I'm Worried About Network Engineers... Here's What I Found

[00:03] you tell the doomsdayers who are like, "Uh, no. AI is coming for us. 50% of jobs in 5 years, it's all gone." There's a lot of people I see asleep on this. They're They're asleep on this, Chuck. They're not paying attention. Do you

[00:16] think the roles we have now with like IT engineers, network engineers will exist in 5 years? Um, I have a question that's kind of been keeping me up at night. Is there a future for network engineers in the age of AI? like in 5 years are we

[00:29] going to exist? That's why I'm here at Cisco Live, the largest networking asking this question. I need the answer. I'm asking Cisco people. I'm asking network engineers who come here, 22,000 of them, hey, what do you think about

[00:42] AI? Are you worried? Do you currently use AI? Is this a thing for you? And let me tell you, this year feels different for network engineers. Everywhere you step, everywhere you look, AI, AI, AI. The opening keynote, Chuck Robbins, AI.

[00:55] G2 Patel, AI. Everything is AI. So, here we go. Let's dive in. Okay. First, walking up to Cisco Live, the conference center. AI everywhere. Every sign, every

[01:07] advertisement, it's the AI era. We're powering AI. Every booth in the world of solutions had something to say about AI. We kind of have a reason to be a little saw a Reddit post today where someone who had just gone to Cisco Live, they're

[01:23] like, "Huh, are we going to be okay?" It's a very question I'm trying to answer dayto-day. And by the way, this video is sponsored by Cisco. They gave me the freedom to make whatever I wanted. Thank you, Cisco team. You are

[01:36] amazing. Now, my first mission was to figure out how people feel about AI. Are they optimistic? Are they scared? Um, I can tell you that I touch AI almost daily for almost everything I do for work. So, AI, how do you feel about it?

[01:50] Are you optimistic? Are you excited? Are you all in? You get all in or you get all out. How do you feel about AI? Oh, I love AI. I use it almost every day. So, about it? Oh, yeah. I think it's the future. I mean, everything's going to

[02:03] don't learn it now, you're going to get left behind. So, I'm optimistic. Great. different form than what we're used to right now, right? So, AI now, I think

[02:16] it'll actually enhance all of our experiences. Right. I'm optimistic about experiences. Right. I'm optimistic about it. I've used it for both college and personal like creating experience. So you use AI now? Yep. Um I know my

[02:29] colleague here is not so keen on it really. I'm indifferent. I mean I think it's neat for security and like medical stuff but right all the GPT and all that stuff. It's neat but I don't care. Now overall people seem to be pretty excited

[02:43] that AI is going to make their lives better because it already has. And I know for me it has in some ways and the more models that come out and the the better I understand how to use it, it does actually help me with some things

[02:56] but not everything. We'll get to that later. Then I wanted to see okay network engineers, IT people, how are they actually using AI? Are they using it with their jobs and not just with their everyday life stuff? With AI, are you

[03:08] using it right now in your job? I am. How are you using it? Mostly for uh code ciloting and stuff that I don't want to do. We actually just started about a year ago and we're actually going heavy into AI. We're a utility company, right?

[03:24] So, we're finding ways whereby AI could help us automate a lot of stuff that's, you know, done manually right now. So, I mean, we talk we talk about about this mean, we talk we talk about about this all the time. Um, AI is an assistant to

[03:39] what you do. And so, you are the gatekeeper of what you allow AI to do or gatekeeper of what you allow AI to do or not do to your infrastructure. Um we I think it's encouraged today to leverage that assistant next to you. I know you

[03:52] and I when we talked first talked about automation back what five six years ago uh we were starting the whole DevNet thing and we talked about okay so is automation going to change the way I manage my network and is it going to

[04:05] take my job and and and and here we are we're just talking about automation as a skill as something that you should adopt as part of your tool set when you're working with your network AI is the same um my company we're we're trying to use

[04:18] AI where we can um I You don't want to rely completely on AI, I'll give you for instance, I had to write a document uh really an ITIL document on that. So, okay, I go to co-pilot. C-pilot, write me a document

[04:32] policy document. So, it gives me a policy document. Okay, make these changes making these changes gets me like 60% of the way there in minutes. At that point, now I can take it and do my own thing. I wouldn't take that document

[04:46] and put it forward and said, "Here's my document." Right. Right. Don't do that. but it's 60s something percent of the way there and now it's a really easy way rather than sitting at a blank page and going what do I AI is just another tool

[04:59] um and I've been talking this week a lot about ways that network engineers can use AI today to make themselves better engineers so an example I I I will you've got some piece of knowledge buried deep in the back of your head but

[05:13] it's buried under dust and cobwebs and too many late nights with pizza um you could go try to dive through it or go into it But I like asking AI cuz I know "Yeah, that's the right answer. I can dive in." So verification of

[05:26] information, great use case for it. Now, speaking of using AI, one of my first interviews I actually had was sitting down with John Capo Biano. This guy is a ball of energy. I love him. Thank you for sitting down with me, John. He's a

[05:39] left and he's working for an AI company called Selector AI. I wanted to sit down with John because he's kind of at the forefront of networking with AI. on Twitter. He's the first guy I see messing with MCP, agent to agent, agent

[05:53] him the same question. Are you optimistic about AI? Why do you think it was that you were more excited versus like we're doomed? Well, I see it as um like the computer is for human convenience. Now, I'm old. I'm very old.

[06:08] My first computer didn't come with a mouse. Oh. The mouse was a peripheral that was screwed into the back on a serial port. And now I could move the mouse around. Okay, I know it sounds hilariously trivial and I'm really old,

[06:21] but it's about convenience, is it not? Mhm. So now instead of menus and gooies and clicking and memorizing commands and I can use my own natural intuition and just sus out questions in a conversational way. Now, should we be

[06:37] frightened? Uh, I think we should be a little frightened, right? There's with the great power comes the great responsibility, but it's an opportunity. Chuck, I see this like like the protocols we just talked about MCP and

[06:50] ADA. Why they've got me so excited is because they're protocols like HTTP or because they're protocols like HTTP or SMTP or spanning tree protocols. Your code can work with my code because we're developing them with standards. It's not

[07:05] the wild west anymore. Right. And it's very much like the late 90s, mid 90s where protocols suddenly changed the world. Right. Right. Think about HTTP, hypertext transfer protocol and HTML. What did that do for the world, right?

[07:21] It brought us online. Yeah, it did. Yeah. So now these new MCP tools are going to bring us even further with artificial intelligence. So you're overall optimistic? I'm optimistic. I'm optimistic. So of course, obviously John

[07:34] works for an AI company. He's pretty optimistic. But I loved one of the last conversations I had with Hank Preston. Now hint Preston if you don't know who he is he when I first met him he was the Python network automation guy back in

[07:48] 2018209 a big shift in networking was happening we as network engineers and admins were being told that we have to become developers in fact Susie we got to the stage and said hey you guys are

[08:01] developers now and that kind of scared a lot of network engineers including myself like are we going away is our job going away guess what it But Hank was at the forefront of that. He had an amazing session that everyone flocked to. He was

[08:15] like the guy those years and he was showing us how to start using network automation. This is a key thing I want you to remember. We're going to touch on that later. But one of the last things that Hank said to me and Jeremy as we

[08:28] were talking to him was that he said, "You have to be optimistic about AI. What else are you going to do?" And that's I kind of feel like where everyone is. We're choosing to be optimistic. you have to to wake up in

[08:40] the morning and continue on with this network engineering journey. Now, do we have a reason to be optimistic? That's still what I'm trying to figure out. The people worried? So, they may be optimistic about AI, but are they

[08:54] worried about network engineering? Are they worried about their own jobs? No, I networks were not what they I mean, licensing what isn't what it is today. progressed since then. It's the same thing. I I think a network engineer

[09:08] that's saying like well I just you know I I build switches I plug that's not an engineer it's a jockey um that whole that whole job has changed over the you know you look at catalyst center you can deploy switches very quickly. So

[09:24] using uh using the tools in catalyst center to deploy how many switches right away. Yeah that's true. So, so the job itself has changed and I think everyone understands that and they need to you know learn the new stuff and and move

[09:38] forward with that. So one thing about AI is yes there might be job loss but also creates opportunity for you to learn new stuff and you know remain relevant. So I'm not worried about that at all. I actually embrace it. I think it's going

[09:52] actually embrace it. I think it's going to be overall net good for everybody. Yeah. you know, do you what's your take on AI as far as the job market goes? Do you think it's kind of scary like we're going to start losing jobs or what do

[10:05] you think? I would say like 25% jobs of the current form will go away. New jobs have come but it will not like gap all the beaches bridges but it will come different things. So like as a current technology like here the tech expert I

[10:22] familiar with it like the way we get oursel familiar with apps smartphones we should get familiar with it and get on top of it and take it like another cuisine right you are just trying out another cuisine just don't be afraid of

[10:36] it right so I don't think that networking as a career is going to diminish I think it's going to become more and more specialized more and more specialized to support you know high performance computing or GPUs or large

[10:51] performance computing or GPUs or large language models in the data center. Um I think we still have 3 to 5 years of a safety net. I don't think that it's going to be that revolutionary and that fast. Beyond 3 to 5 years it's really

[11:03] hard to project or predict. I think the time for skepticism is behind us. I think that everyone who is using AI has an advantage over people who are not using AI. Okay. Right. And as we have this conversation, if you're not using

[11:16] AI, there are people are there are other people using it, right? And there's going to be a parody gap here. And that is what I'm worried about are the people paying attention, that are going to wake up one day to a whole new reality that

[11:31] they're not prepared for. Um, at least today and and it some of this certainly will change as AI models get better. Um, but right now AI is not truly creative. AI are mathematical models that are building on patterns and finding

[11:46] good information and it's fantastic at that. And so one of the things that I think true humans are still necessary is for innovative ideas. Um I'm also I try to be optimistic right the world is not for machines. The world is for people

[12:01] and I think that we will recognize that there is still value in having people involved in the work not just the machines that are there. Again, is going to be okay. Many saying that network engineering will become even

[12:14] more critical as AI becomes more prevalent. More data centers means more network engineers making those data centers work. I like that viewpoint. So, point, I'm feeling pretty good. I've talked to people at Cisco. I've talked

[12:26] to engineers. I've talked to engineers who love using AI, and they're all feeling pretty good about it for now. And that was my next question because right now AI is moving at a pace I can't keep up with. Can you keep up with it?

[12:39] fact, just this week and I haven't looked at it yet or even played with it, we had a new model 03 Pro come out. So my question was in 5 years is the job of network engineer going to exist or system engineer, IT engineer because in

[12:54] 5 years a lot can change. I think they're going to change. They've been changing. Uh I think you're still going to have network engineers. uh obviously you know the roles are going to be there but they're going to change because we

[13:07] have to change with the times you know uh you look at uh you look at cyber security you look at uh all the problems that are out there the bad guys are also using AI so they're getting quicker so we have to worry about that as well it

[13:19] it it's going to be different for sure yeah what's like uniquely human that you can do just that kind of thinking what's uniquely engineer right so you know you actually understand how networks network, right? Right. I mean, we all

[13:33] know, we've all seen networks where you're looking at that and like you're like, who built that? Do do you understand? It was AI. AI built it. Yeah. You know, so you know, there will always be the need to be able to have a

[13:48] deeper understanding of how the network works. Do you think the role of network engineer as we know it right now, will that exist in 5 years? It will. We definitely. especially network engineers, you know, AI runs on the

[14:02] network, so you're going to have to have the network there. So, I just see it if anything, we're but we're never going to go away. We're here to stay. It's more of AI actually assisting actually seeing more visibility of the network

[14:15] and what's going on and give us more instead of working, you know, 12, 16, 18 hour days, it actually allow us to work eight hour days. Do you guys think that Do you think the AI will take over networking?

[14:29] Uh, I don't think so. As it, like I said, as it currently stands now. Definitely not. In the future, it's possible. I think it's it's somewhere in the middle. I think the the mundane tasks that we uh do daytoday are going

[14:43] to go away, which means I don't need to SSH into a box. I don't need to look at way. I'm going to probably have an agent that do that for me. But we talked about this like automation whether before AI you were supposed to do that anyway

[14:59] right with automation um do I think it's going to we're still going to touch the CLI I think at some level we will um it's going to probably be augmented with AI built into the terminal which I think it's out there today where you have your

[15:13] CLI you have an agent on your terminal that's helping you with commands I don't I don't know most of the iOS XC commands and I can't tell you how many times I go out to Cisco Docs or I go out to Google to search for it. Now it's there. So,

[15:27] it's going to speed up. It's going to it's going to make make me have superpowers um as a network engineer, but I don't think it's going to make me obsolete. I think that network engineer looks similar to the one that is today,

[15:40] but they've in uh incorporated uh more security skills because again, there's not a company I know of that wants an insecure network. And and I I would also say someone that understands it not just from a network security standpoint, but

[15:54] a cyber security standpoint, understands red teaming, understands pen testing, understands blue teaming skills like incident response and threat hunting. I also think that that's someone that natively understands automation. I I I

[16:08] think that, you know, everybody thinks about AI almost like it's an easy button. It's not. If you really want to get the benefits of AI, if you really have to you have to automate more of your network uh operations and

[16:22] activities and that's going to be a big part of I think any uh CCNA, CCNP, CCIE engineer and what they're doing. They're going to be thinking about how they can really drive that automation to get the benefits of AI. Otherwise, when I hear

[16:38] of people talk about, oh, it's all hype. It's all hype until you really automate that network activity, and I think that that's going to be a big part of the future work that our engineers are going to be doing. Now, this question, I feel

[16:50] more practical for me. For example, Clark, he said, "We'll have 8 hour days network engineers are normally faced with. Waking up and and getting those calls. Maybe AI removes being on call. Wouldn't that be amazing?" Karine said

[17:05] that AI is going to give us superpowers. It will augment us and make us better. It'll do the things that we don't want to do. And Josh reiterated that AI runs on the network. The network powers it. The more AI there is, the more AI needs

[17:18] us. Now, my take on that, I think it's optimistic and we have to be optimistic. It's really hard to say what's going to happen in 5 years. If we're honest with ourselves, we will all say that we were all surprised where AI is at this

[17:31] I feel like we're going to be saying something very similar in five years. And then I asked one more fun question. It was a meme I saw on X or Twitter. I call it still? Which side of the bus are you on? And this assumes the premise

[17:46] that AI will eventually take half of all jobs. Will you be on the left side where you're saddened by the idea of AI taking jobs? You don't want it to happen. You Or are you on the right side of the bus? Will you be overjoyed by the idea that

[18:01] AI takes jobs? Maybe we enter a postj job world. Sure, there's problems to solve. Oh, a lot more problems to solve with that scenario. Where do you sit? Uh, I'm the girl standing in the middle

[18:14] going, it could do that and it could do that. My jury's still out, if you'll pardon the pun. I'm somewhere in the middle. Somewhere in the middle. Okay. Lots of jobs be happy to hand off to AI. I'm on the AI's going to take all our

[18:28] jobs, right? Okay. And I'm and I am an outlier, right? Like I I have my own outlier, right? Like I I have my own unique uh take on this, but it is it to me it being AI and large language models and natural language and chatting with

[18:42] the computer. It it is like the promise of the last 30 40 years. I feel like we've been on this trajectory and we're sort of reaching the operand of Mo's law

[18:54] and some of these right we're really starting to maximize our compute power and I I again I believe in a utopian outcome now whether or not we it's going to cost a lot of jobs I think that as yes some jobs will be displaced but you

[19:11] know two years ago there wasn't such a thing as a machine learning engineer right so new emerging categories of jobs will will come out. Uh I would say for me somewhere in the

[19:24] I would say for me somewhere in the middle um maybe towards the the left a little bit where is it is AI going to destroy half of all the jobs? I don't think so. And I I I want to say this and this is something that we we we overlook

[19:38] this is something that we we we overlook often which is humans engineers um network engineers have this thing that AI doesn't have which is critical thinking and so you built your critical thinking uh

[19:53] skills from learning from uh troubleshooting networks from experience right that AI does does not have the capability to do that critical thinking skills. Now, I think I'm somewhere in the middle as well. There are jobs in

[20:09] history that don't exist anymore and should not exist. Like lamplighters. of Anthropic used when he was talking about, "Hey, wake up. Jobs might be taken. In 2 to 3 years, we might see 50% of jobs gone." But also, I think that we

[20:25] need to work. People need to work. People need purpose. We should never remove people from these jobs. AI should always be there to help us but never replace us. And that's a common theme I heard many many times throughout talking

[20:38] with everyone in all these interviews is that AI isn't going to replace you but someone using AI will. But then John Capo Biano and he comes in with the craziest notion and many people do hold this view. He was fully on the right

[20:52] happened. The opening keynote at Cisco crazy stuff they're doing. And shocker, they talked about AI a lot. Chuck Robbins opened with saying that the AI race is on. We have to move forward. We

[21:06] have to be quick because everyone is trying to have the best AI. We have to best. We all know that every company, every organization, every government defensive mechanism. There's a FOMO element about what are you doing in AI?

[21:21] And I agree. There's no stopping it. The only thing we can do now is sprint forward as fast as we can. I get that. And then G2 Patel took the stage. One thing he said stood out to me. With the help of agents or AI, we can take the 8

[21:36] them in a way to where we have a throughput, a workforce throughput of 80 billion people. So 8 billion humans will feel like we have the throughput feel like we have the throughput capacity of 80 billion humans. He then

[21:50] a different looking workforce. We'll have humans of course maybe but we'll also have AI agents, AI apps, robotics and humanoids. That's terrifying, but

[22:02] announced something pretty cool and also pretty terrifying. They announced their own purpose-built AI model for networking. It's called the deep network model. It's been trained on a ton of networking related data and tasks. And

[22:16] similar to how they benchmark other AI LLMs on coding task, this one they gave it a CCIE test labs and exam and it performed 20% better than the other announced something similar at RSA conference, but it was security focused.

[22:31] foundation model. I did try it out. It was okay. But then Cisco showed us something pretty crazy. They're putting AI in the data center and they're doing it with a thing called AI canvas. It's a new control panel that's all about AI.

[22:45] which this guy was troubleshooting a connectivity error and this demo was live real time and he simply just chatted with this interface which I believe was the deep network model and it started doing things that man humans

[22:58] just can't do. It's multimodal so it was generating graphs and charts about the network in real time. Graphs and charts that we could drag on over to the main canvas real time dashboard creation which honestly is amazing. But then they

[23:13] showed one more terrifying thing. They had the model figure out a configuration like produce a QoS policy or a cost policy for me quality of service and then he asked it to apply this policy. So at this point this is something a

[23:27] network engineer would have done. We would have gotten into the routers, the switches, wherever this QoS policy would need to be applied. We would have troubleshooted it, figured out the configuration and applied it. The AI did

[23:39] this for the guy operating this. I forgot his name. And all he had to do was just tell it his intent. What do you wanted to happen in regular human language, not CLI speak like we spend so long learning when we're studying for

[23:51] the CCNA and CCMP, which brought me to a conversation I had with Kareem Escander team that actually brought me out to Cisco. Thank you, Kareem. How do you they're using it right? How do you have a CCNA exam that's measuring if they're

[24:06] troubleshooting a network? One of the ways we we think we can do this well first of all we might have to reconsider how we evaluate exams. So in the real world, you would have Google, you would have chat GPT. Instead of doing a lot of

[24:20] multiple choice memorization like you just said, imagine if your exam allows you access to an LLM, allows you access to a search and instead of multiple choice, you're dropped onto a web portal for Moroi or you're dropped into a CLI

[24:35] when it says implement, you have to configure, you have to implement. The second part of your question is towards how do you get them to think critically about what AI is coming back with. We think there's an opportunity to present

[24:50] them with um artifacts like this is the topology diagram of the network. These are the configurations of the network. This is the prompt and this is what the LLM said. What do you think is the right answer? What do you think is the most

[25:03] accurate response the AI has given you? And if they've studied, if they've prepared, if they've played with some of that LLM with something like Cisco modeling labs in a simulated environment, they should be able to

[25:15] answer that. They should be able to say, "Ah, you can't do that interface there." Or it says, "We're not allowed to run this protocol in our network, so I can't trust that answer. It has to be this answer." So, what we're seeing right now

[25:27] is people say, "You need to learn AI. You still need to learn the CCNA, but you need to learn all the new tools." A big focus that Joe mentioned was an AI. So when you're talking with an LLM and being able to see if it's giving

[25:40] you bad data or it's hallucinating. Okay, so let's close this up. I came to Cisco Live to figure out is there a future for network engineers this year. We had an attendance of 22,000 people. 22,000 people that care about networking

[25:56] that were there to learn about networking to kind of get a forecast for learning? The overall take of many people I talked with was they're optimistic. They're using AI every day in their job and they're actually

[26:10] their lives easier, their jobs easier. Some are honest and saying, "Yeah, it might take away some jobs, but it won't take away the jobs of people who adapt, who learn, and use AI in their current roles." Now, I know in many roles I've

[26:24] more people, but we couldn't find them. So, I totally understand the appeal of having an AI buddy to augment your staff. I just hope it doesn't go too far to where they realize suddenly they don't need me either. But I will say

[26:37] this, I was so encouraged being at this conference because what I saw was still a strong desire to learn networking. People are still tons of people are still going for their CCNA, their CCMP. Everyone's excited about learning that

[26:52] about learning the basics. IP addressing, subnetting, OPF, EIGRP. For the time being, no matter how good AI gets, we need people who know those protocols so we can tell AI when it's being dumb. But reality check, real

[27:07] change is coming and we need to be ready to adapt. You need to adapt now. If you haven't already started playing with AI, you need to become very familiar with for yourself right now is just ask yourself, how can AI help me do my

[27:20] current job better? Also, get in community. Talk with your fellow engineers and see what they're doing. But also take comfort in this. This is no different than anything we've ever encountered in it. We've always had

[27:33] to learn and adapt to the new technologies that come up. The people who don't adapt, they do get left behind. We had traditional telefan migrate to voiceover IP. We saw the big change with network automation, going

[27:45] from manually configuring a switch a switch a switch with CLI to configuring thousands all at once with automation. One might say that AI is simply another iteration, the next generation of automation. And as far as what network

[28:00] engineers will be doing in the future, I think that's probably it. It will not be getting in the weeds as much as we are right now, dealing with outages and and looking through logs. It'll be how can I make the network better? How can I

[28:14] automate more of the stuff we already have? And in 5 years, who knows? You can't know. Be optimistic. Keep studying. And from what we can see, the world needs network engineers. And we have a shortage and more AI is being

[28:28] I want to know what you think. The questions I asked, I would love to have those put in the comments below. That's all I got. I'll catch you guys next