Teaching Infrastructure from Experience with Keith Townsend, Principal & Co-Founder of The CTO Advisor

(Intro Narrator at 00:00:03) Hello, my friends. Today, Joel is talking to Keith, principal and co-founder of The CTO Advisor, and they discuss how Keith has taken on infrastructure modernization projects so he can experience the pain involved and better teach about it, why it's critical to be mindful of technical debt you are taking on, and how the infrastructure team can better support developers.

(Joel Beasley at 00:00:37) Here we go. This is the Modern CTO podcast. Finally. It's been like four years I've been checking out your stuff, and it's so cool to actually see you and have you here, man.

(Keith at 00:00:57) Yeah. You've gotten some pretty good interviews in, man. I appreciate all the work you've been doing.

(Joel Beasley at 00:01:02) I've been trying. You know? So why did you start The CTO Advisor?

(Keith at 00:01:09) Wow. That's a loaded question, Joel. You started it right off the top, right?

(Joel Beasley at 00:01:13) Yeah.

(Keith at 00:01:14) So I guess a better question is, what is The CTO Advisor? And then the why will start to make a little more sense. I like to be upfront about how I make money. The CTO Advisor and CTO is mainly via content. We originally set out to create an advisory firm. I have a project and a management consulting background, and I originally thought, you know what? I'll start a business that's an advisory firm, but I know most advisory firms don't succeed because the revenue is just too up and down. It's too inconsistent, so I'll augment it with sponsored content because I've done sponsored content before. Come to find out, the sponsored content business is way better than the advisory business. The turnaround for me to close a deal and deliver a deal on the content side—basically helping the big IT vendors of the world connect with my audience—is way faster. So I've shifted focus and said, okay, instead of focusing on advisory, which we still do some of, we mainly focus on content. And it ended up being doing something that I love, which is just—you do content. It's incredible. Just talking to people about their journeys and doing stuff and helping IT decision makers make decisions.

(Joel Beasley at 00:02:57) Yeah. No. I mean, we didn't make money for the first two years, so we were like 200 episodes in before we figured out that people pay for this stuff. And then we were like, okay. Well, so we created deliverable packages and all that type of stuff. And then we sort of hit a growth limit, right? Because you can only do so many episodes. You can only do so much content. And then we started making content for other podcasts and other companies. And so that I think is the way that we can grow much further.

(Keith at 00:03:29) That's pretty interesting. I've heard a number of people kind of get into whether it's custom podcasts for brands, et cetera. Is that mainly where it's at, the custom podcast for brands, a company says, hey, I want to do a podcast, but I have no idea how to do a podcast?

(Joel Beasley at 00:03:50) Yeah. I mean, that's like the simple version of it. Yeah. But ultimately, you know, we do everything from creating the concept or looking at what's in the marketplace. Basically, all the stuff we do for our own brands, you know. We always have to make sure the content's relevant. We always have to make sure the audience is interested in it and be checking back and forth. And so some of our sponsors have a lot of success with the podcast, and then they started saying, hey, can you make us a podcast? And so we started doing that, and that's been doing pretty well.

(Keith at 00:04:23) Yeah. So we take a little bit different approach. I love the gap you're filling. It took us a while to kind of figure out who is it that we want to talk to and how do we want that relationship to look like. So we narrowed down on specifically the enterprise IT CTOs. We did an interview with the CTO of The New York Times some time ago. I'll go and talk to him specifically about making the transition to serverless and talking about the business of IT versus your angle, which I really appreciate, which is kind of the—I love that you cover some of that, but yours is more of a superset, and you're talking about the product as well as kind of the business of IT.

(Joel Beasley at 00:05:22) Yeah. I just try to think of things that are interesting to me. You know, I use myself a lot. Like, what do I want to listen to? You know? And just to clarify, it doesn't have to be a CTO. We look at it from a content perspective. Like, as long as the content is interesting to the audience, then it's cool. But 80% plus of the episodes are CTOs.

(Keith at 00:05:48) Yeah. So I guess I should caveat that as well. A lot of times it's not CTOs—topics CTOs would care about and guests that CTOs would care about. So we had one careers episode where, and I'm sure you've experienced this, the types of content that do well on the general social media platforms, YouTube, LinkedIn, Facebook, et cetera, like, we're talking scale. The topics that scale are entry-level career topics. If you're talking about how to break into IT or how to get your first job as a network administrator or security professional coming from the support desk, those—because the audience is very big for those types of topics. What is less successful, or I don't want to say successful—what doesn't scale as well—is these mid-career topics. Like, how do you move a senior person on from one senior role to another senior role, or how do you retain them? If you've had ten or fifteen years of experience, how do you move along your career at that point? There's less of those people kind of by rule in the audience, but CTOs want to know this. CTOs, CIOs, and COOs want to know this so they can attract talent. So we will run those types of guests as well. Again, kind of your point is, if I find it interesting, chances are—it's a great thing about the internet. If I find it interesting, there's probably a tribe of people out there like me. So expanding kind of the content beyond podcast, we do papers. We do custom video. We do events, et cetera. So getting people interested in The CTO Advisor content in general has taken, and it's still taken, a lot of concerted effort to make sure I understand who my audience is and why I'm creating the content. So I try and ask myself some basic questions whenever I create content, especially content that I'm investing a bunch of resources into. For instance, we just spent $100,000 on some server gear for the data center. That's—you know, that's—I'm a small business. I only have a couple of employees, like, three employees, and a $100,000 investment is a really big—it's a huge deal. Investment for me.

(Joel Beasley at 00:08:40) Yeah. It's huge. It's a very big investment.

(Keith at 00:08:40) And we ask ourselves some basic questions, like, who are we creating this content for? Like, who's this for? Is it just for Keith to have—you know, I happen to like $100,000 servers, but I'm not so well-heeled that I'll just go out and buy a $100,000 server. But who will consume this content? Why will they come consume this content? And what will bring them back to kind of get back into the story? And then there's the balance of, you know, am I selfish with the information? One of the—I kind of tweeted out, hey, I bought this $100,000 thing. And one of my audience members mentioned, oh, I can't wait to see the write-up on it. I'm like, I had no intention of doing a write-up on this thing. It was kind of a thing we needed, and, yeah, I could make some money off of—well, it would be super interesting for my audience to hear about it, but I wouldn't necessarily make any money off of it directly. And that has been one of my most interesting balances, how to balance—how to be kind of—how to stay open and have a business model at the same time. Because, typically, what I would like to do is for the hardware manufacturer that I bought the $100,000 thing from to sponsor the content because they're getting free publicity. But at the end of the day, I'm here to serve my audience. And if I'm not serving my audience, they'll find somewhere else to go.

(Joel Beasley at 00:10:16) Yeah. Who—what type of server did you buy?

(Keith at 00:10:19) It's a HPE dHCI system. So it's a hyperconverged infrastructure system, a couple of servers, storage array, and a whole bunch of memory and disk. One of the interesting things I like to tell people about that is that one-fifth of the cost is memory. Physically, one-fifth of the cost is for DRAM. And so DRAM is expensive.

(Joel Beasley at 00:10:45) Why did you need this powerful server?

(Keith at 00:10:48) So power isn't necessarily the thing that made us buy the server. We built a data center to tell the enterprise IT journey. What we wanted to be able to do, since we're not doing proper advisory work—and even if I did do proper advisory work, it's really hard for me to take a customer's story, genericize it enough that I can share with my audience. So that's one of the other conflicts of having a content slash advisory business. So we said, okay, why don't we just go on the journey with our customer? Why don't we build a legacy data center, which is what we did in 2020? We took kind of five-year-old technology, deployed it into a proper data center with, you know, guards with guns and super redundant power, 10 gig internet access, 10 gig access to the public cloud, like a proper enterprise data center that looks like much of the environments I've supported throughout my career. And one of our first sponsorships was Intel to help us modernize that data center. So take it from that five-year-old busted environment that's, quite frankly, most people probably still have today to this new environment that's going to have 25 gig networking, 100 gig interconnects, Intel or AMD latest gen processors, a ton of DRAM, NVMe storage, just a modern data center. And so what we discovered along our walk is the same thing that most people discover is that the difficulty isn't in the bits, the speeds and feeds, and all of that. All that stuff is solved. The difficulty is taking your legacy operations, which we had because we built in technical debt from the start, and migrate from that technical debt into this new environment. And we have the same challenges that every other enterprise IT shop has. How much time do I spend on researching firmware, upgrading VMware vSphere, configuring network ports, et cetera, versus learning Kubernetes, internet control planes like—or any internet public cloud control planes like AWS or Google Cloud or OCI, et cetera. How do I balance that? I only have so many resources, but I have all the challenges of an enterprise IT shop. So we narrowed down on a solution that's going to automate some of that mundane—looking for firmware upgrades, firmware compatibilities, automating the provisioning of storage LUNs, all of that stuff that I don't want to spend money and resources doing, these solutions do for me. So we bought it for that. Not necessarily because we have some super fast application or a ton of data that we need to ingest, and we need to process that data. No. This is because we're on the same journey as our audience. We want to feel their pain and experience what they experience and tell that story.

(Joel Beasley at 00:14:18) Well, that is awesome. So you did the project before in 2020 where you built that legacy data center and then you upgraded it to understand all the pain. And then you've recently bought this new $100,000 server. And what is that—an older one? Is it a brand new one? What's your hopes of your experience with that? What lessons will that yield?

(Keith at 00:14:39) It's a brand new one. So the lessons that we've learned so far, a couple of lessons—I'm actually putting together, and it's timely because I just did a short video on Twitter about it right before we hit record for this. Nothing's ever as easy as vendors make it out to look. We're definitely going to get some benefit out of this thing. It is brand new. I bought it from CDW. I arranged financing through HPE Financial. It is the full customer experience. The only thing that we didn't do was pay for professional services to have it installed because who has time for that? I know how to turn a knob. And we've learned the same thing that I've learned repeatedly over my career that none of this stuff is as shiny as they make it out to be in the sales presentation. Fortunately, we had already deployed one of these as a sponsored project with HPE last year, so we kind of knew what we were getting into. But now we're actually integrating it into our environment, and we're running into all of the problems you would traditionally. And on top of that, we're—I'm not going to say we're pandemic world. We're new normal when it comes to working from home. I don't have anyone permanently positioned at the data center. So when there's basic things that need to be done, like, I need a NIC card moved from one machine to another machine, this solution doesn't change those basics—you know, the basics of physics—for me. I still have those challenges, and I have to weigh the value of automating the environment as much as possible versus just waiting a week until we can get somebody out there or paying the $75 an hour to have remote hands handle something to the best of their ability. So there's trade-offs. So even with this $100,000 investment, I'm not going to get to where I want to get to from an operating model, but I will get to it from a lessons learned. So when I share the experience with my audience in a month or two, they're going to get the full richness of someone who kind of knows their pain.

(Joel Beasley at 00:17:05) Yeah. And I'm curious. So do you think it's because technology is moving so fast that these companies don't have time to actually do mock scenarios like yours and make the product easier to use before releasing it to market? Or do you think that they just don't care as much or what?

(Keith at 00:17:24) Yeah. It's hard. I wish they—the answer to the question is I wish they would have more resources to do it, but I'll give you an example of something that happened in our data center about a year or so ago. We did a project for one of the public cloud companies, and we've done projects for all but one of the public cloud companies now. And a team within—you know, these companies are huge. A team within that company needed to validate that a software solution worked on their stack. That software solution extended the private data center into the public cloud, and none of these multimillion dollar, multibillion dollar companies—none of them had any environment to do that validation testing in. They would have had to stand up—they would have had to, you know, go out and sign a colo agreement, stand up the servers, et cetera. And they said, hey, Keith, since we just did this project with you, can we just use your data center to do it? And I'm like, yeah, sure. Why not? Value add, do the project. They called—one of the validation engineers called me and said, hey, is it possible for you to change your whole IP addressing scheme to be this static IP addressing scheme because the test suite assumes a certain set of IP addresses? And, of course, I'm like, no. This isn't a paid engagement, and we're not going to do that. But it shows the level of complexity when you want to do this level of software testing or integration testing. To get the product out the door, you have to make assumptions, and sometimes they make the wrong assumptions. It's happened with every HCI solution or dHCI solution that I've provisioned. It's like, oh—if you're familiar with IT infrastructure in general, VMware vSphere is the biggest solution out there. And they just make decisions like, oh, everyone's VMware vSphere environment looks the same, so the automation is baked on that. And their claims that you can deploy the solution in fifteen minutes or twenty minutes is caveated by, oh, just as long as your environment looks like this.

(Keith at 00:19:44) Otherwise, it's gonna take two weeks.

(Joel Beasley at 00:19:47) Got it. Got it. And it's hard too because there's not a real awesome answer to that. I mean, you have to build it to something, and you can't build it to everything. You would never get the product out the door if you built it to perfectly support fifteen minute deployments on every variation, right?

(Keith at 00:20:04) Yeah. And, you know, technical debt is a thing. I love the saying that IT is additive. Every time you add or buy a new solution, you're expanding the surface area of what you have to support. You rarely reduce it.

(Keith at 00:20:28) I had this argument when Nutanix, one of the biggest or the biggest HCI vendor, came out, and they had this argument that HCI saves money. And I fought back a little bit and said, you know what? If you're a small shop and you're replacing all of your storage arrays and servers with Nutanix and HCI, that is true. You will save money. But if you're a big enterprise IT shop, you know, Walmart—let's pick on Walmart because they're publicly a Nutanix customer—Walmart is also publicly an HPE customer. They're also publicly a Dell customer, and they have all these different solutions. And the HCI thing that was supposed to save you all this money is just another thing, and it's another thing that you have to support and you have to add to your overall IT cost. So when a new solution comes out to market, whether it's cloud, serverless, Kubernetes, it really doesn't matter. It's all additive. I like to go back and talk to the Kubernetes or cloud folks or the serverless folks, and I say, you know what? As we zoom out, let's talk to your enterprise architect about how they're still supporting OpenStack or how they're still supporting services they built via AWS ten years ago and how they're thinking about bringing that stuff back on-prem because it's no longer being invested in. And there's just technical debt, and they need a way to control the cost of it versus continuing to innovate on the platform. So that is the journey that we're talking about. How are we making these decisions and these trade-offs when someone says, "Hey, Keith, you're still on vSphere. Why don't you move to Kubernetes?" It's not that simple. It's not that simple for the real world, and it's not that simple for my world, which is now the real world.

(Joel Beasley at 00:22:31) Yeah. No, you have a deep background in infrastructure. So I mean, I can just tell because my background that I go really deep in is software development, like best practices with writing scalable software or test-driven development, things like that. So I went like a mile deep there, and I only learned enough infrastructure stuff to stand up MVPs, get money behind it, and then hire an infrastructure person that's smart, that knows how to do it, to take it beyond that. So it's like stand it up, get the first customer, and then—so I don't have really any experience. So as you're talking about all this stuff, I was trying to really wrap my mind around, you know, how—you keep saying you have this data center, and I'm like, what do you do with it? Is this like a mock environment, or are you actually legitimately using it for some project?

(Keith at 00:23:25) Yeah. So I love the fact that you have software chops, and this will help kind of paint the picture and comparative. I can go a mile deep in infrastructure and ankle deep in software development, so we'll complete each other's sentences.

(Joel Beasley at 00:23:41) You complete me. Yeah.

(Keith at 00:23:45) So let's say ten years ago you built a monolithic application on AWS's control plane. You're using different services from the control plane to build the service, but essentially it's a monolith. And five years later, the application's mature. There is no new feature. Feature requests have slowed down. It's time to move this application to steady state. You have your new project that you need to do. You know, there's the new hotness, whatever the business process or whatever that you want to write software for. Now you're using modern software practices. You've probably adopted Kubernetes, CI/CD pipelines, et cetera. And then someone taps you on the shoulder and says, "Hey, Joel, go back and update that now ten-year-old app." Your response is more than likely gonna be what?

(Joel Beasley at 00:24:50) I don't wanna do it.

(Keith at 00:24:52) Try someone else.

(Joel Beasley at 00:24:54) I don't remember how. Yeah.

(Keith at 00:24:57) Yeah. I don't remember how. I don't wanna do it. Where's the business value in that? If it's a ten-year-old app, what's the business value in bringing a ten-year-old app to Kubernetes? If I'm not actively developing it anymore, where's the business value? So that is a reality that as an infrastructure guy I have accepted. It is my job to complement you as the developer and make you go faster. So that ten-year-old app that you no longer want to maintain is for me to figure out how to keep that running and keep the infrastructure moving as you're solving problems. So all of that legacy data center that I've built, I put mock apps on that are ten years old. I put some new stuff on there. And then I help tease out and say, when should I put the new stuff on new stuff? And when does the new stuff become old enough that it needs to go on the old stuff? Because the old stuff isn't going away. It's just piling, piling, piling, piling, and it's not reasonable. And it took me a long time after talking to folks like you to accept this. It's not reasonable to ask me to come to you and ask you to keep the old stuff modern. That's just not reasonable.

(Joel Beasley at 00:26:12) Yeah. It's like because the business value isn't there, and then how are you gonna pay people paychecks to do it when no one's paying you to modernize that? I was so blown away when I got into this whole talking to other people, other technology leaders across all these industries, to find out that there are still huge portions of mission-critical systems for us as a people in The United States that still run on these old mainframes that basically nobody really knows how to run, and they're just, like, fingers-crossed type deal.

(Keith at 00:26:48) Yeah. I went into the manufacturing floor of a major pharmaceutical company, and they make a very important medicine. And the way the FDA works, when you formulate a manufacturing process on a computer system, that computer system becomes validated. You can't change it out. Because it was validated, because the manufacturing process was validated on a system in a snapshot in time, twenty years later you still need to—the manufacturing process still needs to operate on that validated computer system. So in this very modern pharmaceutical company—every other sensor, you may not be familiar with the name, but I'll give some context—there's a DEC Alpha system on the floor. HPE bought DEC Alpha back in 1999. So there hasn't been a net new DEC system since maybe 2002, 2003 at the latest. HP—well, actually, HP bought them, then HP became HPE. But to the point, that system still needs to be maintained because of regulatory reasons. I visited during our CTO Advisor road trip, which we hadn't talked about at all, last year where, you know, me and my wife toured the country and talked to CTOs and kinda brought the discussion to—yeah, it was a really fun trip. We're planning on doing it again this year, and we brought the discussion to the various regions of the country. We stopped off in Boston and stopped at HPE Financial Services' reclamation system where they buy old equipment from customers and then refurbish it and send it out. They had dozens of DEC Alpha systems. It's a huge part of their system. It's a not insignificant part of their revenue to lease back these ancient systems and support these ancient systems because they're still mission-critical in a lot of environments.

(Joel Beasley at 00:29:07) That is crazy. Where did you go? So you went Boston. Where else did you go?

(Keith at 00:29:13) So we started—we're in Chicago or in Chicagoland. So we started out Chicago. We headed out Northeast, so we hit—I'll do the quick tour—Cleveland, Philly, Boston, New York, DC. We took a week off in Tampa to visit family. Atlanta, Houston, then Austin, LA, San Fran, Seattle, Portland, Colorado, and then home. So ninety-six days in total from door to door, out the door to back home.

(Joel Beasley at 00:29:57) And you ran interviews while you were remote, or you just relationship building and hung out with people?

(Keith at 00:30:01) Both interviews and relationship building. One of our favorite conversations is in Akron, Ohio. We talked to a small business—they're a manufacturer. They do signage, like metal signage. Anything that has to do with machines—they curate machines to do signage, so anything that has to do with stamping or signage, they do. And I was just fascinated. I asked them, you know, I would've thought that all of this would've been outsourced to China by now. And it was really, really interesting as a business owner talking to another business owner about how they're competing with China, keeping up with their customer demands, reacting to their customer demands, how they had this 40-year-old piece of equipment right next to a piece of equipment they just paid $300,000 for and how both was generating revenue for them, and also adjusting to the pandemic. Probably my most enjoyable conversation—maybe not the most geeky, but definitely the most enjoyable conversation—is we just went across the nation talking to people doing stuff during the height of the pandemic.

(Joel Beasley at 00:31:21) And some of those episodes are on your podcast?

(Keith at 00:31:24) So we didn't release any as podcast, but they are on the YouTube channel. There's a whole playlist. So, you know, if you visit youtube.com/ctoadvisor, you'll find all of that wonderful content. We're looking to see how to more formally do it this year as we look for sponsors and events. What we wanna do is—at the end of this trip, we want—and there's a pretty cool theme to it. It's kinda like "Where's the Cloud?" So we're gonna set out and we will talk to customers and IT vendors and other analysts and ask them, "Hey, how do you get to the cloud?" Then at the end, whatever VMworld is gonna be called this year, whether it's VMworld or some other conference, we'll end up pulling our Airstream trailer onto the showroom floor and say, "Hey, we've made it to the cloud," and have a conversation around our journey to the cloud.

(Joel Beasley at 00:32:22) Oh, nice. Yeah. That's pretty cool.

(Keith at 00:32:25) That's the vision. We'll see if our sponsors pony up the bill to actually make that happen, but that's the vision.

(Joel Beasley at 00:32:33) Yeah. We actually, like, had sold everything, I don't know, like ten months or so ago. My wife and I, we were in Florida, so I'm a native to the Bradenton, Tampa area. So that's cool that you have family down there. And we sold everything, bought a fifth wheel, you know, a camper, and spent seven months just traveling around The United States. I went and got to, like, visit two different people.

(Keith at 00:32:58) Yeah. That is so crazy that we've done similar things. Seven months is probably a little bit more practical time period, and the fifth wheel may be a little bit more comfortable than my Airstream.

(Joel Beasley at 00:33:11) Oh, yeah.

(Keith at 00:33:11) I do really like the Airstream, but it is tight. And living in it for three months while constantly moving with something—that's hard. It's well-appointed, but that was really interesting. And constantly moving with something—the not having the address thing, though, that takes a little bit of getting used to, I would imagine.

(Joel Beasley at 00:33:32) Oh, yeah. Well, there's a thing called MyRVMail, and they give you, like, a legit address in the state that you want. And so we just have all the mail there. We see it come in through the app, and then we just tap on the things that we want them to open up, and they open and scan them. And then if I click a button, they'll actually mail it to whatever address I want them to mail it to.

(Keith at 00:33:51) That's cool. What was the biggest kind of shock or lesson learned from RVing for seven months?

(Joel Beasley at 00:33:59) Biggest shock was that you're technically considered homeless. I was filling out some paperwork and stuff for the kids or whatnot, and I was like, "What?" I was like, "Oh, okay. Fine." There was that. But, honestly—so I have two kids as well, a third on the way. But we just did this with two kids, a four- and a three-year-old. So we started out with a bumper pull, smaller trailer. And then we quickly realized, oh, you have to spend more money. Yeah. And then at first we were driving around. We did one big circle around The US first. And then we realized that the concept of, like, driving two or three days, stopping for two days, driving—the constant moving is, like, untenable, especially with children. So we decided, like, let's go stay in this park for a month, then let's go stay in this park for a month. And then that was way more doable. And then it got cold. And so we're sitting there freezing in the trailer, like, three, four months ago, just going through Zillow. Like, maybe we need a house again. We're freezing. And we found this little farm in Tennessee, and so we bought that. And now we've been here for two months.

(Keith at 00:35:10) Ain't it really cool to be able to do our jobs in an RV? Well, one, the thing that I learned about doing our jobs in RV is internet is really hard. That's a really hard problem. How did you solve your internet connectivity issues?

(Joel Beasley at 00:35:27) I rented hotel rooms when I had to do shows. And I had another host, Adam. So, originally, like, it was gonna be a summer thing. And so we did a summer trip, like a six-week trip. And then we're just, like, at the end of that six-week trip, we went home and sold everything in, like, one week and then went full-time. So in that summer takeover we did with Adam, he was, like, a great host. So I was, like, alright, well, I'm not the only one who has to host it. So he would host some. And then when I could get to hotels and rent rooms, especially for the sponsored ones, that was really easy, right? Then the hotels had really stable internet overall. So the entire time I was like, "Hurry up, Elon. What are you doing, bro? Like, get me one of those satellites quick, man."

(Keith at 00:36:15) Yeah. I finally got approved for the—well, I had ordered it over a year ago, well over a year ago. And beginning of March I finally got my notice, "Your thing is gonna ship. Just confirm it." And then at that point I'm like, I don't know. I have friends that live off of it. And if it works for RVing, it is a godsend. But it is nothing like having a permanent hardwired connection that is not full of latency. Because if you're interviewing folks, the latency that it introduces is quite substantial.

(Joel Beasley at 00:36:55) Oh, yeah. The best conversations are when you can have it in person, but that's definitely gotten less popular in the past couple years. So we've just been doing the best to, like, you know, build an interesting set and increase our mic and video quality and do everything we can do within our control to make it the best experience possible.

(Keith at 00:37:15) So what part of Tennessee are you in?

(Joel Beasley at 00:37:17) So I'm an hour outside of Nashville in a town called Clarksville.

(Keith at 00:37:22) So we were just in Nashville the weekend before last. So we could have done—we could have done this live in the studio. Oh, right? Yeah. We spent the week—we were, for a week, we were—or actually, two whole weeks we were homeless. Between closing on houses, we said, you know what? Why stay at a hotel? Let's just jump in the Airstream and hit the road again.

(Keith at 00:37:49) So we started, you know, it's cold in Chicago, and we just started traveling south. And we stopped in Nashville, and we're like, yeah, this is good. And maybe about fifteen minutes away from Opry.

(Keith at 00:38:00) And we were like, okay. We spent the week in Nashville, and not anything on the calendar. We had pretty decent LTE access, and we hung out for a week. So we could have done this in person.

(Joel Beasley at 00:38:15) Well, next time you're down in Nashville, let me know and come by the studio. I built the studio on my property as a separate building, so that way now I just have to walk 25 yards to my work. And there's gigabit internet out here, man. How crazy is that?

(Keith at 00:38:34) Friend, Chris Collotti, who's in Tennessee. I'm not sure which part of Tennessee he's in, but he also has gigabit internet up and down. I'm like, son of a... I have gigabit down, but I would kill for gigabit up. But again, for what we do, gigabit up would be much more preferable than gigabit down.

(Joel Beasley at 00:38:51) Yeah. And one of the reasons I found is there's a bunch of government programs that actually would pay companies. Like, they could go get money to install this stuff, and the government would help pay, subsidize it for them. So honestly, I'm not a huge... I like small limited government. But when I saw that program, I was like, yes.

(Joel Beasley at 00:39:11) Because what that does is it makes all these places more accessible and you get a better quality of life because you don't have to live in the city for amazing stuff. And it's overall, it's pretty... and Musk even got in on that a little bit too because they qualified under the Starlink system as well for it.

(Keith at 00:39:30) Yeah. I've been following the Starlink thing for a while. As you think about rural Native American reservations, places that just won't get the investment otherwise, and having this type of access... The new Americas field CTO for VMware, I had her on the podcast a few weeks ago. Amanda Blevins, she lives in a very rural area and it is literally her lifeline. Another guy, Joe Onasick, who is on the border of Wyoming and Montana, you know, not exactly third world country at all. He's living off of... we just did a podcast with him last week and he's living off of Starlink.

(Keith at 00:40:19) So I'm a huge, huge fan of getting as much bandwidth as you can to areas that are low in opportunity because I feel like it's just like when I was a kid and I got my first PC, and it was just life changing. I think getting folks high speed internet is one of the best investments the public can make because, again, you know, as you think about areas not too far from you like Appalachia, parts of Alabama, Mississippi, rural Tennessee where there's not the infrastructure for these types of tech jobs, and now you can send somebody to a boot camp and they can actually get a six figure job and not leave their town. How does that transform their town when there's... you know, how does it transform the town when there's only 10 or 15... I say this lightly, but I don't mean it as a slight. I mean, when you add 10 or 15 people to a small town that make six figures, that changes the town.

(Joel Beasley at 00:41:24) Absolutely. It changes the town's purchasing power considerably. And these small towns, they're like small towns too. And it's... I think of it like building the highway system. Like, once you put these internet systems in, once you run the fiber, right, that's like building the road.

(Joel Beasley at 00:41:43) Now people can come out here, now they can drive out here, now they can come live out here and work out here. So I was really excited about that. And also we got the latest technology. Right? So, like, if you're downtown in the city of Chicago, you might have older technology than me out in the sticks in Tennessee because they just purchased it.

(Keith at 00:42:01) That is definitely, definitely the case. You'd be surprised how many buildings are still DSL only.

(Joel Beasley at 00:42:09) Yeah. I haven't heard that in a long time.

(Keith at 00:42:12) Yeah. There are... I did a project maybe two years ago. I reviewed a building getting upgraded to fiber so that they can get fiber from the phone company. It wasn't... it's an old building. It was a mess expense, and they punted and said they'll just do DSL. Yeah. The infrastructure isn't there for... they do DSL and DIRECTV for television because they don't have the infrastructure to run coax.

(Joel Beasley at 00:42:45) Wow.

(Keith at 00:42:46) And this is downtown Chicago.

(Joel Beasley at 00:42:48) I'm not surprised. The stuff I've heard, dude, I'm never surprised. All I... that's why I just constantly refocus on the future. Where are we going? I don't want to look too much over there.

(Joel Beasley at 00:43:01) But I know we're coming up on time. I saw on my producer's note that there was a hard stop here at the top of the hour. So I don't want you to be late to your next meeting, but I also want to make sure, because we are still recording, what's a great episode somebody can listen to on your podcast that you want to send some of our listeners over to check out?

(Keith at 00:43:18) You know what? The one that a lot of people really enjoyed is a conversation I had with Kelsey Hightower about a year, year and a half ago. Kelsey works for Google Cloud. If you're in the cloud native space at all, you're familiar with him. It was kind of this... Kelsey has an infrastructure background as well, but he's also a developer.

(Keith at 00:43:38) So it was this old school enterprise IT guy, me, talking to this Kubernetes guy, and it was just magical. Maybe one of my most downloaded podcasts because we appeal so much to this crossover audience of developer and infrastructure guy in helping make an argument for Kubernetes and when to kind of not do Kubernetes.

(Joel Beasley at 00:44:04) Yeah. All right. Well, we'll check that out. We'll put a link in the show notes so that people can go check out that episode as well when they're listening. And, man...

(Keith at 00:44:13) And so, Joel, we'll do the same thing because we're cross promoting platforms. What's the must listen episode for you?

(Joel Beasley at 00:44:19) So I would say it's the one I've gotten the most feedback lately from. This one I did with John Lennox. And he is, like, super smart Oxford type individual that studies AI, and he's probably 77 or 78 or something like that. He's much older. But he, like, studied under CS Lewis and stuff. And he talked about the intersection of AI ethics and Christianity.

(Joel Beasley at 00:44:53) And I was... and he just brought up some really different views. It was really cool to talk to a very intelligent, pragmatic scientist type person and then have him discuss ethics and AI and how that intersects with, you know, Christianity and religion. And it was a really cool conversation that I still get a bunch of messages about.

(Keith at 00:45:16) Oh, that is super appealing to me. I've got to check it out.

(Joel Beasley at 00:45:20) Yeah. Yeah. So, man, we made a podcast. How do you feel?

(Keith at 00:45:24) I feel great. It was about time. We've been trying to do this for a while and we finally did it.

(Joel Beasley at 00:45:29) Thank you so much for listening. And if you found this episode useful, please share it with a friend or a colleague who you think would get value from it. And if you have topics that you'd like to hear discussed on the podcast, either add me on LinkedIn or send me an email, [email protected]. Every time I get an email or LinkedIn message, it absolutely makes my day and inspires me to keep going.