38. Technology Silos are a Thing of the Past – Tech Field Day Podcast

Hey, tech Field Day podcast listeners. Just a quick program note here. We are going to be taking a couple of weeks off with the Tech Field Day podcast, since it aligns very poorly or very well, depending on your perspective with Christmas and New Year's.

We will be back on January 7th, uh, Tuesday with the next regularly scheduled episode of the Tech Field Day podcast. It has been a discipline full of walls of constraints of people that want to impact the way that technology is being used. But as our technology future continues to accelerate faster and faster, those walls are falling more and more.

And in this episode of the Tech Field Day podcast, we discuss the fact that silos are a thing of the past. Welcome to the Tech Field Day podcast, where we bring together a group of influential IT experts to discuss a single idea about key concepts in the industry. This podcast features a variety of perspectives from members of the Tech Field Day delegate community, and is often recorded in association with one of our Tech Field Day events.

Tech Field Day is a part of the Futurum Group, and this podcast is also published on our sister company Site Techstrong tv. And this episode, as we look back on 2024, we'll be discussing siloed. It is a thing of the past.

I'd like to take a moment for our guests to introduce themselves before we jump into the discussion, starting with Steven. Hello, this is Steven Foskett. I am the founder of Tech Field Day.

The, I don't know, chief Cook and bottle washer of this here community. And, uh, I am a, uh, I don't know, a big nerd. Uh, I love, uh, talking about storage and networking and clouds and AI and all that kind of stuff.

And you can find me, uh, every week on Textron Gang on Tuesdays usually. And of course, every Wednesday on the Gestalt rundown. And I'm Alist Cook.

I am a, uh, event lead here at Tick Field Day as well. I'm the newest member to join. Uh, I focus typically on the cloud and development and things that are, are interesting to me.

co nz. Of course, you can find all of this on LinkedIn as well. And of course, I am Tom Hollingsworth.

I am an event lead here at Tech Field Day, covering all things networking, security, and mobility. And, uh, I'm very excited to be having this conversation because that's one of the things that Tech Field Day is all about, is getting past the siloed nature of it. So let's jump into this discussion, gentlemen, because we've seen a lot of this over the years.

I can remember a time when everybody had to have their storage silos and they had to have their compute silos, and they had to have their wireless team silos. And at the end of 2024, a lot of those things have been put into a, uh, blender and kind of just like shredded and mixed together to create this, um, homogenous it goo, I guess would be a better way to put it. Um, are, we're still seeing some silos here and there, but what do you think has been the most successful thing so far in it as far as breaking down silos?

Well, if I may, Tom, I have a name for that. Gelatinous it goo that you described. I think we should call it gestalt It, meaning the big picture of it.

Instead of looking at the individual elements. It could be all of it. And then instead of having, I've got an idea.

Instead of having networking Field Day and Cloud Field Day and Security Field Day and apta, maybe we could do one that's called Tech Field Day. What a, isn't that a good idea? Tom?

I, you're, you're already way ahead of us here. 'cause you're, you're thinking about how all of these pieces intersect together. But I can already hear the discussion from the people out there.

Oh, tech Field Day is great, but what I really want to do is I want to talk about Technology X and there's a lot of people that just love talking about, I don't know, AI and cloud and 5G, you name it. Every time we think we've kind of knocked down one of those silos, another one crops up to take its place. This is, I think, a, a natural progression, right?

We've, we've seen that new innovation turns up. We've got a new kind of accelerator. Well, there's gonna be a specialist team.

You'll look after that kind of accelerator. Uh, in my career, I've seen it happen a couple of times. Uh, way back last century, I was a Citrix specialist and I was there as it came from being a, a point solution from the old wind frame days.

You can tell why I've got no here left, uh, through to being a core element is now part of Windows. Is is that multi win component. Uh, definitely we see that innovation happens, builds a silo, and then it gets shrunk down again.

We would've seen that with wireless. There was no wireless silo until there was wireless. So I think this is a normal progression that new silos crop up, and then things get pulled into the core, or we discover that the silo was completely useless.

And we don't know until afterwards which ones we're getting. Well, I would, uh, jump in on, on that. I, I recently had the experience of going through some old boxes from previous offices where I found, um, my entire collection of Storage magazine, of which I was a Backpage columnist for 10 years.

Um, I found, uh, some of the conference proceedings from Storage Decisions conference back in the two thousands. You could call them the naughties. Um, and those conferences, it was incredible, uh, looking at the people that spoke a lot of great friends, um, you know, looking at the agenda content, apparently I taught advanced storage school one and two, uh, in, at, at storage Decisions.

Um, it was four tracks. It was thousands of people that showed up. It was, uh, you know, uh, an expo floor with, uh, dozens of companies.

As I said, there were multiple magazines, uh, at least three that I can think of that focused on storage, enterprise storage, San Nas, et cetera, and all of that. I, I hate to say it hurts me. It breaks my heart.

'cause that's my, it's gone. It's, it's, there is no storage industry anymore. All the companies were merged.

Um, most of them were merged with HPE or IBM or Dell. Uh, some of them still standalone. I mean, we still do have some independent storage companies, but most of those wouldn't consider themselves to be storage companies anymore.

They would consider themselves to be data companies or AI companies or cloud companies or something, anything not storage. And it is incredible what happened in the space of my career where storage went from being a component to being a field, to being a component again. And you mentioned wireless.

I mean, sure, there was no wireless, um, back in the day. Another thing I found, Tom was an 8 0 2 point 11, no a no B, no C, no d no z, uh, adapter. Um, two megabits, um, in my basement.

Um, that was the beginning of wireless. And now, again, in the wireless industry, the wifi industry, most of the wireless companies are part of HPE or, uh, Cisco. Uh, am am I wrong?

I mean, is the wireless industry kind of experiencing this? And yet, wireless and storage is still incredibly expensive, important. It has people who focus on it.

It has, um, you know, best practices. It has great technology, it has great products. Why isn't it a field?

So I would argue that one of the reasons why it became less of a field and more of a feature was because we stopped focusing on the individual components. I mean, we laugh about the fact that there's a storage field day and a wireless field day, but there's also a reason why there's not a scuzzy field day or why there's an 8 0 2 point 11 field day. It's because eventually you get to a point where it, the, the lines blur a lot, right?

It used to be that getting that VM Max installed in your system and figuring out the most efficient way to, you know, partition the workloads and put your, your high io uh, uh, needs over here and have your archival over here. Uh, that all got eliminated because of AWS. So now there's this big storage array in the sky.

And I as a developer don't care. I just write to the bucket and the bucket takes care of everything that I need. Likewise, wireless in the day of the 8 0 2 point 11, no a, no B, no C was this newfangled technology that allowed me to take a device and plug an adapter in and, and not be at my desk.

And now it's so ubiquitous that the idea of trying to find an ethernet cable to plug a device in anymore is radical because everyone just assumes there's wireless everywhere. And I think that that is one of the reasons that a lot of the shift in it has been away from components and technologies to use cases to, um, to the way that we're going to develop the technology, which tends to blur those lines quite a bit. Uh, you know, look at something like the CPU, right?

We spend a lot of time talking about how CPUs have changed and shifted over the years, and they've added more cores and they've driven the, the clock speeds up. But now, I mean, does it matter how many cores the CPU has? Or is it more about the function of the device?

You know, an NPU is optimized for doing inferencing. A GPU can do inferencing and other things. Now it's less about the technology and more about the way it's being used.

I think we're seeing this sort of move towards platforms, bringing together all of those pieces. I mean, the, the first fundamental platform I worked on was, was a virtualization platform. And it hid away all of that complexity of storage and all of the complexity of networking allowed us to, to focus just on virtual machines.

Well, progressively we've seen that move to smaller and smaller units as we get towards containers and functions as the place your application runs. So I think lots of these technologies are being subsumed into something where it's managed either by a service provider or by some element of the software stack that we're using, the software collection that we're using. So a, a lot of the time we're just seeing these, these things disappear into an incredibly complex system that we don't need to know about.

It's just got a, a layer of abstraction. And one of the laws of it is there's always room for another layer of abstraction, um, to hide all the complexity below. And then once we've got this collection of, of, uh, abstraction layers that we need to bring together, we just layer another abstraction on top and, and hide again.

Uh, hopefully we're doing some cleanup of the layers below eventually. But it is a kind of natural progression to add, add these layers in and bring in more features in, and start assuming it as you say, Tom, assuming that there is wifi, um, you know, teenage children sitting at the beach, wonder why there's no wifi. And you have to explain to 'em that there's a difference between wifi and, and mobile networks.

And similarly, my, uh, my, uh, parents, my parents-in-law miss this stuff too. So, uh, yeah, they, they sort of, there's, there's always a lot of complexity underneath. None of these things can operate by themselves, but we hide that complexity.

We, we provide what I call advanced simplicity, simple to consume, but it hides a lot of complexity underneath. And I think one of the reasons that that has happened is, um, that we have an embarrassment of riches now. So think about, um, you know, basically my entire IT career, um, was spent as everyone's IT career, IT career was moving bottlenecks or, you know, we would say remov, but the truth is, we were just moving them because as soon as you made the bus faster, or the network faster, or the storage faster, or the processor faster, or the memory faster, or whatever you just revealed, oh, no.

Oh man, I can't believe I, you know, I made the storage real fast, but the network's slow and, or the CPU, whatever. Well, now we have this insane world, an embarrassment of riches where, you know, you, you talk about wifi, wifi is over a gigabit now, uh, storage, you know, with flash, the, uh, the advancements in flash storage even, um, even cheap garbage storage is faster than anything I ever used as a, a enterprise, high-end enterprise it, you know, administrator, um, you know, every CPU now, like my, my phone has a faster CPU, one of the fastest CPUs on the planet is, is in my phone, and it runs on batteries for like 24 7, right? Um, oh, you need more processing power.

Well, here's, you know, you know, the, the latest Apple, uh, system on chip has an NPU, uh, that will outperform, uh, the high-end graphics cards of just a few years ago. And it does it on like three watts of power. We, we, we have this embarrassment of riches, you know, we've got all the memory, we've got all the storage, we've got all the io, we've got all the networking, we've got all the processing.

We just don't have to worry about it anymore. And if you don't have to worry about it, then you don't need a specialist to tune it and configure it and puzzle over it and ponder it. You can just throw something at it.

And one of the things that I've noticed about application development platforms, app application platforms, you know, functions as a service, uh, cloud, is that for the most part, I mean, some of them do specify processor speed and storage speed and type and things like that, but for the most part they don't. And for the most part, it doesn't matter because basically anything you can run on it is gonna run fine. So you really don't have to worry about what it's running on.

You have to worry about what you're doing with it. And that's actually happening as well, um, with laptops. I mean, you look at the latest, um, ARM-based laptops, especially, you look at the, the Mac, you look at the phones coming out of Google and Apple and the rest, and, and frankly, they're good enough for anything you wanna do.

So let's not obsess about the hardware. Am I wrong? Is, am I betraying all nerds everywhere?

I don't think you are. And the fact that we no longer track versioning and things like that as closely as we used to, like, you know, it's, it's an event now when, uh, an old phone drops off of the support list, or, you know, when you find out that, oh, this, this feature that was just released in 2024 is actually capable of being run on a device that was released in late 2020. You just kind of expect that.

And I think that the, the embarrassment of Rich's problem that you discussed, also, the flip side of that is that when you have scarcity, when you have, um, difficulty optimizing things, you create a boundary where people can exert power. And that's what it's always been about, right? It's not so much that the storage team was a group of specialists that were capable of optimizing disc layouts and things like that.

It was that storage was a finite resource, whether it was speed or, uh, availability, and that someone could control that. And through that control, they could effectively force other parts of the business to kind of bow to their whims as it were. You know, oh, you don't like the fact that this application's running slow across the network.

Well, as the person in charge of the network, I'm the one who has to do all the resource allocations. So you better play nicely with me if you want that to happen. And when we got to a world where effectively there was unlimited CPU an unlimited storage provided, of course you have an unlimited credit card, then those problems went away.

And two things happened. One, people realized that hoarding those resources were, was not as important as it was before. I don't need to have the greatest, um, AWS instance out there that cost me like $5 a minute to run.

I can get, I can program one that just runs well enough for the workload that I need. But secondly, by mixing everything together into our Gestalt IT concept, we realized that controlling access to those resources wasn't as important. And so now that we've moved past that, we've focused on the outcomes instead of the building blocks to make them, It's thinking about how is what I'm doing making the business better?

Rather than can I move this knob from six to seven and can I, um, basically some of it is about getting rid of it, little empires that are actually contrary to the benefit of the organization. And, and, uh, you know, that description of being that the arbiter of access to the network, well, that's not necessarily in the best interest of the organization. So there's definitely some, a huge amount of value in that shift up towards what, what is the business value of what I'm doing?

How can I improve the business value of what I'm doing? Because of course, in the end, the only reason we have it and is, and, and computer systems is to help the businesses be more successful, help the organization to be more successful as an, as an organization. So I really like that there's been a progressive shift towards focusing on what value do we bring?

And one thing I'll add there, Alistair, that when you stop focusing and obsessing about the technology, and you start thinking about the business needs and outcomes, you're starting to think more like a manager. So basically, we kinda gave ourselves a promotion. Well, I do wanna maybe throw a wild card in here because I can hear, I know this hasn't actually been published yet when we're recording it, obviously, but I can already, uh, uh, I have a premonition that somebody listening to this, maybe it's you dear listener, is screaming at us saying, you impossible idiots.

How can you possibly be be saying this? Have you heard of ai? You are saying that we have everything we could possibly use.

We have all the power, all the compute, all the storage, all the bandwidth we could po Have you heard what's happening in it? And to that person, I wanna say you were right. What we've, what we've got, what we're describing here in my mind is the infrastructure needed to run ordinary, you know, conventional systems, you know, oh, it's an ERP system, it's a web server, it's a business, you know, CRM system, whatever it is, those things have an embarrassment of riches.

And those things can essentially be run on your phone these days. What we've done though, is we've, by creating this just incredible flood of CPU power and bandwidth and storage and everything is, we've enabled an entire new, like a, just like a quantum leap ahead of applications. And that's what I, what AI is.

And, and if people, I know, I know, I, I didn't mean to talk about ai. Uh, we talk about AI every time, but if pe, when people talk about AI applications, what they're really saying is this order of magnitude different, multiple orders of magnitude, different application that requires just incomprehensible levels of everything. And so if we didn't have an embarrassment of riches, if we didn't have effectively unlimited CPU and accelerators and networking and storage, and we never could have done any of this, and that wouldn't be on the table because it would be impossible, but because we have unlimited resources, we're able to run insane applications on unlimited resources, and then suddenly that makes the unlimited resources limited, and we're back to constraints.

And so if you want to find ban bottlenecks to move around, and if you wanna find constraints, just go to the HPC conference, go to supercomputing, go to an any AI conference, tune in to some of these, uh, folks who are out there talking and writing about, um, AI applications and prepare to be boggled as we talk about gigawatts of power as we talk about, um, more compute than can fit in a data center as we talk about, you know, bandwidth specified in, uh, you know, not just hundreds of gigabits, but, uh, terabits as we talk about, you know, exabytes of data, um, none of the, none of these things are even comprehensible in when you're talking about a conventional application or even a modern application, even a modern web scale application. These things are only comprehensible when you're talking about this other thing that we have called ai. But really isn't AI necessarily.

I mean, AI is kind of a description of part of, of it, it's just mind-boggling compute. Am I, am I wrong guys? No, you're absolutely not wrong.

I, I think about all of the stories that I used to hear when SDN was the, the cool kid on the block 10 years ago about how the network is the bottleneck and being able to provision servers should take hours and not weeks. And, and if we just had access to all of the ability to just automatically provision this, wouldn't the world be a better place? Well, the outgrowth of that is where we're at right now with ai.

The outgrowth of centralizing resources, of eliminating bottlenecks, of creating automations to provision things is where we are with ai, with, like you said, effectively unlimited resources. What could you build if the sky really was the limit? And the wiz kids at open AI said, I can build a human brain.

Yeah. Or At least I'm gonna build if the sky is the limit, I'm gonna build to the sky. And let's look at another area of technology that we do discuss on occasion that I think is in its infancy of solving hardware limitations that will explode when it does.

And that's quantum computing, because if you wanna look like what a silo looks like right now, quantum computing is faced with the ultimate silo, the laws of physics. Like we cannot make a, a quantum computer run at room temperature right now. It must be cooled to within a few degrees of absolute zero.

That's a resource constraint. 'cause there's only so much liquid nitrogen in the world. The cubits only live for so long, they only have a certain error rate, right?

But once we solve those problems, we're looking at the effective end of cryptography, at least the way that we've known it for years. And, and that's the thing that I think that people can't really wrap their head around is when I don't have anything to worry about, when I'm not worried about what is holding me back, where can I go with it? And they can't really envision where that is.

And so in a way, a silo is less a set of walls to break through and more of a blanket to wrap around yourself and feel warm and cozy in the fact that you can still conceive of a world that needs you and needs this technology instead of it disappearing into the ether of fueling the engine of AI or whatever is coming down the road. By the way, Tom, there's plenty of liquid nitrogen, but it's not cold enough. It's liquid helium.

That's, that's the shortage. That's the one that gets you cold enough for the, uh, for the quantum, uh, you know, the liquid hydrogen is that warm comfy blanket. I, I, um, definitely see there's, uh, harking back to a bit of Steven's comments around plentiful and, and everything expands to fill it.

It comes along with Jevons paradox. When you reduce the cost to consume something because it's plentiful. New new use cases are, are invented to fill the excess capacity.

And we've always seen this leapfrogging of, suddenly there's new hardware that is five times as fast as the hardware we used to have before, and we'll never use all of this capacity. And then use case turns up, software changes, and, uh, we use up all of that capacity, and now the hardware is the slow thing. And so we've, we've always seen that leapfrogging consuming everything that's possible.

This is just the human nature. We consume everything that's available to do things that have never been possible before. Uh, I think Quantum is an interesting one.

It's going to be still a little while before we see practical, reliable quantum that doesn't just make headlines in the newspaper, but it is a transformational way, way we u essentially do, do math. And, um, it, it will be really interesting to see over time. One of the things I come back to for AI is that vast amount of resource being used used to achieve something that kind of doesn't seem like it should, need that vast an array of resource.

We do that same kind of, um, reasoning in our heads with a lot less energy than open AI using in its data centers. So I do wonder whether there is going to be a real revolution in AI when the power of demand to either train or to do, uh, inference with AI just drops off because we approach it in a different way. This is, for me, where the future of AI sits is, is in a different way of doing the thing, not having to do this massive amount of resource.

And we've seen some of that. We've seen things like the small language models, so you're still getting a generative AI model, but now it's running on that TPU inside Steven's phone. Um, I haven't upgraded my phone recently enough to have that, but you know, we are seeing those transitions away from the idea you have to use infinite amounts of resource, even though there is a vast amount of resource in that one little fondle slab that Steven has in his pocket.

So I, I, you know, you've made me think of something too, and, and I guess that's sort of, uh, the grumpy old man rule, uh, shaking his hand. I want to quote, uh, the great philosopher Douglas Adams in saying anything that is in the world when you're born is normal and ordinary. Anything invented between the ages of 15 and 35 is new and exciting and revolutionary, and anything invented after you're 35 is against the natural order of things.

I wonder if that's what we're seeing here, because, you know, my, my father, uh, would have seen the integrated circuit as new and exciting and revolutionary. Today we are just like, yeah, whatever. I've got billions of transistors in my pocket.

Um, you know, the same thing can be looked at. You know, you look at, at Quantum, you look at ai, um, in a way, I think that's what we're going to, and, and to get back to the original premise of this, the busting down of silos, maybe what we're doing is we're busting down the silos that were invented between the ages of 15 and 35, which is why suddenly what was new and revolutionary doesn't seem all that new and revolutionary anymore, because that's just sort of life to people who are, you know, just entering the IT careers or in the beginnings of their IT careers. And for them, you know, they're excited and talking about the revolutionary world of ai.

Um, I guess their kids will probably think that AI is just how things work. Um, what are you, uh, am I wrong? Uh, uh, the silos we're busting down are just silos of our own mind.

And given that we are all over 35 fairly clearly, uh, yeah, the, these other silos we've created in our careers that are suddenly disappearing, and we notice them because they're our, as you say, they're our silos. They're the ones that are in our hearts. Um, you know, Stephen, you're a storage guy even though there is no longer such a thing as a storage guy.

And, and so, uh, this, these are our silos that are being busted, and that's the only reason we care. Whereas, yeah, my, my children don't care about building a server or, uh, being able to, to wire the network. Uh, they expect wifi and they expect cloud services.

Yeah, I think that to some of us, those silos look like, uh, unassailable walls and to others they look like hurdles. But in five years, even those hurdles will look like steps for people to step on and, and gain more, right? They, you know, it, it used to be that we thought that getting to the moon was an impossible feat, and now we're planning to go to Mars.

And, and even further than that, and it's not because we have, you know, discovered any radical new technology. We've just tried to use it more efficiently and tried to put things together in a way that makes sense to us. And as long as we have people who are willing to come up and say, I can do better.

I can build faster and I need more resource to do that, that we won't have to worry about reverting to the silos of old, we may find our own new silos, but that just means that all of this has happened before and all of it will happen again, and we'll continue to overcome it the way that we always have. I think that'll just about do it for this episode, gentlemen, but if people want to check out some of the things that you're working on, where can they go to find those, Steven? Well, uh, I will be running AI Field Day and AI infrastructure Field Day, uh, along with Edge Field Day this year.

Uh, you can also find me every, most Tuesdays, every Tuesday, most Tuesdays on the Textron gang, uh, talking about these topics and more and ranting about solar energy for some reason. That's, they got my, they got my number when it comes to that. Uh, and of course on the Wednesday, uh, gestalt News rundown along with my friend Tom Hollingsworth and sometimes my friend Alistair Cook, uh, along among others.

Um, and, and generally you can find me on the socials at s Foskett as an old man shaking my fist at the cloud. Well, if you happen to come to New Zealand, I live in toga, so you can find me here, but that doesn't often happen. co nz for my blog or find me on LinkedIn.

There's been a lot more content going out on LinkedIn. Of course, I'm around on some of the socials as well. And I will be, uh, in the, the States next year doing in 2025 doing, uh, cloud Field Day events and maybe some of the other events as well.

Maybe I'll steal some from Steve. And of course you can find me wherever there is a networking field day or a security field day coming up in 2025. If you like to watch me a little bit more frequently than that every Wednesday I am on the rundown with my cohost Steven and Alistair and a wide cast of amazing people from across the IT industry who have spent their entire lives breaking down silos.

Um, I also, uh, participate in other fu and group things like the, uh, 5G Factor with my friend Ron Westfall. And of course, if you want to catch my daily ramblings, make sure you head over and follow me on all of the socials. I'm more active on Blue Sky now than just about anywhere else as we continue to build that community up.

So make sure that you follow along. net there, and wherever you find me, just look for a networking nerd, and I promise you those packets will come tumbling out of a wire somewhere. All right, well, thank you very much for listening to this episode of the Tech Field Day podcast.

If you enjoyed this discussion, please subscribe on our YouTube channel or your favorite podcast application of choice so that you don't miss an episode. Do consider following and leaving a rating and a review as well, because that really helps people understand what we're all about here. This podcast is brought to you by the Tech Field Day, the home of IT experts from across the enterprise, which is a part of the Future Group.

com/podcast or look at us over on Textron tv. Thank you very much for listening, and we'll see you in the next episode.