Insider's Guide to Cloud Computing

Why An Insider View on Cloud Computing

Prasad Rao: Hello, and welcome all to another episode of "GOTO Book Club." I'm Prasad Rao Rao, author of the book, Cloud Career Journeys." And today we have David Linthicum with us. David, would you like to introduce yourself?

David Linthicum: Hey. Sure. David Linthicum Linthicum, it's great to be here, number one. I think your book was amazing, and enjoyed talking to you about it. I've been in the industry for probably 35 years, the last 20 years in the cloud. An architect by trade. Pretty much have held different technology executive roles. My last one was as chief cloud security officer at Deloitte for about six years. Before that, I worked with a company called Cloud Technology Partners, one of the original employees. And prior to that, I was CEO a few times, and then CTO four times, different companies that I joined and sold. So, as an independent influencer, and analyst, now, and just having a good time talking to guys like you.

Prasad Rao: Cool. And today, we'll be talking about your book, "Insider's Guide to Cloud Computing." So, tell me, like, you know, what inspired you to write this book and share your insider knowledge about cloud computing?

David Linthicum: Because a lot of people told me I should write a book when they would hire me as an influencer consultant, things like that. My persona is to tell the truth and tell them when cloud would work and cloud wouldn't work, and why you would want to use it and why you wouldn't want to use it. So, never a shill for any kind of technology, including the cloud. So, somebody told me, and said, "Well, you need to write a book about the insider information on this stuff. because no one's saying this stuff." So, I took all the years of experience in cloud computing, as a cloud architect, and those sorts of things, and jotted them down in an insider's book that kind of gets to the back-end details that no one was able to or wanted to talk about. Everybody was pretty much gung-ho cloud, which is fine, but there are reasons to use it and reasons not to use it.

And I think that even the cloud providers agree that they want people to use it in successful ways. And the way and patterns people are leveraging cloud these days is not so much. In 2022, we got the big bills, people moved to...lifted and shifted into the cloud, weren't really paying attention to the optimization of the applications in that environment. Now we're quickly trying to re-platform, redo, and modernize those systems to lower the bills and optimize those systems. I think I heard the word optimize from the cloud providers last year for the first time. I think they're all in on making these things happen, the rise of FinOps, all these sorts of things. So it really is getting into, we're well past 10 years of cloud computing, making this stuff work, figuring out where it's a good value, where it's not a good value. And I think people are ready for that message. And that's why I wrote the book.

Prasad Rao: Awesome. And we'll dive deeper into the book. But before that, like, you're a bestselling author, you're a course instructor, a keynote speaker, and you're, like, a lot many more things. Like, you have authored 15 books, and probably more than 7000 articles. So apart from all your professional commitments that you have, like, how do you find time for writing, and maintaining discipline among so many other things?

David Linthicum: Well, it's just part of a daily ritual. In other words, I think if you do something for a month at a time, you force yourself to do it, whether it's exercise, reading a book, or doing yoga, or whatever, it creates a habit. And you're going to kind of stick to the habit. I realized early on when I was in my 20s that I had to differentiate myself somehow. I didn't have an Ivy League degree, you know, didn't go to MIT, went to a normal college. And if I was going to accelerate my career, I needed to present myself as a thought leader in space. And so I started to write articles for local consulting journals, and, you know, user groups, things like that, and some magazines picked it up. And that was the only way to get something published back in the 90s and the 80s, was getting into a magazine. Started writing for "PC Magazine." And it kind of steamrolled after that. It was a nice side gig, but also symbiotic with my job. In other words, I had to keep up with technology to write about it.

So, it was the way to learn in a faster way, to force myself to learn. And I could bring that knowledge into my role as an architect. So, all of these things kind of work and play well together. So I learned early on to get up early, and write 2000 words. I write about 2000 words a day, whether it's my InfoWorld blog, my book, or, you know, even show notes for the video I'm doing later. And I enjoy it. And it's funny, I did not enjoy it when I first started. It was drudgery to me. I just figured out I needed to do it to accelerate my career, so I forced myself to do it. Now it's something I really enjoy. I always look forward to writing my InfoWorld blog every week, just because it's a new idea. And it's a challenge to see what I can put out there, that's going to get the most interest. You know, now that we have different multimedia facets, such as podcasting, making sure the podcasts get out there, and also videos, and all these sorts of things that are new media to us.

And it's just amazing and empowering to me that anybody in the world can sit down and write a blog. Get a Substack account or a Blogger account, probably dating myself, or publish on LinkedIn. And anybody can read it. And so it's not like in the days back when I was writing for "PC Magazine," only I could write for "PC Magazine," and the other writers. It was a very limited world. Now it's an explosion. So I just enjoy the freedom of it. I enjoy the fact I can get ideas out there without them being edited. So to me, it's just an empowering thing that makes me feel great. Just like you guys, in writing your book, I'm sure you had some days where you didn't want to write, but you did it anyway. And you have to have the discipline of going back and keeping writing.

Prasad Rao: But 2000 words a day, that's amazing and very impressive. And probably a note to everyone in their 20s who are not from a big college, right? That's a great tip there, how you can actually make yourself stand out. So thanks for sharing it.

Recommended talk: Cloud Career Journeys • Prasad Rao & David Linthicum • GOTO 2024

Demystifying Cloud Computing: Industry Secrets and Common Misconceptions

Prasad Rao: Now let's dive deep into the book. And you mentioned that many secrets in the cloud computing industry could have, you know, like, or can help enterprises succeed. Do you want to talk about some of them and some secrets or misconceptions that companies should be aware of?

David Linthicum: Sure. I think the biggest one that people are surprised about is that, in many instances, cloud-native systems, containers, and Kubernetes, things like that, are going to be not necessarily a requirement for all applications. Which is almost a religion out there to move in those directions, and to take existing applications and write new applications, in containers and Kubernetes. I've built hundreds of architectures that way, and it's great technology, but it shouldn't always be used. And there's something I call a container tax that comes along with doing that because you need more expensive people, you need more time. It's operated in a different way. And you have to account for that complexity when you build those systems. And I think that revealing that was something that was eye-opening for lots of folks because I just wanted them to evaluate the use case that they're solving when they move into containers.

Same thing with AI. AI, I've been an AI developer since I was 18. I worked with M-One and Lisp, taught as a young college professor back in the day, and watched the whole thing generate over time. We had the same problem back then we may be having now, it's overused. In other words, we're trying to apply it to different problem domains. You gotta remember, in the 80s, there was a big boom in AI. People were, you know, buying the Borland tools, things like that, at the local bookstores. I'm sure it's well before your time. And it was overused. People were using it to write business applications and transaction applications. Now, the AI back then was not as capable as it is now. It's multitudes of power.

But the ability to pick the right use cases to apply that technology is something that I don't think we understand right now. Generative AI is cheap, it's available, you can use it in the cloud, fairly easy, and you can become a generative AI developer in a couple of days. However, you can also reuse it. And you're gonna find that it's very expensive. Leverage is very expensive. Processing, that burns a lot of power. GPUs are great, but they actually need a processor themselves to run, so they take a lot more power to run. And so it's just really kind of looking at this with a pragmatic eye. And not getting us into trouble, which I think we're getting into right now, in a few years' time, when we realize that these applications should not have been written in that particular technology. Also, the big thing was, in chapter one, I talked about finding the value of cloud computing and admitting to the fact it never was CapEx versus OpEx. That never was the way to define the value of cloud computing.

If you're not able to define agility as kind of a core critical value, a hard value, and a sock value of cloud computing, normally, it's not going to have a good business case for cloud computing. And the applications shouldn't be moving in those areas. So, the ability to get people to understand that they're not going to save more costs than cloud costs. In other words, clouds are very expensive relative to the on-premise systems. And that's because they have come down in price a great deal. But there are reasons to use it. And if you put a lot of value on agility and the ability to speed the market, integration with various systems. For example, we're building an AI system and we need access to data tooling and technology that's important to us for this aspect. That's gonna be a business case for using the cloud, but it's also a business case for not using the cloud. If the application is isolated, it does very simplistic things, and uses massive amounts of data stories that are isolated unto themselves. And a lot of applications are like that, they're very monolithic. Then leaving it on-premise in many instances is going to be your best bet.

That doesn't mean you can't move it to the cloud in the future, as the capabilities change. But that's something that should be looked at on a case-by-case basis. The other thing is, that everybody predicted the death of the mainframe. Those things are still around. So, hugely powerful processors. They measure their memory in terabytes, which tells you how big they are. And there are reasons to use those sorts of big iron systems. And in many instances, the clouds are not going to displace the applications that are running on there. It may be in the future, and you can certainly find mainframe analogs on the cloud. And I've built systems like that. But it's not necessarily an efficient process. And it's not going to get you to a better state.

So it's having people step back and look at the cloud industry for what it is, something that can have potential, do a lot of potential good. And something that we can make a lot of mistakes in. So, your ability to avoid mistakes, pick the right use cases, and use the cloud to a fundamental value, is going to help everybody. It's going to help the cloud providers. It's going to help AWS, Microsoft, Google. It's going to help the consultants out there. And the big thing, is it's going to help the enterprises that are consuming this stuff and trying to run the budgets.

Prasad Rao: Totally agree. And that chapter was really insightful when I was reading it. No one says it in that specific terms, that it never was about CapEx and OpEx. It's always about the value. Great way to put it actually.

From Overspending to Optimization: Strategies for the Cloud

Prasad Rao: Now, you also talk about two types of organizations. One, you know, that they know that they're overspending. And others, that they know that they're doing it right. How can the companies transition from that first category to the second category?

David Linthicum: Just really good at planning. I mean, it's not a... I don't think it's a lot of rocket science or an algorithm I can give you, so you do this better. It's just doing additional planning before you start moving into these areas. And we're seeing this lack of planning that's occurring now in the movement to generative AI. So, why are you doing it? What's the business case? How are you leveraging the technology? How are you going to operate it? How are you going to deal with the complexity of operations? Your hiring patterns, the skills changes that you need. All these sorts of things should be known as you get in there. And all this stuff... With many organizations, probably 80% of the organizations out there, it's a big surprise. In other words, they realize that they can't find a lot of cloud talent, because there's a scarcity of cloud talent out there, good ones. There's not enough people like yourself out there looking for jobs. And therefore, they find it a tougher path to go down, much more expensive. And they're not really seeing the benefits there when that should have been on the radar screen early on.

They should have made decisions around application workloads and datasets that are going to be aligned better to what the business is able to do and align to the use of cloud computing. So planning is not an exciting thing to do. You're going to have to, you know, figure out what your data is, figure out who owns it, figure out performance characteristics, figure out cost and budget characteristics, all these sorts of things that architects are supposed to be doing. But come up with a solution, both a macro solution. In other words, a solution for the entire enterprise. And the micro solution, solutions for the particular systems that are there, that are going to be the best logical conclusion to return the most value back to the business. And I state that probably 100 times in the book. We're measured in our ability to leverage cloud computing successfully, the ability to return the maximum amount of value back to the business. If we're not looking at that metric, we're gonna go off kilter each and every time. So it's the dead answer.

You know, do some planning, do your budgeting, make sure you understand the cost, do your due diligence before you move any of these systems into the cloud, and then you'll be part of the team that...the business that does it right. And that's probably 20% of the enterprises out there leveraging cloud, are able to do it kind of right, 80% down. And they're bleeding money, they're hurting their business, they're putting resources often in areas that are going to be less productive. And now that we're moving into the whole generative AI way in the cloud, then those sorts of things are going to be business-killing. They're going to spend money that probably could take down the business. And also the business could be missing out on a key innovative differentiator that could change the business and actually become the business. So the stakes are way higher now. And so, I'm telling people, you don't do this planning, you don't do this architecture, you're gonna run into these walls, and you're gonna kill your business.

And I'm seeing it happen right in front of my eyes, by the way. You're gonna see not necessarily bankruptcies, but lots of businesses getting sold or absorbed in other companies. And when you figure out, do a post mortem on what happened, they had a weakness in their business because they weren't spending their IT dollars in areas that were...to a strategic benefit for them, the competitors did. They were able to build generative AI systems in the cloud or not in the cloud, that were able to provide a key differentiator for them, where they'd provide a better supply chain, better customer experience, and able to do things cheaper. You know, all these things that businesses can do, ala Uber and Netflix, things like that. And they're just gonna basically dominate and kill the business. Do you remember Netflix and Blockbuster? I mean, that's the best analyte you could think about there. So you want to be Netflix, you don't want to be Blockbuster.

Prasad Rao: Continuous innovation is important. And cause should always align with revenue. That's what you also mentioned. You know, when you're thinking about doing something, think about, you know, whether the cause is going to help you with the business case, with the revenue, or it's just like burning money. .

David Linthicum: Don't burn money.

Recommended talk: Patterns of Distributed Systems • Unmesh Joshi & James Lewis • GOTO 2024

Multi-cloud Strategies

Prasad Rao: Talking about multi-cloud. You have a full section in the book about multi-cloud strategies, and in-depth, you know, discussion about the multi-cloud upsides and downsides. Would you like to share a few insights about that?

David Linthicum: It's normally going to happen to you, whether you want it to happen to you or not. Everybody I talk to, they always tells me they're using AWS as a primary provider, and they're not going to support multi-cloud, but they're always going to have a... Every one of those has moved into multi-cloud environments. And that's because of the best-of-breed systems that may not exist in a particular cloud. So in other words, where primary cloud providers, AWS, we find most of their stuff you're able to use in AWS. We may need to use an AI system, however, that only Microsoft hosts for some sort of a special business reason. And so they're multi-cloud. So, it happens to everybody organically normally. It's not planned. They're not, you know, stating, we're going to definitely go to multi-cloud. Different pods of developers are making decisions around what best-of-breed technology to make. And they're empowered to pick whatever cloud providers that they want normally. And they should be. They should be able to use the best-of-breed technology that are fit for their purpose.

Recommended talk: Enabling Developers in a Multi-Cloud World in 5 Minutes • Mauricio Salatino • GOTO 2023

And that's going to return the most value back to the business, again that metric again. So, multi-cloud is going to happen to you whether you like it or not, so you might as well figure out how to manage it correctly. And even if normally you're gonna have an on-premise infrastructure, some edge-based computing, you're going to support at least three cloud providers. If not now, it's going to be in the future. And I guarantee, if you don't think you're there, now you are. Somebody else's leveraging another cloud. You're leveraging SaaS providers like Salesforce, things like that. So, if that's the case, then how do you approach the operations of it. We have a complexity problem. We went from managing, you know, 3000 cloud services under management, to 9000 cloud services under management. And we're using these in different patterns, and by different providers, and different operational queues.

So we can silo around the clouds. We can go, I'm going to use the AWS security system, AWS various operations, AWS governance, and the same with Microsoft and same with Google. That's not going to scale because you have to have people who are able to maintain this, people who are able to maintain this, people who are able to maintain this. Normally, they're not communicating with one another. And you're doing everything in a siloed approach. And certainly the on-premise systems, the same thing. So, the ability to look at a whole super cloud, meta cloud infrastructure, which all it is saying, it's a fancy marketing word for saying that we're going to put a layer of technology, security operations, developments, all those sorts of things, are able to run across clouds. That's all it is. And even on-premise, I recommend you also extend this to the on-premise system. So we're dealing with security in the same way. Now, we may be leveraging native security as the APIs underneath the scenes that are doing these things. But we're not dealing with different layers and doing it. We're doing security in one layer, we're doing governance in one layer, operations in one layer, metadata management in one layer, as much as possible.

The ability to move those systems up to a set of common services is the only way you're going to be able to do it. So I write about that in the book. I even have a course out on LinkedIn Learning called Cloud Complexity Management, which shows you how to build, in essence, a meta cloud and super cloud. Longest course I've ever done, three hours. And it's getting people thinking about how they're managing these multi-cloud solutions. Right now, they're throwing away money. Everybody who's running a multi-cloud out there, I guarantee they're operating in silos. Different people are operating in different specialties, things like that. They're paying too much for managing those cloud systems. They're not getting the same amount of value back to it. It's overly complex and very heterogeneous. So you're gonna have to mediate that complexity somehow. So that's my core message in the book, and I even tell you how to do it.

Unified Security Management

Prasad Rao: So any specific tips that you would have for organizations on how they can manage security across all the cloud platforms they are on and on-prem, instead of using the native security tools?

David Linthicum: That's a great question. I think it is the ability to use security managers that are able to operate across clouds. And we're seeing some tools from IBM, and Splunk, and all these sorts of things, who are able to... It's not going to be one tool that's going to provide a unified layer. It's going to be different tools and technologies that are using whatever native APIs that security providers have, on-premise, Google, Microsoft, AWS, and whatever micro cloud you may have in place. The idea is to pick that layer, but don't necessarily focus on a technology or a tool set, but understand logically what you're trying to do. Because that layer is going to change a lot over time. But the ability to pull up as much as you can to a common layer. And some of that stuff is still going to have to exist in a silo. We live in a practical world. But the layer that's able to use...security managers that are able to use native APIs to control the security on each of those platforms. So in other words, so the platform, they're being monitored, they're being secured. SecOps is going on. And they're getting the security services they think they need.

But to the humans that are managing it, we're not doing so with different dashboards, different consoles, and different processes. We're doing so with a single set of tools, where the orchestration, the automation, the AI capabilities, all those sorts of things exist in this layer. And by the way, it's an old architectural trick. In other words, as soon as we got into multiple systems, distributed systems, the commonality was something that we always wanted to do as an architect. We just lost track of it, because the clouds unto themselves kind of became silos of automation in and of themselves. And the cloud providers don't tell you to use these tools, they want you to use their tools. But in using these tools, you actually consume more cloud, but make more money. But the technology started to rise out there, ops, security, governance, metadata management, all these sorts of things. And, you know, it has a name, it's called super cloud, I call it meta cloud. But, you know, whatever you want to call it. It's the common layer that runs between the cloud providers and also extends out to the on-premise systems as well.

Recommended talk: Decoding Modern Tech: Cloud, APIs, Wasm, Security, & More • Daniel Bryant & Matt Turner • GOTO 2023

Guidelines for Strategic Cloud Adoption and Talent Acquisition

Prasad Rao: Now, these terminologies can be a bit overwhelming, and you, you know, like, touched upon that also in the book for the companies who are starting their journey. So what advice would you give to a company that is thinking about making the move into the cloud?

David Linthicum: I would say, understand your own business reasons for doing so, which is a tough thing to do. Most people... And it certainly was the case back in 2000 to 2011. They would call you in to do a consultation on moving into the cloud without an understanding of what purpose they were doing it. That was the question that I asked, first of all. In other words, why are you doing it? What reason are you looking to move in the cloud? And normally, you get a lot of blank stares. You know, here we are in 2024, and those blank stares still exist. So, there should be a reason or a purpose for doing it. That translates into a plan, which translates into a budget and a business case. And once we get through that, then we can figure out the path to incrementally move into the cloud. I don't believe that every company should have every system in the cloud. Some of the smaller companies out there, believe they can save money, sometimes they can, and sometimes they can't.

So, we have to look at the plethora of platforms out there, to make sure they're making the best choice. What happened to the cloud business case over the last 10 years, the price of storage and the price of processing on on-premise systems just came down like a rocket, a 45-degree angle. You can Google the graphs out there. HDD storage, for example, it's just dirt cheap, almost free. So, if we're leveraging the cloud, that's the alternative to leverage some of these on-premise systems. So it has to be some sort of a compelling reason to leverage the cloud. And you already talked about a few of them. Are we able to find the soft values in the cloud? In other words, the ability to find agility, find business benefits, the ability to have things change on a dime, and the ability to use an ecosystem. What's the reason you want to use things in the cloud? It's easier to build AI systems in the cloud, for example.

However, most of the applications I see out there are monolithic. The data is tightly coupled to the application. And there's not going to be a business benefit from them moving into the cloud unless they make significant changes to that application. And if they don't, then it's going to cost them a lot more money. So, it's going through these very simple, I think, almost commonsensical, you know, kind of questions that businesses need to ask themselves. There's just so much hype and so much marketing aligned with the cloud. People love going to reinvent it. It's like Woodstock for cloud geeks. And I do too. But you have to come out of there with a pragmatic understanding of how it works for you. Not how it works on stage, at the keynotes, but how it's going to work in your organization for the purposes that you're making. Those are the tougher questions. And I hate being the designated buzzkill, but I think someone has to ask the questions.

Prasad Rao: No, and you raise quite the right points, right? Like, there are so many services, which one would you leverage? And why? And how would you do it optimally? Right? That's the main thing. If you're going to simply lift and shift, obviously, your cost is going to rise. But how are you able to actually use it? And as you said, optimization is something that you hear more and more now in the cloud world, along with the modernization that is happening for, you know, benefit... And as you said, everyone benefits. And you mentioned multiple times, that if you do it in the right way, everyone benefits across the board. But, you know, and you touched upon this, that there is a cloud of talent, right? It's not easy to find the right talent. So how should, you know, organizations prepare for the skills and the process for this whole cloud native world?

David Linthicum: Give yourself at least a year to ramp up. Because it's going to take you that long to figure out how to reskill your existing personnel. In other words, sending them to training or retraining them in-house. Lots of good ways to do that now with all the great on-demand training that exists. Also, who you're going to hire. And so you're going to get to the skill sets that are going to make you successful. And you're planning this not after you deploy your system and you start developing it, but before you start moving it. A lot of people, they'll come out with their budget, and they say, "Okay, we got $10 million to build this system. And let's start building it." And if they don't have the skills in-house, it's not going to be a good thing. You're gonna make a lot of mistakes that have to be redone and rethought as the higher-skill people come onto the board. So splurge on hiring the right talent, certainly a cloud architect, security architecture, cloud engineers, AI architects, AI engineers, data scientists, all these sorts of things that have modern skills that are able to make good decisions for you and understand your current requirements and map them into a solution.

And I think that people underestimate how long that's going to take. And also they underestimate how much they need to pay. So, you can find those people out there, they're not cheap. If you're building that talent in-house, make sure you're also, as you're upskilling them, you also upskill their pay, or else they'll leave. So there has to be a retention system there. You have to make your people happy. That needs to occur before you start doing magnificent things with cloud technology. I think people are hindered right now...the number one hindrance in people moving into the cloud is they can't find the people to help them. And so, they end up hiring consultants, very expensive to do that. And even the consultants may not have the talent for you, just because they are finding the same shortages in the marketplace.

And so, it's the asset that you need to secure before you start moving into the cloud. And if you don't have the skills in place, you're not going to be successful. So just figure out how to get the skills in place. But give yourself enough time and allocate enough budget to do it. People underestimate that stuff all the time. And you're going to be in for a penny in for a pound. Either you're going to do cloud correctly, hire the right skills, and pay for the skills that you need. Or you're going to try to penny-ante your way through it. And it's just not going to work and you end up wasting that money.

Very valid point. Like, you know, a strategy, like, we'll figure it along the way will not work here unless you have the right talent with you before starting.

Predictions on the Future of Cloud Computing

Prasad Rao: Cool. Now, in the book, in the latter half of the book, you also talk about the insider's perspective of the future of cloud computing. So can you give a sneak peak of what you foresee that is coming down the line for cloud computing?

David Linthicum: I've talked about this a great deal. I think that we're no longer moving to cloud computing as our end-state destination for everything. If you talk to... Probably not me, I don't think I've ever had this opinion. But if you talk to the thought leaders back in 2000 to 2011, they would tell you that this is the destination for everything. We're going to shut down our data centers, and we're going to move to these centralized processes that exist in the cloud. And we're going to use everything as a utility service. And that's where it's going to go. And you saw businesses out there that will tell you, we're cloud only. In other words, we're only moving to the cloud. That's the only platform we're going to enable. That's gone out the window. And so the future of cloud is going to be ubiquitous computing, heterogeneous computing, where we're leveraging whatever platform we need to leverage that's going to bring back the most value to the business. That's a concept.

That means a mix of modern on-premise systems and, in some instances, edge-based systems. It means we're going to leverage co-location providers as well, to provide some of this infrastructure. And that's basically data center real estate you rent. We're going to leverage managed service providers, which I think are under-leveraged, for people who basically maintain these systems for you, even systems in the cloud, and you're outsourcing a lot of that problem. Just people who know how to do it professionally. And then stuff in the cloud. And I think that we are going to continue to migrate into the cloud, certainly most of the AI stuff is going to exist in the cloud. We just saw the numbers from the cloud providers. Now they're growing like crazy because of AI.

But we're not going to move to an all-cloud world, we're not going to move to an all-on-premise world, an all-edge world. It's never going to be an extreme. It's going to be ubiquitous computing, heterogeneity, leveraging all kinds of different platforms. We have to have the connectivity between these things. Now we have 5G, ubiquitous networking that goes through ubiquitous computing. And also the ability to leverage existing assets to more of a benefit for the business. I mean, the computer I'm talking to you right now, I'm just looking at the processor saturation, it's at 5%. Okay? Well, I rent this out, you know, as available processing occurs, either in my house or even in a team of people. So, they're able to find it and use it securely. And therefore, I'm already paying for the power for this thing. So they're leveraging the power, leveraging this stuff. So it has a huge sustainability benefit. And that's an extreme example.

But the moving to mobile computing, edge computing, moving things out to any kind of a number of different platforms, and the ability to manage them on that point is where we're moving to. And I think that's going to be the end state of all this. It's never going to be moving to all clouds. I don't think it ever was going to move to all clouds. I always saw a saturation point at about 60%, 70%, where lots of applications were not economically viable in the cloud, so we need to keep them where they are. So that's the future. I mean, wish I could say we're moving to this platform and this AI. That's not the case. We're moving to a complex, heterogeneous environment that we need to figure out how to manage better than we can today.

Prasad Rao: Very insightful. And frankly, the book, end to end, that I have read is more insightful than many of the books that I have read. So, thank you so much for writing this book. It was really a great read for me. And I recommend everyone to have a look. Even if you are working with the cloud or not working with the cloud, you will find it equally insightful just to understand the trends and what you should be looking for. So, David, where can the readers or the audience find your work?

David Linthicum: Anywhere fine books are sold. Amazon's obviously the pick that everybody uses.

Prasad Rao: Awesome. And where can the audience go to know more about you?

David Linthicum: LinkedIn is probably the best place. I have a YouTube channel now, it's about 30 days old, called The Cloud Insider, and hence the book. I have about 10,000 subscribers now. So come subscribe. InfoWorld blog, I've had that blog for 12 years. Also my classes on LinkedIn Learning. There are probably 72 classes out there in all kinds of different languages. I just released a generative AI architecture course with Go Cloud Careers. And it's available in May. But you can preorder...move there now. And that's kind of a cool course because it's a long-form course. It's 40 hours long. And there's going to be office hours, where I show up once a week and help architects, and give them problems, and they're able to solve things. So it's more of a hands-on, guide you through its experience. I'm excited about that. I think that's a better way to teach architecture.

Prasad Rao: Awesome. I love the energy that you have in everything that you do. Thank you so much for joining us today.

David Linthicum: You got it. Thank you very much for having me.

Prasad Rao: Thank you.