into. He doesn’t want to code, but he does want to play with it. And when this started right around Christmas, I told him and I would tell any young person, go play with it, go figure out what it can do. What it can’t do. Explore it. When you think, how would I do this search? Maybe go ask an AI instead. Look at the image processing. What can it do? How does it change? When you use different prompts? When I say this instead of that, look at what it does differently. Just explore it, get comfortable with it and don’t be afraid of it because it’s going to be a part of your life and a part of your career and you need to be comfortable with it and start getting it into your kind of routine and your awareness. You can’t ignore it.
Jeff (Host) – 00:50:34: Yeah. Any other advice for our audience or perspectives on the cloud before we wrap up?
Lori (Guest) – 00:50:39: No, just keep thinking about how can you use it just like any of technology, like we just said about generative AI, how can I use the cloud best? What is it best suited for? I have this thing to do. What thing might it work for? Right, there’s sure to be an application. Everybody can use the cloud in some way. You just have to think about it and actually deliberately evaluate whether it’s the right thing for this project or that project every time you go to do it.
Jeff (Host) – 00:51:10: Thanks, Lori. It’s been great having you. Appreciate your battle-tested experience and the depth of your insight and just your personality. It’s been fun to have you on the show.
Lori (Guest) – 00:51:20: Thank you for inviting me. I really enjoyed it.
Jeff (Host) – 00:51:25: The Future Of podcast is brought to you by Fresh Consulting. To find out more about how we pair design and technology together to shape the future, visit us at freshconsulting.com. Make sure to search for The Future Of on Apple Podcasts, Spotify, Google Podcasts, or anywhere else podcasts are found. Make sure to click subscribe so you don’t miss any of our future episodes, and on behalf of our team here at Fresh, thank you for listening.