88. Tech Field Day Takeaways for 2025: Making AI Less Manic – Tech Field Day Podcast

We're wrapping up 2025 on the Tech Field Day podcast, and we thought it would be a good idea to bring you our big field day takeaways for the year. Welcome to the Tech Field Day podcast, where each episode we bring together a group of influential IT experts from across the enterprise IT community to discuss topics and themes. In this special episode, I'm joined by my wonderful coworkers, Steven Foskett and Alistair Cook as we wrap up 2025 and bring you our big tech field day takeaways.

Before we jump into our discussion today, I thought it might be a good idea for us to take a moment to reacquaint you with the Tech Field Day event leads from this year. Steven? I'm Steven Foskett.

I am the overall president of the Tech Field Day Business Unit for futurum, and I focus on a lot of new technologies, uh, principally AI Field Day, and I'm looking forward to doing some other topics in 2026, actually, uh, 25, I did Mainframe Field Day too, which was interesting. I'm Alistair Cook. I'm the newest member of the, uh, tech Field Day leads.

I look after AI infrastructure, field Day, cloud Field Day, also app dev Field day, and each field day. I think it's gonna be an interesting, uh, collection of things to, uh, look back on as we've seen lots of things going on. But what do you look after Tom?

So, I'm Tom Hollingsworth and I look after all of the data transport areas, networking, security, and wireless and mobility. And it has been a very interesting year. There's been a lot of advancements in the technology and I can't wait to see where we're going to be going in 2026, thanks to all of the effort that has been put into driving Enterprise it.

So let's start off with kind of our big takeaways from what we learned at Tech Field Day in 2025. Steven, you wanna kick us off with that? Sure.

Um, you know, my biggest takeaway is related to AI mania. So 2025 has been all about AI infrastructure investments. Uh, it's been in the news, it's been in politics, it's been probably around the kitchen table.

It is everywhere. But that being said, so far we are still at a sort of a, where's the beef stage of AI in the enterprise and in other areas as well. And that was something that I really focused on this year with AI Field Day, with the utilizing tech, uh, podcast with the new a utilizing AI podcast with all these episodes of, of the Tech Field Day podcast that I've done.

It is that, that core question, where's the beef? And the answer is there's beef. Um, there are useful applications, there are productive applications, there are applications that bring in more value than they cost for ai, but they're not necessarily the applications that people are thinking of.

Everything from, uh, what we saw when we went to Click Connect and they talked about how they're using AI to do things like data classification. Um, NetApp talked about using AI to use to do data classification. Uh, we talked to companies like Articulate that are building domain specific AI models that would help out with various, uh, verticals.

And all of these are useful, productive, business focused tools for ai. And I wanna call out something that we actually heard just this week as we're recording this at Commvault Shift. Uh, they had a speaker who I've seen, uh, I've spoken with on, uh, tech Field, a extras I've heard speak before.

Dr. Uh, Roman Chowdry, uh, from Humane Intelligence. And she made it very clear and pointed out, AI is a tool.

It is a new, um, arrow in our quiver. We can use it to do productive things. Most of us expect it to some sort of be this magic wand.

It's not magic, it's just a thing that we can do things with. And that's really what I focused on all year long. And that's what I'm gonna be focused on next year as well as we continue AI Field Day as we continue the utilizing AI podcast and all of the, uh, AI related field day things.

So, uh, little bit of background knowledge for those of you watching this. Um, we have a joke, an inside joke at Tech Field Day that we could name every Tech Field Day event, AI Field Day, AI, networking Field Day, ai, security Field Day, because that's what everybody's talking about. That doesn't mean they're all the same topic.

It just means that everybody's trying to figure out what I've just talked about, which is how do we use ai? And we see the AI adoption as being very much of a layered element. So my AI infrastructure field day events have been quite diverse.

We've had events that were very centered on how do you build the networking to supply large numbers of GPUs as as you're building out your, uh, training infrastructure. I know a guy who could tell you about That. Yeah.

And, and then once you've got all of the, the hardware deployed out, oddly enough, we haven't really had a lot of conversations about how you put GPUs into servers. 'cause that's not particularly hard until you're trying to put out a lot of them into racks. And then the hundreds of thousands of, uh, kilowatts that, or hundreds of thousands of watts that you, you need to power these things becomes a challenge.

But a lot of the time, the real challenges, getting that business integration, getting from we can experiment and we can build a few little things using whatever development system we have to, we can address 200 different use cases within the enterprise organization and we can do it continuously for the next three to five years in a way that is not gonna bankrupt our business. But as you say, say Stephen is gonna bring more value in than is being spent on achieving that value. And that transition from experimentation to large scale production is where people are, are struggling a little bit as they try and find the ways to actually get the value.

Because there have been huge spends. And we're seeing, uh, as we've seen on the Rundown podcast throughout the year, a continuing trend of these large vendors that are expecting to run your AI workloads, building out massive infrastructures, making contracts for three and five years for power and for data center space for the expected huge growth in a I use. And all of that spend has gotta come with some revenue coming back.

I think one of the things that's been really apparent to me over the year is the fact that a lot of these companies really are trying to not only search for the value in ai, they're trying to explain why there's different value to be had. And all you've gotta do is look back at several of the presentations that we've seen to look for the commonality of the phrase AI for something and something for ai. Because there are two very different ways to approach this problem.

A lot of people in the industry are taking the first track, which would be something more kind to your event, Steven AI for something. And that's using algorithmic uh, processing, that's using AG agentic AI to make a service or a product better. Uh, whether it's rewriting your words to make sure that you are better understood, whether it's using AI to collate statistics or to detect threats before they become massive breaches.

We are harnessing the power of what people are building to make operations life better. That has been a huge theme throughout the year, because that is one of the most easy things to see when you put a button in a dashboard to say that people can pick up on that very quickly. But also it's that understanding that for a lot of companies, many of the ones that Alistair speaks to at AI Infrastructure Field Day, building your infrastructure the way that we have done it for so many years is not going to work because these machines are so hungry for resources that it is nearly impossible to run them the way we would like.

And when you look at the shift that's been going on in the industry away from legacy technologies like InfiniBand to more modern technologies that we call ethernet but aren't really ethernet, it, it helps people understand really that AI is a unique challenge to the industry. And we can't think about it in the way that we have been thinking about things before because we will fall down if we do. Yeah.

And and I, I think I should point out too that the core question that you just brought up there, Tom, AI for X or X for ai, that's the, the differentiator between the field day events that we're doing. Hmm. So my events are AI for, and that's what AI Field Day is for your events are for ai, which is what AI infrastructure Field Day is for.

And so we have a common question sometimes from the delegates, from the companies, from the, uh, viewers remotely. And it is sort of like, what is the differentiator? What is AI infrastructure field A, what is AI field A, what should we present at these?

And again, behind the scenes for the folks that watch Field Day events, um, we do a lot of work to help these companies prepare for their presentations, help them to understand what they should present and how they should present it. And and that's really what we tell them. You know, I I have no problem with a, you know, for example, HPE coming to AI Field Day because they came in with a extremely credible story about how companies can use AI to achieve goals.

They brought in partners, they talked about use cases, they talked about the ways that AI was helping specific companies in specific areas to achieve their goals. HPE could also easily come to AI infrastructure field day or study field day, or you know, networking field day, and they would tell a different story. Mm-hmm.

And I think this goes as well to this sort of core question of ai, which is, it's not that AI is a thing, it's that AI is a lot of things and it depends on how you wanna present it, how you want to y your audience to adopt it, and what they're looking for and how you can help them achieve their goals. It very much depends where you are in that we responsibilities lie in this. So my AI infrastructure events clearly are for the people who are building infrastructure out the things that you need to do before you can do something with ai.

And we have seen that absolutely. There's some commonalities to all of the previous infrastructure, whether it's buying cloud services or deploying services in your own infrastructure or some combination of both of these things. But also that there's a whole lot of infrastructure around feeding that machine, feeding in particular good quality data into it.

Now, some of the AI providers say, you just dump us all of your data and we'll work it out. Well, that means they're doing the data processing stuff inside ai, and maybe that's not the best place to do data labeling, data ca categorization, data cleansing. It might be an expensive way to do something.

You could do better using a different platform. And so building out that pipeline of how you get from the data you have to the data that's going to provide the answers through your AI system, that's been one of the, the themes of the AI and infrastructure side. And I imagine it's also been quite a strong thing a across on the AI field day side.

Mm-hmm. And, you know, there's actually another thing, Tom, that I wanna get into here, and that's related to you. So AI for security and security for ai.

And so when it comes to security field day, I'd say if there's another mega trend in 2025, it's security and cybersecurity threats. So let me throw that to you. What's your feeling about AI for security, security for ai and the impact and the impact of the current just, uh, apocalypse of, of cybersecurity incidents that have been happening?

It's Funny that you say that because one of the things that we are seeing is the true application of AI being used as a force multiplier, but maybe not in the way that we want it to. The first point, AI for security is pretty cut and dried, right? We now have these new systems that we need to understand in order to secure them, but we don't actually know how to do that.

And we're trying to find that out. I have talked to a number of companies over the year that have told me about how they're doing prompt injection security. They're preventing people from jailbreaking online models.

They're preventing you from being able to harness a agentic AI concepts to attack other systems. And that's not an easy challenge. Something as simple as shadow ai, can you track how many people in your organization are potentially feeding sensitive or critically, uh, compartmentalized data to an ai?

Do you know how to look for that? And some people have come up with some very novel ideas and created a way to prevent that from happening. But you may not have even known that that was even possible because how do you know what people are using?

Are they using a platform that you've approved of? Are they using a platform that was number one on the app store for a week? There's no way to know that for sure.

Now, let's talk about the other side of that, using the AI to build better security, because one of the things that we know that human beings are very good at is pattern recognition. It's also one of the things that we are extremely bad at when you throw enough data at us. And when we started to see this in the machine learning, um, algorithms that we're kind of the nascent AI that we might see today, we were starting to surface these ideas.

We were starting to see, oh, well perhaps someone logging in at an off hour from a bad location is something that we should deal with. But as we've surfaced more and more of these data points, and we've created this ocean of data, not just a lake anymore, but a huge ocean of data, we can't make sense out of it. And the only thing that can is a dispassionate program that works 24 7 to give us the answers.

As people though, we still have to go back and look at it and decide if this is valuable reliable data that we can use to base our protection infrastructure on. Because if we don't, we are going to find ourselves not only drowning in the amount of information that we're getting, but as we're starting to see towards the end of 2025, people leveraging AI tools to become better attackers, whether they are using AI to generate the attack tools that they are building, or in some cases they're leveraging the AI platform itself to make the attacks happen automatically. Yeah.

Persistently permanently. Yeah. And we heard that from Fortinet.

We heard that from Microsoft, uh, very strongly. I mean, these are companies that are confronting this stuff every day. They are.

And, and the, the traditional stalwarts that are building the infrastructure that Alistair talks about quite a bit are racing to develop protections against this so that their customers feel like they are being taken care of. And the startup culture is doing the exact same in the hopes that they will either land a couple of really big customers that will help pay for more development or be purchased by one of the larger organizations that realizes they can't catch up with their own development. It, it's a great time to be a security researcher who specializes in these areas, but it's also a very important time to understand that you have to have a perspective and not judge how things should go based on what you believe the outcome should be.

Because AI will take any inherent bias in what you build into it and amplify it to levels that could potentially cause catastrophic loss exposure. And quite honestly, ending up on the news, which is what a lot of these CEOs really don't want to have happen. We're starting to see even what amounts to a supply chain attack through AI tools.

So we were seeing some news recently of attacks that came through Claude Code and that were very targeted manipulation of Claude Code. So it was kind of like prompt poisoning it. It appears, uh, to inject this, we want you to do something if an somebody from another company comes into Claude Code.

And that's absolutely terrifying. Um, leads me towards wanting to own my own destiny in, in the world of ai to have my own AI system that's completely isolated. The idea of sovereign AI is absolutely coming along for us, and that drives a need for a very different kind of infrastructure than if you're building your AI applications using entirely cloud services.

Yeah. And that, that idea of sovereignty is not a new one. I mean, we, the three of us have been talking about that with, with regard to cloud mm-hmm.

For two decades now. I mean, Tom, you know, all of us have been to these events, you know, we, but, uh, one of the things that was very interesting this year that is not visible to our audience either is the salon that we've been doing. And so we have another offering for Tech Field Day that is completely, you know, behind the curtain.

And that is that companies that are developing products have us come onsite with not the typical delegates, but with CIOs and CTOs from, you know, end user companies. And we do a day of discussions about what is concerning to them, what they're looking for, what their goals are, what they, what they're lacking in the market. Um, and we did a couple of those this year where the topic really was what you were just talking about, Alistair sovereignty.

Hmm. It is CIOs and CTOs saying, I am terrified of data. I'm terrified of spiraling costs.

I'm terrified of what AI could do with my company. I want to own it and control it. And these are not tiny companies.

These are the biggest companies, the companies that can do it. I mean, we had banks, we had health insurance companies, we had, you know, financial services of all sort insurance. And, and they are very interested in working with ecosystem partners and building AI platforms, building data platforms, building secure clouds so that they can have sovereignty of their data.

And of course, there's geopolitical issues as well that are leading people to say that. So we, you know, been international. I mean, you, you've, uh, obviously, uh, went, uh, overseas for Cisco Live.

You've, well, I think you've been overseas a couple of times and, um, and people in other areas are increasingly worried about being overly reliant on the United States. Um, there's obviously a lot of worry about the geopolitical issues of revolving around Russia, around China. Um, some concerns about India, um, you know, Europe going its own way with data protection.

And all of these are leading to an increased, um, focus on sovereignty, which is, again, this is not a new thing, but we're seeing it actually happen. Yeah. I think the biggest thing there is that these companies understand that there is inherent risk in the market.

Whenever anything new is developed that they have no clear control over, it introduces that uncertainty of risk. And that's something that traditional companies do not like at all. If they don't have a proper control that can accommodate that risk and do something with it, they feel like they're twisting in the wind.

And so they're trying to create risk barriers in those places, but they don't know where to build that. And unfortunately, it's not as simple as asking an AI make this less risky, because by asking that question, you have effectively made things more risky because you're introducing additional data points, the possibility of hallucination, uh, opening yourself up to a possibly, uh, having a, a prompt injection or prompt, uh, surfacing attack. And, and this is the thing that will keep your CISO awake at night trying to accommodate for all of this, while also being able to produce a report for the executive team that says, yes, we have managed all of this unaccounted for risk.

I wanted to pivot us because we haven't only talked about AI this year. We have had other topics. There are other things that people want to talk about, security for ai, but there've been plenty of things that haven't been all about ai.

Mm-hmm. Where we've seen, for example, you were at Convault Shift very recently. I was at CubeCon very recently, application development, data infrastructure.

These are all things that continue to be important for organizations, the ability to manage change, the ability to accommodate changing markets, changing geopolitics. These things all continue to be present as, as requirements for IT organizations for development organizations. And we've seen continuing, uh, presentations around how to achieve agility, how to, uh, enhance agility within your organization, how to approach new technologies.

These are all topics that we've seen in various presentations through the years. So I didn't want us to get lost in the idea that everything is ai, even though the way you get funding is to just slap an AI label on anything. Well then to that point, I think it's kind of, um, amusing.

As I said at the top, uh, one of the biggest challenges when it comes to AI is that there just isn't much of a market yet for some of these products. Now, there is something of a market, but it's definitely not commensurate with the amount of money that is being spent globally on, uh, AI infrastructure, um, you know, CapEx. And yet there's a huge amount of money still being spent on really conventional stuff.

Uh, storage systems have been just rocketing forward. Mm-hmm. Um, you know, we're seeing incredible numbers from the hard drive members, vendors, the SSD vendors.

Um, the server vendors are still selling like crazy. Um, we're seeing Intel making a bit of a comeback, even as a MD is rocketing forward, even as ARM is being bought. Now, some of that is for ai, but a lot of it isn't because there's still a lot of people doing a lot of stuff that doesn't say ai, right?

Yeah, It is. And there's a lot of money to be made in the encapsulation. That's one of the things that I took away from a lot of the presentations this year.

Yes, we can focus on these big ticket items, we can see the billions of dollars that are being thrown around between the largest players in the market. But at the end of the day, the largest AI cluster, the most impressive inferencing system out there still needs to communicate with the rest of the world. It still needs to be able to be secured.

It still needs to be able to have users talk to it through their workstations. And that encapsulation doesn't exist outside of an AI cluster, but someone still has to maintain it. So as much as I would love to sit here and say that the only thing that you need to worry about is how many, uh, checks you're gonna be writing to the large organizations, you still have to have wireless access points.

You still have to have a cohesive security strategy that involves things like microsegmentation and zero trust. And you still need a network to be able to transport all of this data back and forth from where you want it to be, to where it needs to be. And if you, as someone who is making that development decision, believe that that's some kind of an encapsulated bubble that you don't need to worry about, it still needs to be fixed.

And that's where a lot of these companies are making advancements. Maybe they're not the billion dollar, uh, magical headlines that you see in any tech news site, but they still make the lives better for everyone who's using that technology. Whether we're upgrading enterprise wireless deployments, whether we're finding new ways to use things like DPU and Smart Nicks to accelerate server traffic, or something as simple as being able to classify a user when they log onto the network and automatically assign them resources to prevent them from being able to be used as a reconnaissance attack.

All of those things to me say that whether or not you are thinking that you need to jump into that world of ai, there is still a lot of money to be made in the market if you're willing to do the things that people do every day. And I, I think one of the, the sort of ongoing themes, again, we, we come back to themes that we've hit for 20 years, is that if, if you are good at it, you can be good at it with whatever the new thing is. But if you're fundamentally bad at it, some new technology is not gonna miraculously make you your organization better at it.

And In fact, I would say that it is probably gonna be exponentially worse. If you're bad at it, Yes. You'll, you'll layer more bad things on top of your existing bad things.

And so the, the fundamental idea of understanding how to use these new products, how to use your even older products in a production real world environment, I think drives to the core of what we want Tech Field Day to deliver to the audience for tech field data practitioners out there who are building, designing, uh, integrating these solutions. Fundamentally, it's, we, we want you to be better at, at doing it as a professional practice. Um, and very much one of the, the themes has been towards mass production rather than handcrafted prote, uh, perfection that you, you need to be a sort of manufacturer of it within your organization rather than a craftman of it.

And, you know, that I think is really a good summary of sort of where we are in this time, because yes, AI is making all the headlines, but you still have to be good at the fundamentals. You still need good data protection practices, good data management practices, good networking, uh, you know, appropriate approaches to building applications and deploying them and managing them, and all of these things that we've been talking about forever just because it's AI or not doesn't change that. So I wonder before we wrap up, should we talk a little bit about what we're looking for in the next year?

Um, what can people expect from Tech Field Day? When we come back from the hiatus here, we're gonna be on hiatus for a couple of weeks with the podcast, then we're gonna come back in 26. What are we gonna be doing?

Well, I'm, in terms of events, I'm kicking off the air with more AI infrastructure. And so we're, we're Absolutely all about ai, isn't it? Once again, it is all about ai, but before you can do the ai, you need the infrastructure for it.

And so we'll be looking at AI infrastructure, but we will continue to look at the thing that AI is making better as well. So through Cloud Fields a and, and our other categories, it's still the, the thing that the business does is going to be made better by ai. It's not going to, it's not AI that's going to do the thing.

I'm starting off 2026 by going on the road to talk to companies that are doing the building, whether it's gonna be at Cisco Live Europe or RSA Con, there is where the people who are making the big announcements are gonna be talking about all of the cool things that they want to be discussing. And I think that's super important because when we lead off the first quarter with all of those discussions, it gives us an opportunity to frame how the rest of the year is going to go, because the leaders in those markets are the ones who are going to dictate the pace of change in technology, and the rest of us are gonna have to play catch up. And I'll say that from, from my sort of, uh, the other things, uh, angle here at Tech Field Day.

Uh, I actually am not done on the year yet. So, um, again, as I said, we have this salon, uh, offering, which is a behind the scenes, no publicity, no recording kind of thing. Well, I, I've got one of them coming up with a major company that, uh, even your mother has heard of.

And we are going to be doing that, uh, in December. And then we actually are very excited to be returning, as Tom said, to do more of these tech Field day outside Silicon Valley hotel rooms kind of events. So we will be on site at more, uh, industry events in the coming year.

We will be on onsite doing more of these salons. We'll be onsite doing a lot more topics as well. So we'll be back looking at data.

We're gonna be looking at, uh, security. We're gonna be looking at infrastructure events. Uh, we are gonna be doing, uh, private events with companies, public events.

Um, another big, uh, topic, as I said is, is, is kind of broadening the areas that we're looking at. We've always, Tom and I have always joked that we could do Tech Field Day on any topic. Uh, well, we're planning on doing that.

We're planning on doing Tech Field Day on a variety of different topics, um, that may, you know, kind of surprise you. I'm not saying mainframe again necessarily, but, you know, let just wait and see some of those. Also, um, I want to give a shout out as well that if you're listening to this and you wanna be part of Field Day, we would love to have you, if you are a C level exec and you want to come to one of these salons, which again, is, is completely private, um, let me know if you are a technical implementer.

If you're a thought leader, if you're somebody who's out there doing interesting things in networking or security, let us know. Go to Tech Field Day website. Uh, there's a form that you can submit your name or just reach out to us.

We're on social media And we're easy to find via email. I mean, uh, our email addresses are all up on the website. You can find us easily.

And we absolutely love having a community of delegates. I think we all, uh, find the time that we spend with our delegates to be the most rewarding part of our job. Apart from, of course, being part of a wonderful team and being here in, in Ohio, my first time here at head office in Ohio, it's been a delight to be surrounded by my favorite team for the whole time.

But the wider, uh, the wider tech field day community is absolutely, uh, very much of a family feel. Um, many have visited here in Ohio, which is not on the mainline place that people go to. Um, being part of that family is definitely a great thing.

If you are somebody who likes to learn about technology and likes to share the things that you learn about technology, And I think it's important that as we kind of wrap up 2025 here, we help that you understand that what we do here at Field Day is only possible because we have a great team that works with us. Whether it's Andrew and Corey who are constantly making us look good on the podcast or on our field day presentations. They're right there.

They're, they're right behind the, the camera, as a matter of fact. Or if it's, uh, Clara and Rachel and Matt who are doing amazing work with, uh, all the scheduling and coordination, we get nothing but compliments about each and every of them about how smoothly these events run. And we know that you love it too, because we're always bringing you the best content and we're able to do it in a way that coordinates with your schedule, and you can consume it in any way that you, you feel free to do so, especially with our partners at Prime Image Media who have done such an amazing job for so many years making these Field Day events so wonderful.

We wanna say a big thank you to all of them. And we wanna say a big thank you to all of you for continuing to consume Tech Field Day in the way that you do through all of the platforms that we are now a part of, whether we are on a new podcast, whether we're bringing you our videos through a new application, we know that this is the kind of thing that you wanna see and trust us that we will be bringing it to you each and every week in 2026, so that you can stay on top of each and every one of the advances in technology. Whether it is a Titanic shift in the way that we build things, or something as simple as a new protocol that might make things a little bit faster for you.

We also wanna say a special shout out to all of our friends at the Futurum Group, what Tech Field Day is a part of. com. This will just about do it for the 2025 episodes of the Tech Field Day podcast.

Please make sure that you are subscribed to our YouTube channel or to your favorite podcast application of choice, because when we come back from our hi hiatus, we're gonna have lots more episodes that we are gonna wanna share with you. But do us a favor, leave us a rating and a review on these episodes because that is the best way for other people to find us and understand what we do here. And hopefully join in the fun that you all have with us, al all the time.

We'll be back in 2026 with more great information. Thank you so much for being a part of the Tech Field Day community. We'll see you next year.