Tom Gillis [00:00:00]:
Think that we are in a good position right now, and we’re gonna get in a stronger position going forward. You know? Yes. The threats are, you know, more sophisticated, but our ability to create intelligent response with systems that used to be quite brittle, I think that’s a big step forward.

Karissa Breen [00:00:29]:
Joining me today and back on the show is Tom Gillis, SVP and general manager, Cisco security data center Internet and cloud infrastructure group from Cisco. And today, we’re discussing the future of cybersecurity and the network. So, Tom, thanks for joining, and lovely to have you back.

Tom Gillis [00:00:51]:
Yeah, Carissa. Good to be back. Thank you.

Karissa Breen [00:00:54]:
Okay. So I really wanna start with I interview you at Cisco Live, so it’s kind of a bit more of a follow on, bit more of a deeper dive discussion from our previous interview. But I really want to understand, in your mind, the need for cybersecurity innovation and why it’s important.

Tom Gillis [00:01:12]:
Well, it’s a it’s a fascinating time we’re in and that the the toolkits that are available to all of us in the tech industry broadly, specifically the combination of advanced silicon as well as these AI capabilities. You put those together. Wow. We can do amazing things. Like, computers can talk. Right? And they can talk in a sensible way. Well, that’s that’s amazing. So there’s kind of a good news, bad news to that, which is the good news is we’re applying a lot of those techniques to security defenses.

Tom Gillis [00:01:42]:
So the bad news is the attackers are applying those same techniques to security attacks. And if you read the popular press, there are organized nation states behind a lot of security stuff that are very ambitious with very broad goals of of penetrating major infrastructure and disrupting, you know, their adversaries ecosystem, which means everything from bank accounts to, you know, the water pumps in your cities to transportation to, you know, sort of petrochemicals, all this stuff is is a target. And that is really raising the game in cybersecurity. So lots of change, lots of activity that creates opportunity.

Karissa Breen [00:02:26]:
So would you say in your experience that from how I see it in media, obviously, being the coalface and looking at different vendors like yourself, would you say more vendors in particular innovating more than ever?

Tom Gillis [00:02:38]:
Absolutely. The the level of innovation is unprecedented. It’s unlike anything we’ve ever seen. And and I’ll just say, you know, for me, not necessarily as cybersecurity. Like, I run a technology engineering team, a very, very large product development team. And every year, that team gets maybe 5% more productive. Right? And every year, we make some incremental investment based on on our growth in the market. And so let’s say that, you know, maybe we add another 5% of engineering capacity.

Tom Gillis [00:03:09]:
Maybe it’s 10%. So total of 15%. We as leaders and managers spend a lot of time managing, you know, on this kind of marginal increase in in engineering capacity. Now all of a sudden, my team has these AI tools that make us 70 or, you know, but twice as productive, a % more productive. Like, it’s just incredible. So so so the ability for us to come up with new ideas, to do new things, it’s it’s unlike anything we’ve been through. So, yes, there’s a ton of hype around AI, but I think it’s pretty warranted. I mean, it’s it’s just an explosive time in almost every domain.

Karissa Breen [00:03:47]:
So then the other side of that is you guys are obviously releasing a lot of new things, like AI and defense came out last week. But would you say that there are probably vendors sort of, you know, lacking behind in the innovation space? And then what what does that actually mean when we say innovation? Is it new products? Is it new services? The way in which we do things. What do you think that means to you when I ask you that question?

Tom Gillis [00:04:08]:
Yeah. So, you know, innovation to me is very clear. There’s one definition of it, right, which just means doing something new. And I I I always get a little annoyed when I hear mature companies feel like, oh, you know, there’s so much innovation, and they’re talking about future development. That’s not innovation. Innovation is new. That’s the in no part. Right? Like, we are innovating.

Tom Gillis [00:04:24]:
And so AI gives us opportunities to innovate in many many layers. We can put AI interfaces on top of existing products. So we’ve already done this with a firewall. We put a natural language interface into it. And so instead of trying to manipulate hundreds of thousands of these very archaic firewall rules, you now can ask kinda common sense questions of like, hey. Can Tom access this, source code repository? And the firewall will tell you, yeah. He can. But I noticed over here that he also has access to a sales repository that maybe he shouldn’t have access to.

Tom Gillis [00:04:56]:
Right? So that level of of interactivity, that’s real. Like, we’re we’re shipping that today, and customers are delighted with that. Now I think AI also allows us to think about building products fundamentally differently and kind of picking up on that firewall thread. This is what we did with Cisco’s HyperShield. So we took what you think of as a traditional firewall, and we said, what let’s let’s completely reimagine this and do it in a manner that wouldn’t even be possible without AI. So this is what I call an AI native product, where it’s a product that can only exist if it’s leveraging AI. And so everybody’s doing this either or they should be. And, you know, you asked the question, you know, is is the whole industry embracing AI? I think by definition, no.

Tom Gillis [00:05:46]:
Some are gonna embrace it more than others. Some are gonna focus on the, you know, kinda let’s stick an AI interface on top of our existing product and keep focusing on our existing products. But others, I think, will be have the the vision and and the and the courage sometimes to, you know, completely reimagine their products in a world of AI because AI is so transformative. A lot of times, things that we’ve done for ten or twenty years no longer need to be done. And that can be hard to let go. So, you know, I said this earlier, but I I really believe it. Change creates winners and losers. And this is a time of unprecedented change, which means I think we’re gonna see a whole bunch of new winners and a whole bunch of new losers in the industry, you know, that’s different than just a few years ago.

Karissa Breen [00:06:29]:
Okay. This is interesting. We’re gonna have this demarcation between winners and losers. So would you say based on what you’re saying here, Tom, is, you know, if companies are not really embracing AI, what whatever their description of it is, so companies who are not doing that at all, is this when they’re gonna get into the loser category, would you say?

Tom Gillis [00:06:46]:
Actually, most part, yes. And I think you would be hard pressed to find a company that’s like, yeah. No. We’re not doing any AI. You know? Like, it’s so transformative. Any process that you have in business that could be written into a destruction manual or, like, a, you know, a run book or a playbook of some sort. AI can do that really well. So even a mundane task like scheduling, AI can schedule way better than we used to be able to do with kind of traditional applications.

Tom Gillis [00:07:13]:
So, really, to me, the the meaningful question is to what degree is a company embracing AI? That’s really the question. You know? There, we will definitely see shades of gray because it’s it’s new and it’s transformative and it’s explosive, but at the same time, there’s all kinds of stuff has not worked out. Right? So you can’t bet everything on it. You have to find the right level of investment resource, focus, vision, executive sponsorship. And, you know, that’s gonna vary depending on the organization. That’s gonna vary widely, I think.

Karissa Breen [00:07:47]:
And going back to the word degree as the operative word, what would you say? Where would that degree be? Would that still be going back to what you said before about AI native in terms of or what does that look like in terms of AI and the use of it?

Tom Gillis [00:07:59]:
Yeah. Well, so you you noted at the beginning of this podcast my very long and kind of cumbersome title. So my my role recently changed at Cisco where I run not just security, but I run all of our cloud and data center infrastructure. So when we think about a data center, a data center is a very procedural environment. Like, there’s a thing in a data center called a runbook, and it literally used to be a book that you would write. You know, here’s the 500 or a thousand steps to launch a workload. You push this button, do that thing, go to the system, get this number, wait for an IP address, you know, open a ticket, update these firewall rules, update a load balancer. It’s a whole series of steps.

Tom Gillis [00:08:36]:
Some of those stuff can be kinda complicated, but it’s just a series of steps. You know, I believe that AI can do all that stuff really, really well. So the opportunity, but also the imperative for Cisco is, like, you know, jump in with both feet, and we’re doing that. Right? So I have very, very large investments around this. And and and I’m not trying to solve every AI problem in the world, but I feel like we can use AI to make our existing products remarkably better. Like, totally change the way a customer thinks about how they operate a data center. And that’s fun and exciting and interesting.

Karissa Breen [00:09:10]:
So I wanna get back on the losers topic just for a moment. What was going on in my mind as you were speaking, do you think okay. Just say twelve months, you come back on the show, you talk to me about, you know, the winner, losers sort of landscape. Do you think that the losers or the people that are, you know, underperforming, do you envision that companies like Cisco will just go and buy them, or do you think that they’ll just phase out and we just never hear of these organizations ever again?

Tom Gillis [00:09:32]:
I have a very, very, strong bias towards this, which is in a world of AI, you have to be willing to blow up your existing products and rebuild them. And because the AI stuff is so is so transformative. And so I don’t wanna buy somebody else’s legacy product that that as in I have to go transform. I think there’s opportunities to to do organic development and because it’s kinda like everything needs a rewrite. You know? So might as well just build it yourself rather than trying to buy someone else’s, you know, 20,000,000 lines of code and refactor it. That’s a bias I had. It’s not to say we don’t do acquisitions. We sure do.

Tom Gillis [00:10:06]:
As we’re doing these acquisitions, I’m thinking about the larger architecture of the system and how do we make this AI native. How do we make this stuff, like, really sing? Like, kinda blow people away with the capabilities that we can deliver with AI. That’s my first thought on the afterthought.

Karissa Breen [00:10:20]:
Okay. This is interesting. So you said blow people away with the AI capabilities. So in your role, what do you think really blows people’s hair back here?

Tom Gillis [00:10:29]:
This isn’t like, oh, five years from now. I’m talking about right now today. Yes. I need to mature it. I need to, you know, sort of prove it out. But, like, I’ve got customers that are putting their hands on this. They’re like, oh my god. This thing writes its own rules.

Tom Gillis [00:10:41]:
This thing tests its own rules. It deploys itself. It upgrades itself. So so that frees those teams up to go do other more high value tasks of which there are many. So this is what I was talking about earlier. Remember, like, this kind of this this notion of abundance? AI is gonna create an abundance of intellectual property, like, you know, capacity, an abundance of knowledge worker output, an absolute abundance. And great companies will harness that abundance and and put it to great use, and others may miss it.

Karissa Breen [00:11:14]:
So the word abundance, I mean, you’re right. So there still seems to be people that are rattled by it. Like, oh, but I’ve been doing this for for twenty seven years, and I’ve built every line of code in this company.

Tom Gillis [00:11:24]:
%. Hundred %. Yes.

Karissa Breen [00:11:26]:
So what happens now?

Tom Gillis [00:11:27]:
Well, this is the point. This is why, like like, there will absolutely be winners and losers. And everyone’s gonna pay lip service to it. You’d be crazy not to. But I don’t think it was gonna be a thing. That would be, you know, for real. I’m sure someone says that, but, like, you’d be hard pressed to find that because it’s so obviously powerful. Question is degree.

Tom Gillis [00:11:43]:
Like, you know, like, do you really believe are you really able to capitalize on on these new ideas? Are you able to let go of the past and, you know, build for tomorrow? And how fast and efficiently can you do that? You know? And it’s hard to do. I don’t wanna trivialize it. I said, that’s very hard to do because you can’t overreach. Right? You know? If I tried to, you know, let’s say I built a machine that could turn water into wine. You go, wow. I read about that, you know, in a book sometime. Well, it turns out that’s actually very difficult to do. Or turn lead into gold, another one, difficult to do.

Tom Gillis [00:12:14]:
So don’t try that. Put something more, you know, constrained, like, hey. I’m gonna build a firewall that is highly autonomous. Yeah. Highly achievable. Very, very interesting. And the first of a series of steps that I think will add up to, you know, an absolute revolution in the IT and infrastructure world.

Karissa Breen [00:12:30]:
So based on what you’re saying, do you think this is a point in time where even small little startups could start to maybe not overturn big players, but have a really good opportunity because they are, you know, AI native. They are listening to, you know, people blown away by turning, you know, water into wine. They are thinking like that as opposed to perhaps some other, you know, traditional companies out there that aren’t thinking like that. Have you seen this before in terms of opportunities in your time being in the space?

Tom Gillis [00:12:58]:
Yeah. Over and over and over again. And I’ve started three companies. And so, you know, market timing is everything. I I my first startup was a content delivery network, a CDN. And I was out drinking beers with my buddy, and we were sort of accounting. It was a marginally successful project, wildly successful at first and then and then but not sustained. And I was like, god, you know, so many ideas we had were absolutely right.

Tom Gillis [00:13:24]:
We were just too early. If my buddy, this is over beers. He’s like, dude, we were too early, but we were too early by a decade. And if you’re off by a decade, he’s not he’s just wrong. Even if it’s the right idea, like, if you’re if you’re off by a decade, you’re wrong. So the point is market timing is everything. And this AI thing is one of these disruptions that is it is right upon us now. And some big companies are gonna be slow.

Tom Gillis [00:13:47]:
Some big companies are gonna go at it too hard and miss. Some big companies are gonna be slow and and and miss it there. And it’s gonna create space for for new players, undoubtedly. Absolutely, undoubtedly. I mean, OpenAI is a company that didn’t even exist a few years ago, and now, like, you know, it’s looking like a a, you know, a pillar of the infrastructure. There will be others.

Karissa Breen [00:14:06]:
So you said there might be big companies who go too hard and miss. What does what does missing look like in your eyes?

Tom Gillis [00:14:12]:
Oh, it’s a turn building one of those machines that can turn lead into gold. Right? You know, you you could convince yourself that this is a good idea. You could you know, you run a little prototype where it’s like, look, lead that kinda looks like gold. It’s yellow. And then you put, you know, hundred million dollars of OPEX to work for two years, and the lead never turns to gold. So, you know, execution is always a hard part of all this stuff. So you gotta be able to execute. And and there inevitably will be some that systemically failed to execute, and and I think that’s gonna really be impactful in this time of great change.

Karissa Breen [00:14:47]:
Okay. So the other thing that I’m curious to know now is, you know, being a practitioner historically in this space and then, you know, doing media stuff now, but one of the things is, you know, even going back a decade, which wasn’t the velocity just wasn’t there in terms of, you know, AI innovation deployment. How do we sort of balance now? We’ve gotta do these things, like you said, like, the, you know, the market’s here. It’s the timing. How do we then balance, like, still staying ahead of the game with security to make sure, like, oh, we’re trying to innovate so much, but we now we don’t even care about our customers because we’re not even protecting them at all. How do you balance that, what you say?

Tom Gillis [00:15:21]:
It’s hard. And that there is no one answer to that. That is judgment, and that is what management does. Well, that’s what defines great management from, you know, sort of good management.

Karissa Breen [00:15:32]:
So do you think there’s always gonna be this ongoing challenge between, well, we obviously wanna be innovative. We wanna have, you know, all these great things that we’re doing, but then there’s still gonna be the security team in the back that are like, oh, well, hey. Like, we need to look at the risk assessments and privacy impact.

Tom Gillis [00:15:46]:
And Always. Every single time. Every every single time. Right? Like like, you know, the old saying, the most secure server is the one that’s powered off. Right? Like, okay. Got it. Right? So we have to find that balance. And AI is particularly tricky.

Tom Gillis [00:16:00]:
Here’s why. With with a traditional application, you know, you have data that lives in a database, and you have application logic, and then you have what we call the presentation layer, the web the web interface. K? So you you can put controls in place and sort of understand all that. Now with an AI based application, there’s this new layer. Right? There’s a layer called the model that sits between the data and the and the applications. And the model sucks all the data in. It reads it, and then the data’s gone. You can’t touch it anymore.

Tom Gillis [00:16:27]:
It becomes like a like a gas that’s in the atmosphere. So so now this model knows a lot of stuff. You wanna make sure, like, hey. How does it not divulge that information? And I don’t know if this translates, to where you’re from, but I grew up. There was a game we used to call a hundred questions. So I’ve got a secret, and you’re gonna ask me a hundred questions to figure out if you get my secret. Usually, you could get you could guess the secret in, like, 20 questions. So that process happens with AI.

Tom Gillis [00:16:56]:
So the AI models understand sensitive information, Tom’s credit card information, Tom’s, you know, Social Security number. But if the app that I’m building on top of these models is trying to just administer a firewall, I don’t want that firewall talking about Tom’s credit card number. What the trick is gonna be that that if you play a hundred questions with that AI model, it might actually inadvertently divulge my credit card number. And so so this is why you mentioned earlier that we just launched a product called AI defense. And AI Defense uses AI to protect AI. So we use AI to ask not a hundred questions. We ask a billion questions. And we ask this model over and over and over again, looking to try to trick it to see if we can get it to inadvertently divulge something they shouldn’t.

Tom Gillis [00:17:43]:
It’s a radical new way of you know, that’s not a security control that anyone’s familiar with. And that’s just one example of how the industry is going to change dramatically in this AI world. Yes. AI introduces whole new classes of threats. Yes. The the nature of an application itself changes dramatically, but it’s not you know? These are changes we’ve got to embrace. Got to.

Karissa Breen [00:18:05]:
So then I’m curious to know from your perspective, what do you think rattles people the most about AI?

Tom Gillis [00:18:14]:
There’s a couple answers to that question because it’s it’s a broad question. In a security sense, you know, this idea that attackers are gonna use AI to trick me to to, you know, fool me in some way to, like, giving up some sensitive information, that’s a very real threat. Right? It is not unreasonable that, you know, you could get an email. Hey, Christopher. It’s great seeing you on the weekend. I took some pictures of you at the game. I posted them here. Click on this link from someone you know, and you were at the game.

Tom Gillis [00:18:41]:
Right? So so that’s that’s sort of scary. I think the larger thing, what we’re talking about kinda like shape of the industry stuff, is that this is great fear that, like, AI is gonna create this level of change. It’s gonna wipe out eliminate jobs. And, you know, that’s what I said earlier. A cynic might think that is true, but I’m an optimist. And I’m like, are you kidding me? Like, we are gonna find ways to do new things in more creative, more valuable ways than ever before. That’s why I think it’s gonna be just an explosive few years where the the productivity, the ability to to accomplish sophisticated tasks is gonna, you know, skyrocket in a manner that, you know, we haven’t experienced really ever.

Karissa Breen [00:19:20]:
Yeah. That’s interesting. So I wanna, unpack that a little bit more in terms of explosive few years, which what what you just said. So what do you think that looks like then? Because people often ask me, like, hey. It’s great to interview people about what’s happening today, but what’s happening tomorrow and beyond that? Anything you can share?

Tom Gillis [00:19:35]:
Yeah. I mean, you know, like, I think I was feeling like this as a kid. You’d you’d hear about these predictions of the future. You know, when I was a kid, I was like, oh my god. I’m gonna have a jet pack. And I’m, like, seven years old. I’m like, that thing is gonna kick ass. Like, I’m gonna get my jet pack, strap into that thing, and I shoot shoot down a school.

Tom Gillis [00:19:51]:
It’s gonna be amazing. Didn’t quite happen. Right? Never really got a jet pack. So this isn’t jetpacks. Like like, I’m looking at the components that go into these solutions, and, you know, AI is gonna allow computers to talk to you and and just and just think. And so everything from the way medicine happens, you know, like like, right now, AI agents can produce medical diagnoses better than humans. Now that doesn’t replace a doctor. That’s just the diagnosis part.

Tom Gillis [00:20:21]:
That’s part of what a doctor does. Doctor does a lot of other other things. But if we can have incredibly accurate medical diagnoses at our fingertips, what an impact that’s gonna happen. You know? And think about the, like, kind of robotics applications. We we have all the, you know, mechanical hardware and the servos and the the cameras to build a really cool robot butler. You know? I would love a butler that could do my laundry and get me beer. The pieces we’re missing is the the logical processing to figure out how to, you know, get me a beer and not get me a Diet Coke. And that’s suddenly arrived.

Tom Gillis [00:20:58]:
So I I think it’s really gonna be a transformative transformative few years.

Karissa Breen [00:21:03]:
Do you think there are many people out there that really get AI sort of to, like, the foundational level? Or do you think, like you said before, everyone’s just saying, oh, we do AI. We do this. You know, these terms that get, you know, floated around whether they do it or not is is the next question. But do you think that people, at the end of the day, can hand on their hearts, say, we thoroughly get AI?

Tom Gillis [00:21:23]:
Oh, I think it’s a continuum. You know? I think there’s some people that would be like, oh, yeah. AI is cool and don’t even really know what it stands for. You know? And then I think there are folks that, you know, like, someone built this stuff. Right? So and then there’s a whole spectrum in between. And and I think as a leader, I feel this way. I know my boss at Cisco feels this way. It’s like, you know, leaders set the tone.

Tom Gillis [00:21:44]:
And leaders need to say, like, hey. This this is not a small thing. Understand the magnitude of of the change that’s possible and then drive the organization to embrace that change. Ride that wave. Be on the right side of the problem. Yeah. That’s exciting.

Karissa Breen [00:21:58]:
So what tone are you setting, Tom, as a leader?

Tom Gillis [00:22:01]:
Hopefully, you’re just you infer that. Right? Like, things that weren’t possible two years ago are suddenly possible. I wanna be the company that that brings that stuff forward. And there’s a double whammy here in that that in the data center, I believe that we can use AI to make infrastructure that actually powers all these AI based applications. So as this AI revolution takes place, you know, I really firmly believe that Cisco can be an engine that drives it.

Karissa Breen [00:22:28]:
So I wanna get back to we spoke before around, you know, people talking about, you know, a very real fear of them worried about, you know, AI and what it can do. So then, as you know, you probably heard people talking about, like, you know, cybercriminals are using AI, then we’re defending with AI. How do you think we’ll ever properly get ahead? We’ve always sort of been behind, but even if we’re not permanently there, what does that sort of conundrum now look like?

Tom Gillis [00:22:54]:
Yeah. This is this debate goes on a lot. You know, the explosion in AI capabilities is actually gonna help the defenders more than the attackers. And the reason for that is, like, there’s no question AI can make that super scary email that looks just like a real email from a real person, but people were clicking on the goofy emails anyway. You know? So it doesn’t really make the problem that much worse. Whereas on the defense side, you know, having an a a firewall that can write its own rules, that can look at these anomalies and automatically take action, that’s net new capability that didn’t exist in the industry. So I think it’s gonna advantage the defender more than the the attackers. So I think it’s kinda overall good news, but, you know, there needs to be seen.

Karissa Breen [00:23:41]:
So with that being said, would you then envision that we will stay ahead then in terms of the the criminals, or do you think it’ll sort of oscillate?

Tom Gillis [00:23:49]:
I think that we are in a good position right now, and we’re gonna get in a stronger position going forward. You know? Yes. The attack the threats are, you know, more sophisticated, but the our ability to create intelligent response with systems that used to be quite brittle, I think that’s a big step forward.

Karissa Breen [00:24:06]:
What do you mean by brittle?

Tom Gillis [00:24:07]:
Well, you know, a lot of infrastructure. And, again, pick our example firewalls. Like, you set a bunch of rules in the firewall, and the the things that they’re runs with those rules. So it’s hard to update them. You know, you update them very carefully. It takes weeks to update them. So, you know, in a world with where you’ve got these kind of dynamic threats that are coming in and changing all the time, it could be hard for traditional systems to to keep track. Or like that application that I talked about, you know, where where you got an AI model in it and you’re looking only for a credit card number, you’re not playing the game of of hundred questions.

Tom Gillis [00:24:37]:
You’re gonna miss those types of things. And so so with AI powered security controls, now all of a sudden we can see it.

Karissa Breen [00:24:44]:
So I wanna flip over now and talk about, you know, security and networking together. Now I know that you guys are doing a lot of work in this space, but I wanna get into your mind about your experience with this. What does this look like? What are your thoughts? And it comes to mind because I think this is really interesting in terms of, you know, the more holistic approach.

Tom Gillis [00:25:04]:
Yes. For a long time, there’s been a lot of interest in kind of the intersection between security and networking. And, you know, I think that there’s two things. One is these AI management tools. The second is advances in Silicon itself that are enabling an absolute, you know, sort of acceleration of security and networking coming together. So I talked about a product called HyperShield. HyperShield is is a firewall, you know, or really kind of a reimagined version of a firewall that can run-in a data center switch. It’s not a box.

Tom Gillis [00:25:41]:
It’s not a separate appliance. It’s just a feature on a a platform. That’s one of of a number of examples where we think we can make security far more effective than it was when it’s separate by integrating it into the network.

Karissa Breen [00:25:54]:
So can I just ask, why was it so separate for so long?

Tom Gillis [00:25:57]:
A lot of reasons. A lot of it was the the the kind of fundamental building blocks. So a network processor wasn’t capable. It wasn’t programmable. It wasn’t capable of doing advanced functions like a firewall. But as the silicon has progressed, now it suddenly it is.

Karissa Breen [00:26:13]:
And so then that being said, and now that things are sort of, you know, working in harmony together, what do you think then what do you think we can see then as a result of doing this? Obviously, it took a while to get to this level, but now we’re at this level with everything else that you said that infused with, you know, AI and everything else that you’ve discussed already with me today. What else can we sort of see on the horizon now?

Tom Gillis [00:26:35]:
Yes. I think the next wave of computing is clearly going to be much more distributed. So as we dream up these AI applications, we’ll do robotics and and you get in that robot tumbler that I was talking about, You have an ability to have the workload, and I’m running everywhere, all in AWS, is super, super interesting. But that’s gonna gonna require security that is infused into the fabric of the network. So we kinda gotta have security everywhere in a world where these devices, you know, are also becoming ridiculous. Right? In a car, in the infrastructure, sell to a hospital environment, and the factory sort of. And the intersection of security networking, I think, is, you know, really, really, out of many ways, the only way to do that.

Karissa Breen [00:27:20]:
So I like the word infuse. So what do you think businesses that, you know, don’t have this fused infused, you know, security, you know, is part of the the fabric of the network. What do you think happens then? Do you think that these companies have a lot more problems than we know and that we hear about? And so this is gonna be the new way of how businesses have to to move, you know, in in the future?

Tom Gillis [00:27:43]:
Yeah. I think usually what happens is is someone makes a mistake. Right? They’ll be like, you know, there’ll be some sort of high profile disaster where, oh, I’ve got this, you know, AI based system that wasn’t protecting them with a distributed security system, and it led to the to, you know, some significant consequence. That mistakes usually only happen once, and then the whole industry take notes and everyone, you know, needs to embrace this type of solution. And so, you know, that’s a cycle that we’ve seen kind of over and over.

Karissa Breen [00:28:11]:
So what do you think now for 2025 excites you the most? I’ll probably see you again in another Cisco event this year, but, obviously, a few months will pass since I speak to you next. So what do you think really, you know, wakes you up in terms of the industry, what Cisco’s doing, what clients are saying? Anything you can share?

Tom Gillis [00:28:31]:
I think I’ve already talked about the, you know, significant opportunity that we have to infuse security into the network to make it part of a switch or router. I think the larger thing that I’m excited about, what what I’m thinking about is this AI revolution, this wave that is already breaking is driving a screw on the renaissance in the data center. And that’s yeah. Data centers had become kinda commodity. It wasn’t really changing that much. You know? It It was it wasn’t necessarily exciting. And, man, it’s super exciting again. And I think that, for me, I find that personally very, very engaging.

Karissa Breen [00:29:08]:
Okay. So, Tom, Cisco celebrates forty years. Is there any sort of learnings you can share from your perspective as you, you know, look on reflection?

Tom Gillis [00:29:17]:
I think the big opportunity for Cisco is to just focus on things that are naturally improved by integration with an aggregate. And I already talked about how we can really save the way people think about network security by by building it from the beginning to work in a virtual router. The same thing is happening in the data center. All this AI movement is driving really a renaissance in in data center infrastructure. And there’s a whole bunch of hard problems that have to be solved. And the network, all the way from the silicon, the processes, to the optics, to the systems, the network can be a key component of how we can make the infrastructure to power these AI workloads and and continue to drive this this revolution that’s happening all around us.

Karissa Breen [00:30:00]:
And so, Tom, really quickly, any closing comments or final thoughts you’d like to leave our audience with today as you, episode 300?

Tom Gillis [00:30:07]:
Yeah. I would I thank you for including me on episode 300. I really enjoyed it. Super enjoyed the conversation. And, yes, it’s gonna be an interesting time. I I look forward to watching it all play out in the next few months.