— Mike Bushong Live From AutoCon 5

SPEAKER_00 0:16

This is the art of network engineering. Where technology meets the human side of IT. Whether you're scaling networks, solving problems, or shaping your career, we've got the insights, stories, and tips to keep you ahead in the ever-evolving world of networking.

andy lapteff 0:30

Welcome to the Art of Network Engineering podcast. My name is Andy Laptop. You wouldn't know it by the background, but I am in Munich, Germany with this guy. You remember this guy? This is Mike Bushong. Hi, Mike.

SPEAKER_02 0:40

Twelve beer here.

andy lapteff 0:42

You're such a show-off. I don't know how to say anything in German. Of course it does. That's all I've got. Is this the first in this might be the first in-person podcast we've done? I'm going to say it is. I don't think we've ever done this in person.

SPEAKER_02 0:56

Usually we're singing. Oh, yeah.

andy lapteff 0:57

Are you going to sing?

SPEAKER_02 0:58

Happy to sing.

andy lapteff 1:00

Uh, what would be appropriate for Autocon 5 in Munich? Do you know anything in German? Any Rammstein?

SPEAKER_02 1:07

We could do 99 lift balloons.

andy lapteff 1:09

Yeah, that'd be good. Little Nana.

— Why Hasn't Network Automation Taken Over?

andy lapteff 1:11

I know we have limited time, so I'm just going to get into it. Where are we? We're at Autocon. What are we talking about? Network automation. How does that relate to network engineers? If you look at the data that NAF collects, I don't know if it's 60 or 70% of most networks aren't meaningfully automated in any way. Does that sound semi-accurate?

SPEAKER_02 1:29

Yeah. So most people look at the full market and they say that uh up to 37% of companies are or have mature automation practices, which means a full two-thirds or slightly under two-thirds do not have mature practices. Even if you look at mature practices, the question is how much of their operations are automated? Um in there, you're typically looking at less than 50%. So if you were to take out the high flyers in our industry, the Googles, the Microsofts, the Amazons, the top half of like the neo-clouds and SaaS companies, if you take all of those out, then we're left with a number that's shockingly small, especially given that we're at Autocon and everyone here is either a practitioner or an enthusiast, or in some way is trying to make money off of the practitioners and enthusiasts.

andy lapteff 2:13

Now I know I didn't like or care for network automation prior to it impacting me. And I don't know if those dots are necessarily connected, but it seemed like years of publicly declaring that network automation wasn't for me, didn't seem like it was great for my career. One of the biggest things we've always done on the show, but what I've been focused on the past year and a half is how can we help network engineers not only get into the business, to the industry, but how can they stay in? I think in tech it's always learning the new thing or being skilled up enough to do it. And I was just speaking to somebody yesterday about this. You know, I thought once I got into networking and became a network engineer, I was done. And I don't know where that came from, but like I got the thing. I have the experience. Like this job is hard enough. I shouldn't have to learn other technology stacks like software dev and automation and stuff. So the reason I'm mentioning it here, and it's someone in our Discord server the other day, I don't know if it was a complaint or a joke, but they were like, is this the art of network automation podcast now? Because I guess I've been talking about it more. What's behind that is A, I don't want what happened to me, like not being in the thing. You've been doing this. How long have you been interested in network automation? 10 years, 20 years? Like you were in early.

SPEAKER_02 3:29

It's like 20 years. I've been doing um automation product management since probably end of 2006, early 2007. So I've been in it like been a long time. Yeah, it's been a

— Can Network Engineers Keep Ignoring Automation?

SPEAKER_02 3:39

minute.

andy lapteff 3:39

It seems like all those folks that haven't automated, that aren't embracing it, that are hiding out, are like I was. And it would be easy to go, well, this doesn't apply to me. Look, Mike's been doing this 20 years. We still have crap adoption rates, so who cares, right? I'm gonna hide for the next 20 years or whatever, not have to learn this and just keep my head down. What I wanted to ask you, I guess, and I don't know what insight you might have, but do you think we might be reaching uh reaching a what's the right term, not a tipping point? Uh is there an inflection that you see in the market and in our industry that for those two-thirds of people, give or take, that aren't automating anything. Can they hide for another five, 10, 15 years, or do you see headwinds of change that could disrupt not their job, like their career?

SPEAKER_02 4:24

I guess I come at this a little bit differently. There's still a large population of people that program COBOL. I just I think in general, when we predict the demise of industries, we tend to overstate how quickly things are gonna fall off. And then we tend to exaggerate a bit on kind of what the overall impact is gonna be. This happens with almost any technology wave, by the way. When people come in and they look at AI, you know, what we'll do is we'll grossly overestimate what's gonna happen over two or three years. We'll tend to underestimate what's gonna happen over like an eight or 10

— We Overestimate Technology in the Short Term

SPEAKER_02 4:52

year period. And so generally speaking, we're gonna overstate in the short term and we'll understate in the long term. Is there a point at which you know network engineering um you you can't survive if you're not automating? I think the answer is yes. Uh, I think for me, there's there's a couple of different skills. Uh, network engineering generally, if you kind of put the emphasis on network and engineering, it's really how do I create the infrastructure required, both architecturally and then operationally, to go in and connect things together, right? That's kind of the practice. And and I think that no matter what happens, whether it's you remember, remember when cloud came out, there was a huge prediction that everything would go public cloud, that you would then wipe out all of the architects as well, right? The network engineering job would go away. That ended up proving not to be true, right? One, just because you're hosting in the cloud doesn't mean that architecture is not a thing that has to be done. And then for those companies that abdicated, abdicated full control over their environments, what they found out was that they were in a poor position to even survive inside cloud when there was a you know a major outage if they didn't necessarily have like dual availability zones as an example and they hadn't thought through the implications of that, then they were out of position. And then those companies were adversely impacted by what really should have been kind of a failover type environment. I think the same thing is going to be true in operations. I I think automation is important. I'm not trying to diminish the impact. Obviously, I'm at Autocon here and delivering the opening keynote, so it better go well. I'm very, very pro-automation. I have been forever, but I just I think there's a set of things that have to happen, whether they're automated or not, around you know architecture and and kind of setting things up. The second thing is there's uh bifurcation in in what the the companies do, right? So I think there's a set of companies that are gonna exist for a long time doing you know what it is they do. And so I'll give you like an example. In a previous job, I was engaged with a company. I'll leave the name of the company out just because I don't want to, I don't want to drive any business. No, I mean uh, but they they like literally this was a company that they they they took rocks, like they created they when people need construction, they were the ones that got the rocks, you know, dug the rocks out and shipped the rocks. And so they had built like a small little data center. And like, do I think that that environment, you know, is is gonna go the way of automation? I don't. It's not part of their core business. They're they're not, even if they grow the business, they're not growing their IT needs. And so I think there's gonna be a bunch of companies that are like that. And so you're what you're gonna see is there's a pocket of you know, network engineering roles that that you know will be lightly changed, if at all, for a relatively long period

— Automation, AI, and Creating Career Optionality

SPEAKER_02 7:16

of time. Now, when we say that that the future involves automation, what we're really saying is that when things that when things grow, when there's a net new addition of you know, new people doing new things across new, you know, infrastructure properties, what we're saying there is that those are gonna, that new practices are gonna heavily favor you know the more modern stuff. I think that's true. And so uh the way I would look at it is I think we're at this point right now where technology is moving us forward. What we're capable of today is different than what we were capable of, let's say two decades ago when I started. I think with cloud and with AI, there's probably more dollars rolling this direction than there was maybe even five years ago. I think what AI is doing is it's a lot of people are looking to AI all the things. AI from a network engineering perspective is, you know, largely a proxy for operations and automation and whatever. And so I think there's this moment now where growth is going to come this direction. And I think the way I would think about it if I'm an individual is there's a lottery, okay? And you have a chance to purchase lottery tickets. And the thing that's different in this case is that you can actually, you know, purchase more and more lottery tickets. You have some control over your odds of winning. It's not that you buy one ticket and and they sell, you know, a hundred million tickets. It's that you can actually, you know, kind of participate in this and and you can control the outcome to some extent. And so for me, I think it's optionality. And so if you're at a point in your career where you don't care about optionality, what you care about is stability and you want to ride out the last seven years of your career, then you know what? I mean, by all means, go for it. Is there some risk? Sure, there's always some risk. But do I think that you're gonna be left without a career? I I don't, because even good automation is gonna be predicated on do you know the underlying engineering principles? Like architecture still matters, knowledge still matters. And I think, you know, the more that we get into like vibe coding and people sort of trying to abstract everything away, the the people who actually do know the details underneath, they're more valuable, not less valuable in that environment. Because things do go wrong. Someone still has to know how these things work. But if you're someone who's like earlier in your career and you're looking at, you know, where's the proverbial puck headed, then yeah, I think understanding these types of technologies, you know, we had obviously with like John Capobianco, we talked about MCP servers, you know, some number of months ago. You know, I think there's a lot that's going on in in sort of the AI space. You've got someone else keynoting this week about spec-driven automation or spec-driven development. I think if you get involved in those types of things, I just think it creates options. And so if I'm managing my career and I'm early in my career, then I'm like, look, I want to create some number of options. Now, do I need to create infinite options? I don't. So I don't need to follow everything, I don't need to chase everything. But do I want to chase nothing? Nah, you know what? I probably want to I want to chase something. And this is this is it, right? And I'll I'll give you, and I know it's a long freaking monologue here, but this is a podcast.

andy lapteff 10:04

This is helpful. Thank you. Keep going.

— You Don't Have to Learn Everything

SPEAKER_02 10:06

So, like, look, I sit in business type roles, right? I I I have to figure out where we're going. And so, like, the the the analog for me, right? Like, there's we're going through crazy supply chain shenanigans right now, right? I mean, like, I don't want to turn this into a supply chain podcast or anything, but there's like crazy stuff that's going on around parts availability and wafer capacity and you know, the golden screw, this, and and you know, how do you do supplier diversification and what are all the strategies? Like, this is an area where, like, do I am I am I personally like interested in supply chain? I'm not. Like, I think it's like it's not like the thing I I want to go do, but it's such an important edge right now that I have become more than conversational about it. Like I've learned enough to go do some damage. And then what's happened is in learning that that's created optionality. I'm more valuable in some tasks than I otherwise would be. And so that creates opportunity for me personally and opportunity for my employers, by the way. So if I look and say, like, is the, you know, what's the in in automation, right? Or in network engineering, I should say, does automation create more optionality? I think it does. I do. And then it's just people have to have to evaluate based on where they are in their career, do they want to partake in that optionality or not? Either one's fine. And then within that, are there like specialties or there are like you know, subtopics that they want to kind of major in? Understanding that knowing a little bit about everything makes you marginally useful. Knowing a good amount about a narrow set of things that makes you very useful in the right context. And so you got to find what's the intersection of what you want to do, what are you interested in? And then where's the opportunity? And that intersection that requires some investment. And

— Do We Still Need Deep Technical Knowledge?

SPEAKER_02 11:45

then the question is not really, are you willing to learn or not, are you not willing to learn? It's are you willing, do you want to invest time into whatever the new thing is? And then, you know, the story we should tell is the story of my friend, I won't say his name, but the story about foosball. Like that's because to me, that's the like that's that's the question. It's like, do you want a lot of people want upside, but they're not willing to put in the effort to get the upside. And I think that that to me, that conversation, that's that's one that's worth having.

andy lapteff 12:15

I think the foosball story is a good one, and maybe we will tell it. And it kind of ties into what I was going to ask you next. The lottery tickets you talked about earlier, I think are the skills that you develop in your career, right? And what impresses me, the most impressive people I find know a little bit about a lot of things. What I'm amazed at by you is that you can answer any question. I never see you stumped. You seem to know a little bit about a lot of different things. And one of the things I wanted to ask you was, you know, do we still need deep knowledge? Because there's a whole like, you know, the world is headed toward generalists. And there's CCIEs out there that are really pissed off, I think, at the commoditization of intelligence. Right? Like,

— Why AI Could Make Deep Expertise More Valuable

andy lapteff 13:56

yes, maybe you need networking experts, but can you ask an LLM and get an 85%? Like, I do we need deep knowledge? I guess we need deep knowledge. You said it earlier. Like there are going to be weird things that happen and we're going to need experts to dive in and figure out the weird thing that happened.

SPEAKER_02 14:12

What's going to happen? So when access to information becomes commoditized, when it becomes easier to get access to a lot of surface level information, then what people do is they invest less in learning. If I can get easy access, if I can use, you know, a prompt and I can get some meaningful information back. And I know people will talk about hallucinations and whatever else, but like let's be clear, AI is improving. And if you're still talking hallucinations, it's not to say they don't happen, but they're not happening like they were two years ago. And so some people they had like an early experience, something happened, they cling to that experience, that experience becomes their their norm, another norm story. But but it becomes like their norm, and then they cling to that even if the the time passes them by. I would say, like, look, if information is relatively easy to access, then then people are gonna they'll somewhat paradoxically, they don't invest more time in information. They say, I can get it when I need it. And so I they invest less time in in getting information. And when that happens, then then what you end up with is this kind of a barbell distribution. Um, for the non-uh stats geeks, a barbell distribution just says that you get a separation of people. So you get a one set of people who are, you know, say very shallow but know a lot of stuff, and then you get like nothing in the middle, and then you get a set of people who are like really, really deep in their knowledge. And what's gonna happen is the number of really, really deep people, it will diminish over time. People will spend less time. In networking, that's gonna be because of a couple of reasons. One, um, there's a lot more that's happening up stack, and so the people will look and they'll chase the AI dollars as an example. And so then you get fewer people specializing in networking. The second thing is that, you know, if the networking knowledge is codified or or or captured in tools that do it for you, then you don't you let the tool do it. And so you start to trust less. This would be like your typical car example where you know most people aren't driving manual transmissions anymore. Most of our kids are will never learn how to drive a stick. I can't drive a stick, just to be clear. But as that happens, what you get is fewer people kind of diving in. And so then the the the diminishing of the diminishing pool, the question is, is there a greater demand on those folks? And I would argue the answer will be yes. There are people, like even you know, like across the board, there are gonna be there will be companies that require that kind of knowledge. And so if you're deeper in knowledge, you will be in demand. You've got to find those positions, but you will that you will be rewarded for that. And so I think even like if you take the certified engineer path, if there's a smaller number of certified network engineers as people move towards, you know, cloud and and sort of a different set of tools, more let's say the CICD pipelining type tools, there will be a market for the certified engineer. Are there going to be tons of jobs there? You know, though I would say in absolute terms, they'll probably there'll be fewer over time. Um, because most of the growth will be, will be in other areas, but the ones that are there, and there's still people making bank off of uh fiber channel. I mean, like those jobs exist. So, and my my point in all of this, just to be, I guess, fully transparent, I think we need to temper the industry discussion about, you know, do this or you're out of a career. I I don't think that's what moves us forward. I think what we have to do is we've got to be somewhat pragmatic about it. We have to understand that a lot of the conclusions we reach are absolute, but the market doesn't deal with absolutes in the same way that we do as people. I think we gotta be somewhat pragmatic about people what people are doing. And then instead of of yanking the wheel left and right and forcing corrections, I think we got to talk to people like how things actually are. Because when we give advice, whether it's from a podcast or from you know some um lectern or whether it's from a a uh a blog post or or whatever, when we give advice, I mean, people will listen. And what's happening right now is the advice is becoming more binary and more urgent because it it stands out in a sea of AI drivel. And so what we gotta do is we gotta like give real advice because people are actually making decisions on this. And if you can be a lone voice of clarity, then maybe you save a couple of people not going down the wrong path, and maybe you get them to where they where they wanna be, which is a variable anyway.

andy lapteff 18:12

So this is something we talked about earlier, and I think it's in your keynote, but change is super hard,

— Are We Scaring Engineers Into Learning?

andy lapteff 18:18

right? Just individually. And I've probably connected some dots just from my own experiences and my own, you know, biases. And then I think, uh-oh, well, this thing happened to me, so I gotta tell all my people, my community, like, listen, you know, skill up before you and I have said that, and in hindsight and and with your feedback too, I don't know if I've been too binary about it too. Um, you know, my experience, I tried to use it as a cautionary tale, but because it was so emotionally charged for me and was such a experience in my life, I wanted to let people know because two-thirds, all the data that you and the the analysts share about the lack of automation. I that's my experience, my peers, all the people I talk to, the people working, like that's and the other part of the narrative I think I connected to my layoff unemployment that may or may not have been related to screw automation dialogue of my past was the massive disruption that could occur as a result of AI. I think we're spending a bajillion dollars to get to AGI, which would render, I think, all of us kind of right. If there's a thing smarter than all the humans combined, I don't know what we do. But in my mind, I kind of conflated the AGI narrative that we'll all be out of a job. I think I related AI to automation, like that they're related and they might not be. But in my mind, AI is like automation on steroids. And if they're talking about massive global job disruption full stop, and I'm looking out for my network engineer people, in my mind, I'm like, everything you've said, like you gotta learn some kind of automation. Like if I'm hiring, let's say I'm running a networking shop and I got two people, one of them is like old Andy, like, nope, I don't care. It's not for me. This isn't what I went to school for. And then I got someone else who's at least curious. Maybe they've run some Ansible playbooks. They've tried Python, they've stumbled, they got a thing working, maybe, you know, they're dabbling in a couple LLMs, maybe they have an account with one of them. Why would I choose what today would be looked at as a dinosaur if they have equal knowledge, like the basis? Oh, I but you know, we both have this certification, we both have two years experience. This person is like, I want nothing to do with automation. And this person is like, eh, I'm open to it. I don't know why you would hire that person, which I think is two-thirds of the people working right now. To your point, there's plenty of jobs for them. I got um friends in like the public sector, they're like, this is never going to be automated. I'll be safe. I'm gonna retire. So I guess all that babble is just to say that hopefully I didn't oversell and hurt the conversation by being too binary, right?

SPEAKER_02 20:49

There's a couple of ways to frame it. The base question is does investing in you know time in your go

— The Problem With Fear-Based Career Advice

SPEAKER_02 20:55

forward technology, whether it's AI or automation or you know, insert, whatever the the current thing is, is that an investment worth making? And I would say generally speaking, those investments create opportunity, right? I I that's why I use the lottery ticket analogy. There's two ways to frame it. So one way is that that creates optionality, it gives you a potential path forward. Another way is that it creates or it it it's required, otherwise that you, your, your career is over. And I think what I'm trying to do is is remove the downside rhetoric. I just I don't think that you motivate a population to go forward if you're constantly reminding them that they're at whatever wrong on the totem pole, they're always at risk, they're they're gonna be, you know, out of a job, they're gonna be out of a career, they're gonna be replaced by a script. I mean, these are all like it used to be a joke that, like, oh, you're you know, one day you're gonna be re replaced by a shell script. I mean, like ad sysadmins used to say that kind of stuff. I just I think when you put that kind of fear into a community, I think what you create is irrational outcomes. And what I'm trying to get people to do is like we should be have an open conversation. I've never done any of my best thinking under duress. Like if I'm pressed, if you if you invoke fight or flight in me personally, so I'm reasonably well spoken. I I handle myself well in debates and arguments, but I still struggle when someone comes at me and it feels like they have got a lot of information and then fight or flight kicks in. I can't always get to rational arguments and and then it it frustrates me and I'll feel my fists clench. I mean, like I will actually like clench up and then I'll make mistakes. And so I think if you put if you take that fight or flight and imagine like invoking fight or flight, not on an individual level, but doing it on a on a like an industry level. And then you say, well, how many people have the the patience, the presence of mind, the self-awareness to push that down, suppress it, and then to act rationally. I'm like, I what is that number? Is it is it 10%? Is it 20%? It's not 80%. And so what we do is we put this entire industry under duress, and then I think that everyone just I think it slows everything down because now you have people subverting, and you've got it's just like I mean, this has been 20 years of me looking at automation and saying uh I was reasonably convinced that the thing that I was pushing was the right thing, and then people were not picking it up. What if the actual problem we have to solve isn't what's the next tool? What if the actual problem Have to solve is how do you make it more accommodating? How do you meet people where they are? How do you get them to set down their cognitive biases or or just their outright objections? Like to me, like that's the thing we have to go do. And so with your audience, I just I think they're all on each appear each person's on their own individual path. And there is no over-the-top, like again, at a podcast level, there's lots of over-the-top proclamations we can make, but it's individuals that take advice. It's not collectives. And so as an individual, how do I talk to you and then say something that's relevant to your journey and then get you past whatever thing is is stopping you from progress. Like that's in my mind, that's what we have to we have to do. And I think it starts with toning down the rhetoric. I think it's okay to talk about the the future and the options and the and the possibilities. I think it's also okay to talk about the fact that you can create advantage if you look at stuff. I think that's fine. What I want to do though is is at least create a little bit of separation between you know people and and sort of the immediate fear of irrelevancy. I don't one, I don't think it's healthy, and two, I think it's probably not realistic.

andy lapteff 24:31

Have we identified the thing holding two-thirds of the industry back? And you know, planes can fly themselves,

— Is Resistance Holding Network Automation Back?

andy lapteff 24:38

cars can drive themselves. You don't need many humans at a supermarket. You can scan your own food, right? Like uh Linux servers have been automated for 20 years. If networking is the last greatest holdout in automation, then you've been doing this longer than I have. I think everything you're saying makes sense and it's right. But do we know that like so so let me let me challenge your examples, right?

SPEAKER_02 25:00

So, how many people like why if planes can fly themselves and we know they can, or even if they don't fly themselves, you could have remote pilots. Why don't we? Because people are still buying that service and people want someone to be in the cockpit, whether they're mostly on autopilot or not. When you look at people at self-checkout, like how efficient is self-checkout? It's awful. It is awful because you take a bunch of people who do this thing with this one interface, you know, one time a month or whatever, right? You know, I so I fly out of San Diego and and there's a a place, um, City Tacos. I go to the City Tacos, there's these two boards where you order yourself. There's a person sitting there, right, by like a cadre, but they they look at me and I gotta order the thing. Then it takes me like three, four minutes to just, and I'm just ordering two tacos. I out of the same damn thing every time I fly out.

andy lapteff 25:42

The faster if you just talk to her, right?

SPEAKER_02 25:43

Yeah, and it totally would be because but it's but but they they're committed. Like, yes, it's true, these things can be self-service

— Automation Isn't Automatically Better

SPEAKER_02 25:50

and they can be automated. They're rarely, you know, at least initially better. They will be over time. I'm not saying I'm not I'm not like a uh a human apologist, but it's it's not, it just isn't the case that things are always like automatically or automagically better. I just I think we gotta be, and it's that that delta and experience from what's expected or what's promised and then what actually happens. And that delta and experience is what's gonna create delay, and that delay is why I don't think that you're gonna see the precipitous decline of of the market. I don't think that's true. And so then if that's if that's not true, then how do I get people to work on a on an appropriate timeline? Now I want to use my, I want to use the foosball story because I think as I'm talking, it it it it almost sounds like I'm I'm I'm giving people permission to not act. I don't think that's the case. I think people need to take agency. I just want them to feel that they're in control of their story. And the foosball story, I think, is the to me is gonna bring the other side of it and pull them both together. So I'm gonna leave the names out to protect the innocent. But I had a coworker of mine when I was at previous gig, and this guy was a developer on the the kernel team, the OS team. And he desperately wanted to be in leadership and he would talk to me every maybe three, four weeks. We'd have a two-hour or more like discussion in my office, like after hours. And he was always like, you know, he they weren't giving him opportunity or they were setting up to fail and whatever. And and I and I gave him advice, you should learn this or you should tune that or whatever. And and he just he never seemed to take the advice. So after some amount of time, he comes in my office and I was and I said, Hey, you know, so-and-so, like you're pretty good at foosball. And his eyes lit up. I mean, this dude, he's this cat. I want to say cat because it sounds like like I'm on a radio show. This this cat um was like good at foosball, right? And and we had, and I said, like, you play a lot, right? He's like, uh, you know, I'm like, I used to beat you all the time. And he's and I'm like, I could I even beat you now? He said, probably not. Like, I agree. Like you play it, like you know you guys do your coffee morning thing, right? And you like you play at lunch, and he's like, Yeah, and I'm like, in the afternoon, you guys get coffee again. Like, you've got to play like what, like, like probably two hours a day. And he kind of

— 500 Hours of Foosball, How Many on Your Career?

SPEAKER_02 27:51

I'm like, maybe. I'm like, God, we've had that table how long? He's like, Yeah, about a year. I'm like, oh wow, it's like 50 weeks, two hours a day. I mean, what is that, 500 hours? He's like, Yeah, I'm like, you know what? I don't want you to complain to me anymore about your job until you go home and talk to your wife and you explain to her that you've spent 500 hours in the last year improving your foosball skills, and then tell her how many hours you've spent improving your career. And he looked at me like I was dead to him. He was pretty mad. But I'm like, look, man, because I'm not gonna say his name, I'm like, look, like it's the truth. You're not putting much time in. And so, my my advice here, like, I'm not saying ignore the future. I'm not. I'm I'm just saying that that you don't have to work under duress. However, on the other side of that, I mean you you should ask yourself the really difficult question. There's all kinds of ways you can expand your career. You can talk about leadership skills, you can go learn, you know, CICD pipelining and you can go learn, you know, AI or automation or or whatever. There's like all any manner of things you can go learn. The question I would have is how much of that time, uh, you know, of your time in the last year, how many hours? I mean, like, and be specific. How many hours have you put into it? And if the answer is like fewer than 20, which is uncomfortable, but probably true for most people, like have you read a book recently? Have you, I mean, like, have you tried some stuff? Have you gone out and like be honest? And if the answer is like less than 20 years, then are you expecting that you're gonna see some kind of career growth with that kind of input? And I'll tell you what I told my kids, because we have the same conversation around soccer. I said, look, you can have your expectations be here, but if you're if your effort is here, you have two choices. You can increase your effort or you can lower your expectations. But you cannot maintain that because that that uh discrepancy is unhealthy. And so for Andy, for your listeners, it's this you can choose to invest or not in your future. You could that investment can be across leadership skills, technical skills, networking skills, programming skills, like whatever. But fundamentally, right, right size your expectations to your effort. Because until you do that, then I just I don't know what you're gonna expect. And so whether that's like the current automation wave or whatever, like to me, that's just context. But the the big disconnect I see is this expectations and effort. And I think most people don't put in the effort. And that's uh not an indictment, but like uh you just if you're not gonna put the effort in, then don't raise the expectations.

andy lapteff 30:18

I'm glad you popped up a couple levels because it's not about automation. And I think that if you can invest in things you're interested in, there's

— Your Expectations Have to Match Your Effort

andy lapteff 30:25

probably a greater chance of you, you know, becoming motivated and having the discipline to do the work. I spend at least an hour a day of my life, usually late at night, wasting away on whatever, right? Like dopamine chasing on a phone, thinking I'm learning something. And that's 365 hours a year, roughly. And if I invested that in anything else that had value, because no one's I am not gaining any skills by looking at the the the shiny thing in my hand, what could I do with that? And I I know you're a sports guy, like I I I love the parallels with the physical world. I started running and investing in in that. And I've run, I don't know, 1500 miles in the past three years, and I hadn't run for a decade prior to that. But because I've been doing that thing, which is a hard thing, there's this part of the brain that I learned about the anterior cingulate cortex, I think it's called. The more hard things you do, the thing actually physically grows, becomes stronger, and allows you to do more hard things. So it's almost like if you get started, get off the foosball table, work on something that you're interested in, and then keep moving forward. And there's so many things in tech we could do, it doesn't have to be automation. And I'm not particularly in love with Python, which is probably why I'm not a Pythonista, but I really love what I'm doing with LLMs and coding. And they're using Python on the back end anyway. So I I have found something in tech that I'm passionate about that I'm working hard on, and it's actually not proper coding.

SPEAKER_02 31:48

So well, and and what what all that gives you is it gives you a little bit of perspective. And then when you have a couple of different

— Build Exposure and Let Knowledge Compound

SPEAKER_02 31:55

areas, I mean, look, we've talked about imposter syndrome in various forums, but but imposter syndrome is the idea that you measure yourself against the the aggregate of everyone else. And of course, you're not gonna know as much as the aggregate. But what you do is if you increase your surface area, like the things that you do know about, and you you kind of dip into, you get exposed to different bits. You said, like, I'm not, I'm rarely um stumped. Uh the reason I'm not rarely stumped is that I have a little bit of experience in a lot of things. And so I uh it gives me some exposure uh to a lot of different areas, and then I can double down, triple down on the areas that I think are meaningful, either because they're important in what I do, or maybe they're just really interesting in what I do, but it let then it lets me be broader in these areas. That's what you got to do. You've got to figure out like how do you, you don't you don't need to improve along every axis, but you do need to be constantly be growing. In material science, there's this idea of um there's a thing called a dendrite. A dendrite is like a crystal. And so the way that crystals form is that they, if you take like a clean surface, right, it's it's flat, but at a at a microscopic level, at a molecular level, it's not flat. There's little protrusions, and those protrusions attract more material. So imagine you got a little bump, and that bump attracts more material. So now you have a slightly bigger bump. And then that attracts even more material, so it becomes a bigger bump and a bigger bump and a bigger bump. What you get is these long lattice structures that come out of that. That's how they grow. And so when you give yourself a little bit of opportunity to know something, you create a little bit of surface area around which knowledge will accumulate. And so then what you now have is lots of areas where you possibly can grow. So if information flies by you and you know nothing about the topic, you don't grab it. But if you know a little bit, then you grab a little bit of that information. And the more you know, you grab even more. And so what happens is you grow over time. What people need to do is they need to expand what they have in knowledge around, what do they have exposure to. And then in this whole automation space, it means like learn a little bit because learning a little bit is what's gonna allow you to learn, learn a little bit more. One of the challenges we have in our industry is that we expect people to be experts before they can be beginners. It requires too much investment of like there's areas where it's like the barrier to entry is so high that unless you know how to like install all this stuff, you can't even start. What we've got to do is you've got to break that down. And so I would tell people, like, look, you know, start to get a little bit of exposure, invest a few minutes before you invest a lot of minutes. And then over time, what that'll do is compound. And if you do that, you know, collectively, you know, daily or weekly, not monthly, then at the end of a reasonable period of time, you're gonna be way better positioned. Like that's what I want my kids to do anyway.

andy lapteff 34:35

I think that's a perfect place to end. I could talk to you forever, but I know I've probably held you over anyway. And just from my own experience, if

— Finding the Time to Keep Learning

andy lapteff 34:42

you're like me and you can say to what you just said, well, I don't have time. I'm busy, I work a job, right? I I have kids, I have a wife or a husband, whatever. If you take a hard look at yourself and you know you're you're honest with yourself, you can find time. Don't, you know, don't look at the whatever that thing is that tracks how much time you have on your phone. Like, you know, there's time there if if you look. Thanks for coming on, Mike. I can't wait to see your keynote. We'll link to it in the show notes. Um, for all things, Art of NetEng, you can check out our Linktree, Linktree forward slash Art of NetEng. And we have a brand new shiny website at artofnetworkengineering.com, building out a resources section, which has a ton of information on uh things to do to get in the industry, communities, people hang out in an RFC section. We're we're adding to it by the day. Speaking of learning things, I've learned enough about dev that I have uh GitHub and we have a repo up there that the website's code that Claude Code helped me generate on is there. And if you see any problems with the website, you can submit an issue in the GitHub and I will fix it. Or you can do a pull request. These are things I've learned recently and you can propose the fix. And I could look at it, merge the branch into main and ta-da! So look at that. The guy who for years said, no, I'm learning new things. So thanks for being here. Thanks for coming on, Mike. Thanks for listening and watching, and we'll catch you next time on the Art of Network Engineering podcast.

SPEAKER_00 36:00

Hey folks, if you like what you heard today, please subscribe to our podcast and your favorite podcaster. You can find us on socials at ArtofnetEng, and you can visit Link Journey forward slash Art of NetEnd for links to all of our content, including the A1 Merch Store and our virtual community on Discord called It's All About the Journey. You can see our pretty faces on our YouTube channel named the Art of Netflix. That's youtube.com forward slash artofnet.