Reindustrialization and the AI Network Revolution

[00:00:00] Bill Long:  Way people are designing data centers today and they know the bandwidth needed, they say, okay. This data center is 100 megawatts. And you think about the bandwidth needed at a 100 megawatts, and when you think about the data center, the amount of power is a fixed variable. AI models and AI hardware are getting exponentially more efficient. So the bandwidth that you need today for a 100 megawatt data center is probably one one thousand of the amount of bandwidth you’re gonna need ten years from now. The planning horizon needs to be not percentages, two, three orders of magnitude different for the same amount of power. So the planning horizon as you’re thinking about AI is just a very different kind of model.

[00:00:35] Scott Kinka:  Technology is changing faster than ever, but the most important conversations aren’t about the tools. They’re about the ideas and the people shaping what comes next. On the Bridgecast, I sit down with the leaders driving change from AI and cybersecurity to the infrastructure powering modern business. Together, we explore how technology moves business forward, what they’re seeing, what they’re building, and what it means for the rest of us. I’m Scott Kinka, and this is the Bridgecast, Presented by Bridgepointe Technologies. Hi, and welcome to the Bridgecast. Today, I’m joined by Bill Long, who’s the chief product and strategy officer at Zayo, one of the leading providers of bandwidth infrastructure and network services globally where he leads product strategy and road map. We’ll talk quite a bit about that today. Um, and he sits at the center of a conversation that every technology leader needs to be having right now, but few are prioritizing. Whether the network underneath all of their AI ambitions is actually built to support them. And that’s what we’re gonna chat about today. Bill, welcome to the Bridgecast. Great.

[00:01:49] Bill Long:  Thanks for having me, Scott.

[00:01:50] Scott Kinka:  Well, let’s just jump right in then. Um, you carry it’s funny. I’ve been in your shoes, uh, in in multiple roles. You carry both a product and strategy title at Zayo. Most organizations or many separate those functions and treat them as different disciplines. What’s the strategic advantage in holding both?

[00:02:09] Bill Long:  That’s a good question. I think it’s, you know, it’s a little bit of, like, you get to come up with what the recipe for the dog food should be, and then you have to eat your own dog food. So I think there’s you know, with the strategy role, you have to be looking around the corner, see what’s coming, and you’re like, oh, there’s a big opportunity or big risk we should be thinking about. But in the product role, you have to actually go and do the things that you said you were gonna do, um, whether that’s building a new product or mitigating a risk. So you get to look around the corner and be figure out what’s coming. You get to build the solution to take advantage of that opportunity. And then if you’re right, great. But if you’re wrong or you you do fall down on execution, there’s no one to blame on but yourself. So it’s you know, as a product guy not owning strategy, you could always be like, well, we’re executing the wrong strategy. There’s a strategy guy who didn’t know an execution, you you can always say, oh, they got the execution wrong. Having both of them, um, you know, you kinda get to look around the corner and you get no one to blame. You get all the credit and and you get all the blame. So it’s, uh, it’s good to have both.

[00:03:04] Scott Kinka:  I can empathize with that, having been in that role. Um, but, you know, you did say something that was interesting in product strategies about making bets on the future, um, and I’m sure we’re gonna talk quite a bit about that today. Um, what’s one trend that you’re investing in today that you believe maybe enterprise leaders still underestimate?

[00:03:24] Bill Long:  Yeah. It’s a good question. I mean, I think obviously the the AI trend, um, I still think, um, we’re at the very front end of what that what that really looks like. Um, and everyone’s kind of focused on that one. The really interesting one that we’re seeing that I don’t think the market might not be paying as much attention to is really the reindustrialization of The United States. So a lot of the the workloads, whether it was chip fabs or factories or whatever, is coming back to The US, but it it’s being solved in a very unique way. So when a factory repatriates The United States that might have been in China or somewhere else, instead of it being rows and rows of people, you have tons of automation. So you kinda have these converging and compounding use cases with its AI and the reindustrialization, and they’re actually accelerating each other and they’re compounding each other. You’re applying lots of automation, and AI makes the automation a lot easier in the universe of what you can apply that automation to, um, becomes bigger. But what you’re from a networking standpoint, what that looks like, a great example is, um, we have a a a major retailer we’re working with who’s automating all of their their warehouses. So in the past, they used to have just, you know, 10 meg Internet connections into their warehouses. Now they’re putting, you know, a $100,000,000 plus of automation robots into each one of their fulfillment centers, and they wanna have a 100 gig redundant fiber connections into those warehouses because it’s robotics. You gotta be connected to the robots. The robots are sending data back all the time back into the cloud into AI. So it’s that real industrialization, I think, is not getting as much, and the tech impacts of that are not getting as much, you know, attention in the market as I think as they should be.

[00:05:03] Scott Kinka:  Wow. I would agree with that. I was not expecting that to be your answer, and it it was immediately clear and and made sense. I mean, you’re not putting factories in metroplexes. Why you can’t afford it. Right? So largely speaking, you’ve got the compounding effect of you wanna automate all this stuff. You wanna build AI. You also want a factory in a place where you can afford to put a factory, And that is not, generally speaking, where we have significant network infrastructure built out.

[00:05:29] Bill Long:  Yeah. And I think you you have that trend along with with AI as well. Like, the big growth that we’re seeing are in locations where networks normally haven’t been. It’s not in the in the Central Business District with big shiny buildings. It’s in industrial parks that are on the outside of town. And even these data centers that are in the middle of nowhere like West Texas, who would have thought that you need terabits of connectivity in West Texas or in the Dakota? So where demand is showing up is very different than historically it’s been, which is a a big challenge but big opportunity for us too.

[00:05:57] Scott Kinka:  I also think maybe one of the reasons why it’s not getting as much play is because, you know, reindustrialization takes time. You know what I mean? So it’s certainly been the, you know, uh, one of the political storms slash stories of the last couple of years, but we the effects of that haven’t necessarily compounded themselves into the economy yet. Some people are like, oh, look at all look at everything that’s been brought back to America. It’s not you know, maybe we’re in the construction phase to your point on a lot of that.

[00:06:24] Bill Long:  We see a ton of it. Like, I think it’s just happening quietly. Because in in a lot of ways, when a factory Patriots and it used to be a thousand people in China, and then it’s 20 people in The United States, it’s only 20 people. But it’s still a factor or fulfillment center that has tons of automation, tons of bandwidth needs, interesting kind of use cases that you wouldn’t you wouldn’t think about. And and and again, I think that it is certainly a good example here is we’re working with a, uh, a fuel station, a convenience store operator. And the convenience store operator now wants, you know, redundant gig connections into each of their into each of their filling stations because they’re automating all the checkout. They’re automating, um, the the gas. They’re automating so much. So it’s it’s really permeating throughout throughout a lot of stuff. We are at the front end, but it’s it’s permeating across a lot of different sort of verticals in the enterprise space.

[00:07:12] Scott Kinka:  So that’s an amazing start. Um, before I get into I wanted to ask a little bit about Zayo just to kinda ground everybody in that story, but let’s do you first. How did you get here? You know, why’d you get in tech? Just give us the quick and dirty, you know, from getting started to Lenny in your current role.

[00:07:29] Bill Long:  So classic. Like all good tech people, I was a philosophy major in college.

[00:07:32] Scott Kinka:  Yeah. Atta guy.

[00:07:34] Bill Long:  Right. And so I moved to Colorado and it was you know, because I love Colorado, and it was just fell into fell into tech. Um, and then just have did ten years, um, working at level three building, uh, you know, nationwide fiber networks, uh, with them, then spent ten years running product at a data center company called Equinix. And then I’ve been at Zayo for four years, um, sort of, um, helping with the you know, sort of taking those those two together. Zayo, from a from a asset standpoint, we do fiber connectivity. Um, you have long haul and metro network fiber connectivity throughout The United States. I’m actually out throughout all North America. So, yeah, just kind of securing this route, and then I don’t think anyone, you know, comes out of college. It’s like, I’m gonna I’m gonna be, uh, I’m gonna work in networking or in infrastructure. That’s been my path, and I’ve been in products and strategy kind of the for that whole path.

[00:08:20] Scott Kinka:  Jump into what’s going on for a moment, but I wanted to zero in on Zayo for a moment because, you know, a lot of people do network. And you mentioned this kind of congruence of your history in data center and your history in network and kind of the role that you’re filling now, sort of where Zayo was focusing their energy on this AI economy and really kind of, you know, modular, um, and networking as a service, you know, on large networks. Can you just explain a little bit about that? Um, we’ll get into why it’s important when we kinda get into philosophically what’s happening to enterprise networking, but just talk a little bit about Zayo for a moment and sort of where your initiatives are there.

[00:08:54] Bill Long:  I think everybody puts network providers in the single bucket because Zayo is very much more a digital infrastructure provider than a standard network service provider. So in the past, your network service providers came really from a history of providing voice and data services and really the the layer four through seven kind of higher level services and and and the underlying infrastructure was just the substrate upon which they delivered those services. Zayo’s coming the complete opposite direction. Our core competency and our sort of key asset is the actual physical network, the fiber in the ground. You know, we operate and build, you know, hundreds of thousands of of route miles, tens of thousands of buildings on that, you know, 2,000 data centers on that. But, really, our core competency is building and operating the fiber at the base level. We think that is the the most different that’s sort of what we’re good at. Um, and and we think that, like, you know, smart people buy dumb pipes. I think that for a lot for the longest time, um, you’ve had kind of network service providers who wanted to monetize the content and mon you know, go up the stack. We think that really focusing at the infrastructure layer, um, it solves meaningful problems for our customers, allows people to innovate on top of that that base infrastructure, um, in a in a way that provides a different outcome and more flexible outcome for our customers while allowing us to sort of make good money, um, doing it while we can. So it’s it’s a very different approach. And if if you if you kind of think about it from, you know I, you know, spend a lot of time in the data center business, it’s very you know, data centers are infrastructure as well. So in this in the same way that you have a data center and then you can choose what kind of servers you wanna put in that, we we view the network in the same way where it’s a digital infrastructure asset that if you wanna run voice over, if you wanna run AI, if you wanna run cloud things, wireless, you know, do what you want over the physical asset. And that permeates all of our DNA. It you know, we come at it from a physical infrastructure standpoint. Now now we do believe strongly that we wanna make infrastructure available at software speed, and we think that, like, you should expect more from your infrastructure. But, fundamentally, our job is to get the right infrastructure at the right place at the right time and remove friction to make that as easy as possible to do.

[00:11:03] Scott Kinka:  Does that intrinsically make your target customer upper mid market and enterprise to your point around smart people buy dumb pipes and then, you know, maybe customers who are listening right now who are in the mid market are like, they’re using Zayo. They just don’t know it because there’s a, you know, there’s a middle tier provider who’s actually providing layer four through seven services over top of your fiber.

[00:11:22] Bill Long:  That’s absolutely right. We do a lot of the four through seven services as well. We do we have a managed services offer, but a lot of those higher level services we we do is just to make it easier for people to consume the base infrastructure. So to your point, you know, 40% of our revenue is through carriers. So we do you know, we connect enterprises that a lot of other network providers where we could be the underlying fiber provider, but we we do it, um, you know, mid market’s a big part of it for for us as well. Um, but it but but you’re right that we our our core focus is is really getting the fiber and the physical infrastructure where it needs to be.

[00:11:57] Scott Kinka:  Let’s turn our attention to to the part of your job that, um, like you said, where you’re placing the bets. You’re looking over the horizon. And I think that’s probably what the listener is thinking right now is, alright. Um, you know, Bill, what’s gonna happen, right, at the end of the day? So let’s just get there. I mean, there’s there’s, um, everybody’s talking about AI investments. You can’t, you know, turn on Squawk Box or or, you know, CNBC and not talk about how it’s being invested in and what’s happening. And, of course, every board meeting or c level is going, go get some. It feels like cloud all over again. Right? We’re GPUs and large language models and copilots, but there’s not a lot of attention being paid to the network that connects at all. Um, in our preshow, we were talking about the stat that, you know, this is the this is the year I think that we’re expected to have AI based traffic supersede human traffic on the Internet. Why do you think that the networking remains kind of the most overlooked portion of this AI conversation?

[00:12:57] Bill Long:  You mentioned that people are like, hey. This looks a lot like the early days of cloud. In a lot of ways, it is. And if you remember in the early days of cloud, when you were trying to connect to cloud, you have a you you build something in AWS, then you use the public Internet to get to it. That that was fine for a little bit, um, but there became real problems that it doesn’t scale. The cost for getting traffic into and out of the cloud has has cost to it. Having a sort of broad IP, you know, publicly addressable Internet space, um, is problematic. And performance, there are issues. So if you take each of those items, whether it’s the volume of traffic, the security profile of the traffic, or the performance of the use case, each one of those vectors is on steroids when it comes to AI. The datasets that you’re passing are the largest datasets that you have because you’re training on your full corpus of whatever data you have. Good example there is we had a company that, like, takes satellite images of the whole world multiple times a day, and so they wanted to have AI train on that. How do you get the data to where it needs to be trained? So the scale of the data is so huge. It’s the most valuable data that you have. If it’s valuable enough for you to be training on it, it is your you the crown jewels of your business, and so it has a much, um, more you know, the security risk of that, you do not want that to be addressable on the public Internet, and it’s a lot more performant. If you have an agent, um, who’s working with a customer, you can’t have delays in that. So it takes the same trends that were there for cloud that puts them on steroids. And, really, if you look at what triggered the the really big takeoff of cloud, it was private connectivity. So you you had you had your developers who are using it in a sandbox environment and were doing it, but, really, once you got the whether it was AWS Direct Connect or or once you got your VPCs, you could do private connectivity and you could be connecting an ecosystem of it. That’s when it really took off. The big financials, the big, uh, retail, once you have that private connectivity, it really started to take off. And at the very beginning of that for for AI. We can talk more about what that means.

[00:14:56] Scott Kinka:  You’ve said traditional enterprise networks were designed for a different era. I love that thought process. You know, AI is forcing organizations to rethink networking from the ground up. I thought a little bit about that statement that you made and really thought about it from COVID forward. Right? Like, COVID was the first time we had to fundamentally think about the network. But in reality, what happened is our networking changed from being office centric to people centric, you know, big pipe centric to more, like, you know, authenticated access centric. And, you know, and now it feels like we’re almost going back the other direction. We’re replacing the headquarters with the model. Like, I gotta get my data to the model instead of my people to the office, but it’s a little bit, like, of a structural return to the way we almost previously operated except that the office is is where you know, who’s crunching my data. Does that analogy fit kind of with your statement around networks being designed for a different era?

[00:15:51] Bill Long:  The big thing that’s different though is that when your data center was at your headquarters location, you could put a mode around it. It was like you had your your data center and everything sat in your basement. And now with both cloud and AI, we’re moving to that distributed hybrid multi cloud environment, especially multi cloud with with AI, you know. Um, so so what that means is you have instead of servers sitting in your basement that you can put a security mode around and you can have people connect to that, you’ll have servers in multi multiple data centers in multiple clouds, and then you’ll have the data that’s available that needs to privately get to wherever the GPUs are that are gonna that are gonna be processing it. And so that how do you get the networking to work when you have a state sprawl? You have stuff all over the place. So if you have stuff in multiple data centers and multiple clouds and multiple locations, you know, creating a bunch of different modes and the problem, uh, compounds on it. And so I think one of the things that we’re working on, I mentioned earlier about you gotta make the infrastructure easy to use, the level you can no longer put infrastructure everywhere and manually configure it. The amount of automation you have to have goes up. You know, the the architecture means that you’re gonna have infrastructure all over the place. That’s a given. But But in order to make that manageable, you the level of automation you have to have has to go up. Historically, network providers have been terrible at providing you APIs, tools, automation to make the job of of both provisioning and managing and monitoring that network easy. And that’s where we’ve that’s where we’ve fallen down, and that that’s where we’re focused on making that architecture easy to to build and operate.

[00:17:27] Scott Kinka:  You are listening to the Bridgecast, proudly sponsored by Bridgepointe Technologies. When it comes to customer experience, Bridgepointe delivers innovative solutions that empower your business to stay ahead. Visit bridgepointetechnologies.com to learn how we can transform your customer experience today. Is that the Dynamic Link product that you’re talking about?

[00:17:47] Bill Long:  It’s Dynamic Link, and and that that’s the first of basically overhauling our whole portfolio when you’re waves on demand is the next product that we’re launching where it’s a the same level of automation that we did with layer two, layer three services. So what what Dynamic Link is is it’s our NAS offer that anywhere you the Zayo network connects, you can you just, you know, order a port. You’re ordering that port. We just put a mid there. There’s not, like, some big heavy box that puts there. Just put a mid. And then we have our edge we sort of a near edge where we put a dynamic link server that can do everything you want. You can set that up to connect to the Internet. You can do route filtering. You can do load balancing. You can do whatever you need the network to do. You can configure it to do it, and it’s all available with APIs and even MCP server. And we built Claude’s skills that anything in a CCIE can do, you can now ask Claude and it can interface into our network. So if you’re like, hey. I’ve got five offices. I wanna create a a VPN layer two network between those five offices, and I wanna create a connection out to AWS. You can just tell Claude that, um, and it could pull your inventory from us, and it could configure that in real time in a matter of, like, minutes.

[00:18:55] Scott Kinka:  And for people that are maybe not familiar as much with NAS and thinking as network as sort of static connection from a to, you know, from point a to to location b, the way I would kinda conceptualize it for them is, you know, think cloud. Like, used to have a server, and then there was Germane infrastructure that was already there, and you just stood up an instance. It’s kinda the same idea. This is your Germane network that’s connecting metros and all the fiber and things that you have in the ground right now, and you’re just building, you know, logical network connections over preexisting infrastructure. So in their world, they’ve gotta get onto your network last mile, but then once they’re on, it’s highly and completely configurable from there.

[00:19:31] Bill Long:  That’s right. And we can even do it on type two. So they can we can offer this NAS because it’s just a NED. You we can use someone else’s network. It even works on Starlink, where you can say you can say as as long as you get into the Zayo network, um, you know, as soon as you can hit our network, we can apply this this logic. So it works on on type two and even and those types of services. But I I love your description. Right? It’s similar to when servers virtualize. You know? You just have to go in you’d have a physical server, and then when you move it up into the cloud, it was just a virtual server. But everything the server used to do, you can still do. But you can just go into the AWS portal and configure it and set it up and tear it down and put an operating system on it. It’s that’s the same way with with NAS. Um, but instead of having to, like, order a physical DIA service with a single port and then a cloud connect port and then a layer two or layer three IP VPN, you can just order a port, and then you can stripe it to be whatever you want it to be.

[00:20:23] Scott Kinka:  Truth be told, for those of us who are in this space and, you know, people that work with Bridgepointe know that it’s you know, a lot of what we do is optimize network infrastructure to find money to go invest other in other places. Right? Like, in the traditional kinda layer two, layer three kinda portion of the of the overall expense. But the concept of networking as service isn’t new. Right? I mean, the concept has been in the industry. It’s kinda circled for years. I feel like it at least hasn’t fully landed. I’m still defining it when I talk to customers. What’s different about this moment?

[00:20:52] Bill Long:  I think there has it’s been, you know, NAS is whatever someone says it is. Like, to your point, there’s been a poor definition on it. I don’t think I think this is one of the sort of the pure vision is only now maturing to the point that it actually is what people wanted it to be. A lot of people were calling MASV, and it’s just the same old services. I think that we’re only now getting to a form factor that it it truly is, you know, you turn up the service and you can go and do whatever you want with it instead of it being something that was is just a veneer on the same old thing. Uh, this is a fundamental rethink of how the service is delivered, how it’s commercialized, the user experience. So it’s a it’s a complete overhaul of of just making it easy all the way from the commercial model to the operating model to the support experience.

[00:21:38] Scott Kinka:  Let’s put ourselves in some listener shoes for a moment. Like, you know, you just dropped in as the head of IT at a, you know, mid market or small in a smallish enterprise business, let’s just say. Fortune 500, you know, between four hundred and five hundred. Right? You get the idea. You’re in there now, and, of course, you know, you’ve dropped you’ve jumped into somebody else’s network probably in various stages of configuration. Some of it’s been rebuilt, some of it hasn’t, some of it’s legacy. And the CEO hasn’t said the word network to anybody in IT in five years and frankly doesn’t care. They just expect it to be there, but it’s saying AI AI AI, cloud cloud cloud, you know, SaaS SaaS. I want efficiency. Go get me some even though I don’t really know what that means. But you’re looking at the network, you know, nobody, like, snaps a finger on the network as new tomorrow. It’s always in various stages of, like, legacy to new. What’s the first thing you would do? Just like let’s just on open up the practical advice bucket and just say, like, you’ve inherited somebody else’s network or maybe you’re in the middle of it. How do you just start preparing for what’s coming your way?

[00:22:43] Bill Long:  The only constant is change. So I think the first thing is to figure out what how can you start with a solution that gives you the option value that you need that you know you’re gonna need for the future. So I think step one is making sure that whatever you do next is gonna enable you for where you eventually wanna go. We sit down with our customers. We have this exact conversation. We’re like, let’s understand the estate that you have right now and then where do you wanna take it? Again, we know that that the future state that you wanna go go to is gonna be distributed, hybrid, multi cloud environment. So where can you start? Um, a lot of places that we start with our customers first is in the data center. So how can we how can we get your data center to data center connectivity operating more efficiently to reduce cost and increase your option value? How can we make it easier and cheaper to connect to clouds, connect the data centers to each other, and then that creates an attach point that is you wanna add headquarter locations, as you wanna add branch offices, as you wanna add other things, it’s where to start. Um, and then that that also if you get your data center to data center connectivity going well, that’s gonna be the landing point for AI. Do you if you talk about, you know, we we at Zayo, because we have a nationwide both metro and long haul network, um, we see where all the the AI labs, the hyperscalers are going. Um, and they’re coming back to the to the major sort of data center interconnection points and making sure that you’re thoughtful about your strategy of who am I gonna need to connect to in the future? Um, you know, I need to connect to you have, obviously, my own data centers, my own headquarter locations, a couple clouds. I’m gonna need to connect to Anthropic. I’m gonna need to connect to CoreWeave, and I need to be able to to do that in a flexible way starting with the data centers, getting a good scaled solution, um, for what you need today, but also doing it away on what you’re gonna need to connect to in the future is super important. Typically, you get a lot of cost savings out of reimagining how you do your data center connectivity that can then go fund, how do I overhaul my headquarters locations by branch offices, etcetera. So that’s kind of the pragmatic

[00:24:40] Scott Kinka:  approach. No. I like that. Data center first. I mean, what it’s assuming a little bit of right now is that that the AI importance in the stack is coming, but is tangential to the data center. You know what I mean? So you’re like data center first, consider what that was from a model, cascade that out to locations in humans. Right? What’s changed if you go through that model I laid out earlier around COVID, right, and then take your answer now. Now I’ll ask you to look forward. Um, and, obviously, the answer is probably all, but I’m just having fun with the question. Right? Like, any plan has a base unit of measure. Right? The denominator in the equation. In five years is that humans, locations, data centers, models, or data. Like, what’s the first base unit to measure in your network design?

[00:25:30] Bill Long:  I think machines talking to each other is gonna dwarf anything that has a person. You can only look at so many CAD videos. Right? When we look at where we’re gonna build our network, the opportunity to enable a a use case where it can be machine talking to a machine, we are much more interested in than a use case that is a that includes a person on any end of it. So I talked earlier about the automating to factories, um, and to even convenience stores. Machine to machine use cases are are gonna be are going to dwarf any sort of use case that involves a human at the end. That’s kind of the first. Now what that unit is, I’m tempted to say tokens. I’m not sure that’s the the the right measure, but it’s gonna be what are the inputs that intelligence is gonna need to act on and then how do they act on it. One thing that we’re be we’ve been putting a lot of thought into is that when you’re connecting the way people are designing data centers today and the amount of bandwidth needed, they say, okay. This data center is 100 megawatts. You know? That’s a that’s a big data center. And you think about the bandwidth needed at a 100 megawatts, and and that’s kind of and, you know, people think that that’s a fixed when you’re when you look about the data center, it’s a fixed the amount of power is a fixed variable. AI models and AI hardware are getting exponentially more efficient. So the bandwidth that you need today for a 100 megawatt data center is probably one one thousandth of the amount of bandwidth you’re gonna need ten years from now. So we’re looking at it the planning horizon needs to be, you know, two, three orders, not percentages, two, three orders of magnitude different for the same amount of power. So it’s the the planning horizon as you’re thinking about AI is is just a very different kind of model.

[00:27:08] Scott Kinka:  It’s like, uh, you you take Moore’s law in the at the advent of broadband, and then you just turn it up by 10 x, right, in terms of what it’s gonna need to and this is the year. Right? I mean, I’m sure you’ve seen that same data. At least the estimates are that this is the year that the machine to machine traffic’s gonna exceed, you know, human based traffic on the Internet. So here it is. I wanna go into a segment we call snippable moments. I’m gonna challenge you with a couple of questions where, um, some of these may replicate some of the conversation that we’ve already had, but I’m gonna ask you to answer me in a sentence. I know as a strategy guy, there’s always a but and you know what I mean? I I sit in that seat now, but, um, no buts. So what should leaders ask a network provider that they are probably not asking today?

[00:27:55] Bill Long:  How much are you investing in your network?

[00:27:57] Scott Kinka:  Okay. Fair. What’s one belief about networking that you think will age poorly over the next five years?

[00:28:04] Bill Long:  This will make people’s heads explode, that you need people to do it.

[00:28:07] Scott Kinka:  Okay. I like that. If a company wants to be more AI ready over the next twelve months, where should it start?

[00:28:13] Bill Long:  Understand the ecosystem you’re gonna need to connect to. And you you’re not gonna be able to do it over the Internet, so you need to start thinking differently.

[00:28:19] Scott Kinka:  Last one. What’s one lesson you wish every first time CIO already knew?

[00:28:25] Bill Long:  Don’t let the network be an afterthought.

[00:28:27] Scott Kinka:  I love it. All good. And we could have explored any one of those, but the good news is I think we did in the earlier part of the conversation. Alright. Next segment is gonna it’s called this or that. This one’s even more fun. Alright? I’m just gonna tell you two things, and you gotta give me an answer. Alright. World Cup or Super Bowl?

[00:28:44] Bill Long:  Super Bowl.

[00:28:45] Scott Kinka:  Mountains or beach? Mountains. Careful on this one. Podcast or audiobook?

[00:28:52] Bill Long:  Oh, podcast for sure.

[00:28:55] Scott Kinka:  Chat, GPT, or Gemini? Uh, Claude. Uh, there you go. I like it. More compute or more bandwidth? This is a loaded one.

[00:29:07] Bill Long:  Um, more compute because that means more bandwidth.

[00:29:09] Scott Kinka:  I love it. Speed or resiliency? Speed. Okay. And yours is the last one. I’m totally setting you up. Product or strategy? Oh,

[00:29:20] Bill Long:  um, products. It it’s the ideas are easy. Execution matters. Got it.

[00:29:26] Scott Kinka:  I love it. Alright. Well, that was fun. Last segment we’re gonna go into, just a couple fun questions for you before we wrap up. First, I need a hot take. Um, just give me something contrarian about how the industry is approaching AI infrastructure or networking, network modernization, what have you, that you think most of your peers would openly push back on. Be controversial for you.

[00:29:47] Bill Long:  Edge compute is never gonna happen. There are too many diseconomies of scale of distributing computer bunch of places. It’s getting right silicon at the right place is always gonna happen in large large pieces. You’ll it’ll be in, like, major metropolitan areas, but it’s never gonna be at the base of the cell tower or somewhere like that.

[00:30:05] Scott Kinka:  And I mean, is that really the let’s go even further. Is the edge compute discussion the the marketing engine of the silicon, you know, of the chip manufacturers really at the end of the day?

[00:30:17] Bill Long:  I don’t know if it’s the chip manufacturers because they don’t care that much about where the chips go. It’s that network equipment vendors and others who wanna be like, you could repurpose your central office and put GPUs there. No one’s good. Like, that’s not gonna work. It just there’s too many diseconomies in scale and, like, the power density and floor loads and the security of a central office is just it’s not gonna work.

[00:30:39] Scott Kinka:  I love it. Um, give me a shameless prediction. It does not need to be about technology, but it can, um, for the next twelve to eighteen months. It could be who wins the Super Bowl since you mentioned that earlier. But give me something we could look back on.

[00:30:51] Bill Long:  Um, next twelve months, I think this the we’re just at the beginning of it dawning on the rest of society what AI means to us. I think, like, the Hugging Face hack and the data center pushback is there’s it’s gonna be the defining I think the election’s gonna bring it to a head, but I think it’s worth the beginning of a sort of, um, societal shift, and I think it’s gonna come into focus in the next twelve to eighteen months. Interesting.

[00:31:18] Scott Kinka:  And that manifests itself in are we embracing AI versus railing against the cost and infrastructure to be able to support it kind of model? You’re on the you’re on the we’re gonna do this or we’re gonna go back to our green roots and save the planet?

[00:31:35] Bill Long:  I don’t think we’re going backwards. Like, I I think it’s just either it’s proven to be too valuable, but I think it’s hugely disrupted to society. I think it’s largely the first thing that it’s gonna do is, like, the crap part of our job we don’t really like, we’re not gonna have to do because AI can do it, which is gonna be great, but it’s also gonna be scary to to a lot of people who, like, that was their job. Um, so I think there’s we got a lot to digest as a society. I think it’s gonna come into focus over the next twelve to eighteen months.

[00:32:00] Scott Kinka:  We’ve had a lot of conversations about that on this show. I mean, I don’t even I don’t even know that we have in The US, we still value workers on, like, a post World War two style of work ethic. Like, you showed up early and left late, so you must be really good. Like, what do we do with, you know, gen zer who pops in and pops out 8,000 lines of code in an hour and stops working? Like, is are are they a bad employee? I don’t know. You know what I mean? We certainly don’t have the mechanisms to evaluate their value today compared to the way that we operate. Um, that’s interesting. Uh, what’s on your nightstand right now? What are you reading?

[00:32:37] Bill Long:  Oh, I read voraciously. If if I can remember the names of the books, I’ll I’ll there’s one on the Nazi Germany before World War two and the economy. I can’t remember the name of it. One by by, uh, William Harrison. Yeah. I wish I could remember the names. I can’t remember the names.

[00:32:52] Scott Kinka:  We’ll we’ll check back in whether you put it in the liner notes. I can surprise you with that one. Um, alright. Last one. I’ve asked every guest. I think you’re guest, like, a 108 or a 109 somewhere in there since we started doing this thing. Are you a sci fi fan? Not at time. Not really? Okay.

[00:33:05] Bill Long:  I can roll with you.

[00:33:06] Scott Kinka:  Well, that’s alright. You had mentioned Nazi Germany. There’s been pretty good pretty good, uh, dystopian futures written about that. So you can choose a dystopian future, some sci fi epic, some alternative history, whatever it is. Either way, you are alone in some alternative future, um, and only one application is still working on this phone. Which one is it? Text? Okay.

[00:33:29] Bill Long:  So I can chat with people.

[00:33:31] Scott Kinka:  I got it. Your it Yeah. The future will be very boring. You need to keep in case somebody. Employer. Does it have to have MMS in it, or if you just, like you wanna be able to send photos and things too?

[00:33:41] Bill Long:  No. No. Text is fine.

[00:33:44] Scott Kinka:  I love it. Alright. Well, that this has been a really, really brisk conversation, but a super impactful one. For listeners who wanna learn a little bit more about your work, um, about your the products that you’re putting out, you know, for the for for Zelle, you know, what’s the best way for them to do that?

[00:34:01] Bill Long:  Just go to zoe.com. There you can get to all of our products, see more about our strategy. Um, we’ve got big announcements just announced and more coming up. So lots going on.

[00:34:11] Scott Kinka:  Fantastic. And then finally, um, if you had to summarize your leadership philosophy in one sentence, what would it be?

[00:34:18] Bill Long:  Do big things with people you enjoy.

[00:34:20] Scott Kinka:  I love it. I love it. Um, we could explore that one for an entire episode, and I agree with that. Um, love it. You’ve been, um, you know, a great guest. It’s been super insightful. I got plenty of good content out of this for sure. And if you’re listening at home, um, or in the car or on the treadmill, I’m sure you were wildly entertained. Um, that is a wrap, uh, for this episode. And if today’s conversation gave you something to think about, what we would appreciate is a subscribe, a review, a share, and certainly pass this on to the members of your IT team for a listen, and we’ll catch you soon on another episode of the If you made it this far, there must have been some good content, or you’re just a fan or a relative of mine. In either case, we appreciate you spending your most valuable asset with us, your time. We don’t take that lightly. We’d also like to say one more time that we’re appreciative of Bridgepointe Technologies and their belief and sponsorship of this show. We hope that if you or someone you know is thinking about your company’s digital transformation or simply the next IT project that you may not have the resources, budget, or time to get to, I’m sure that Bridgepointe, one of the country’s fastest growing technology advisory and procurement firms can help. Check out bridgepointetechnologies.com. Don’t forget the e on Bridgepointe, or simply reach out to me at Skinka, skinka,@bpt3.net. Also, take a minute, please, if you would, to give us a five star review on your favorite platform. It helps give us the visibility to reach other people like you. Thanks for listening to this episode of The Bridge. I’m Scott Kinka. And until next time, there’s a lot of noise out there in business and in life. Do what you can to be the signal. Uh