DOP 58: Innovation And The Sunk Cost Fallacy

Transcript

Darin Pope 0:00
This is episode number 58 of DevOps Paradox with Darin Pope and Viktor Farcic. I am Darin.

Viktor Farcic 0:06
And I am Viktor

Darin Pope 0:08
and I am awake. And Viktor is about fall asleep. Because we're recording this in the evening for me and super early in the or what is this for you late at night? Yeah, this is almost almost bedtime for you. So

Viktor Farcic 0:23
almost a couple of more hours and I will be asleep.

Darin Pope 0:27
Okay. So the time we're recording this, it's around 630 Eastern. It's midnight 30 at where Viktor is. And today we have another guest. Viktor, would you like to please introduce our guest?

Viktor Farcic 0:40
Yes. So our guest is Nirmal. I forgot the last name. He will say it probably. Soon. How do I describe him? He's a conference buddy. Would that fit the bill? So yeah, we met in a conference. I don't know which one and then we met in another and a third and a fourth and stuff like that. And he's a really interesting, interesting guy that I've been trying to get to speak with us. So, how long did it take?

Nirmal Mehta 1:13
A couple of decades?

Viktor Farcic 1:13
A year. Yes, yes.

Nirmal Mehta 1:16
In the tech world.

Viktor Farcic 1:18
Yeah, exactly. So basically what he did is that he went to speak with all the other more important people. And then when there was nobody left on the list anymore. He said, Okay, I can I can do you guys as well.

Nirmal Mehta 1:33
My people finally got back to your people. I could get on your schedule. Thank you. Thank you, Viktor.

Darin Pope 1:43
By the way, introduce yourself for real

Nirmal Mehta 1:45
Yeah. Yeah, no worries. I'm Nirmal Mehta. I'm a Chief Technologist at Booz Allen Hamilton and I'm also a Docker Captain which is probably the main thread where I've met you Viktor is is through the Docker Caption program around DockerCon, and the, you know, CloudBees Jenkins World, kind of that world, right. So it's absolute pleasure.

Viktor Farcic 2:12
We were the first batch I think of Docker Captains, right?

Nirmal Mehta 2:16
Yeah, yeah. Yeah. Like, really early on. Probably we were probably already all hanging out before the Captain program became a thing. And then they were like, oh, that group over there. Like, we should just give them some official kind of support.

Viktor Farcic 2:34
Exactly.

Nirmal Mehta 2:36
Well, I'm very happy to be here. Nice to meet you Darin and and excited to have a conversation today.

Viktor Farcic 2:43
So what what will the conversation be about?

Nirmal Mehta 2:47
I have a couple different topics. I would actually I'm gonna do a little bit of a role reversal here because I wanted to pick your both of your brains on on some topics today and the first one is what what do you think drives innovation in a company? Like how do you create new things? How do you you know, get the creative juices going? How do you is that something that you both can relate to?

Viktor Farcic 3:17
First of all, this is strange. Usually it's us asking questions. This is going now off the rails. I will I will challenge that.

Nirmal Mehta 3:28
Okay.

Viktor Farcic 3:28
I will challenge that. I would say that most of companies are not actually doing innovation. They're calling innovation attempts to adopt things that others innovated.

Darin Pope 3:45
Yes, also called imitation.

Viktor Farcic 3:48
Yeah.

Nirmal Mehta 3:50
So, so that the idea there is that there is no there is no true. So, so innovation is is maybe one of two different paths. One is you combine multiple existing things in different fields together. Or you have an actual true leap in like you take in an invention and it's like an actual technological leap forward, like a true, novel new thing. Is there is there any other?

Viktor Farcic 4:24
I don't believe that any of anybody ever invents something truly new, it's always improvement of some on something that we already did before. Right? And I'm not even talking about software industry in general. But going back to the initial question, what to me is funny when when I see you know, a department in a company says, We're innovation company, we're going to install Kubernetes. So, but yes, if we improve something, it's innovation. I, I do believe that. Just that I don't think that many of us are improving something, but just adopting things.

Nirmal Mehta 5:14
Interesting. So you don't think that like all these things can be a platform for actual, like I'm not maybe we have to separate away like research and development, right, like true, kind of just, you know, mixing different chemicals together to, you know, make a Coronavirus vaccine or something, right. Like that's, there's true invention there maybe, maybe. There's then the systems that support those and there might be a combination of like, one industry kind of paradigm with IT and you get some new way of doing something. That's innovation, but what you're saying is like, most, most places that say they're doing that aren't really doing that, right?

Viktor Farcic 5:59
Yes. I mean so let's define innovation. What would be innovation?

Nirmal Mehta 6:04
Yeah, I mean, that's an endless.

Darin Pope 6:07
Actually, let me ask a real question here, which is easier to define. What is innovation? Or what is DevOps?

Viktor Farcic 6:15
Innovation. Innovation.

Nirmal Mehta 6:20
I don't know, actually. I mean, I, well, I feel like I know sometimes, but it's more like if I see it. It's, I can call it out. Right. But it's hard to define when it's, you know, without any context.

Viktor Farcic 6:35
Would you agree if I would say that innovation means that you did something new on top of something existing. Doesn't matter whether it's 0.1%. Right. But you, you, you move the needle, right, no matter how much.

Nirmal Mehta 6:51
Correct.

Viktor Farcic 6:52
Yes,

Nirmal Mehta 6:52
I think that's important. So now the other thing is in your own roles, you might be asked upon you know, might be asked to, you know, create something new or move that needle. Right? How do you? What's that process you go through? Is there even a process or just comes to you? Is it is it like, you know, sometimes when you hear interviews of artists, some of them, you know, sometimes they have like a, they have an answer where the song just came to me at night, you know, like, it's just there. Fully formed, whatever. And then there's some times where it takes like 20 years for someone, you know, they maybe they do something for 10-15 minutes. And then like, three years later, they actually do record the podcast, you know, like, there's like different speeds of this thinking and creative juices going. What in your own experience is, is the process you've been through, when you're when you've been called upon to, to solution or, or create something new or move that needle?

Viktor Farcic 7:59
I will let Darin answer that simply because I'm a product manager now. So innovation for me means Hey, you, innovate.

Nirmal Mehta 8:09
Well, this is important to you.

Darin Pope 8:11
It's not even that. Yours is you send an email saying go innovate.

Viktor Farcic 8:17
Exactly. Okay. So you answer.

Darin Pope 8:20
So I it's it gets back for me to doing the work. And innovation means putting your butt in the seat and doing it. It goes back to I can't remember who the who the writer was. Maybe it was Stephen King or something. I wasn't Stephen King. Anyway, it was the concept of I can't write until I get inspired. But inspiration shows up every day at 8am when I sit down in my chair. So it's it's a matter of taking that and doing that work. That's what it is. It's just doing it. And it's just sitting there and doing the grind or, you know, playing out the old Edison thing. Okay, you know, how many times did it take to get to the light bulb right? Well, I don't know but I figured out 1000-10,000 ways that it didn't work.

Nirmal Mehta 9:15
So it's sharpening those heuristics, so that, you know, the more you do it, the more you sharpen your own heuristics about what paths will probably lead to nowhere, and which paths you should take so that you get a little bit quicker at that, right?

Darin Pope 9:30
Yeah, I'm going to use the little I'm going to use the a word agile, but I'm not using capital A, I'm using little a meaning did this work or the other example I use a lot with people as you know, watch Jurassic Park, the first one, not the other ones, because they all sucked. But the first one was good. And it's when the raptors were testing the fences.

Nirmal Mehta 9:52
Uh huh.

Darin Pope 9:53
Right, they're just going around testing each part going around, it's like, okay, I can't get through this way. I can't get this way. Looks I can get through this way. I'll come back here again in a minute. Let me go check something else

Nirmal Mehta 10:02
Mm hmm.

Darin Pope 10:03
That's how innovation in my head happens, and typically how it works for me.

Nirmal Mehta 10:09
Interesting. I tend to agree, I feel like I have the same patterns too. But what I rely on is my skill and talent to be able to get to those parts of the fence really quickly, and iterate really quickly. So this kind of loops in with the DevOps kind of conversation, right, so DevOps is about increasing velocity. Right?

Darin Pope 10:32
Well, it depends.

Nirmal Mehta 10:34
Okay.

Darin Pope 10:35
Let's assume that's true for a moment,

Nirmal Mehta 10:37
yeah. Okay.

Darin Pope 10:38
So let's suspend disbelief and make sure that's, that's true.

Nirmal Mehta 10:44
Yeah. So hold that thought for a second because I don't want to go I'm not trying to debate that right now. What I'm trying to say is, is is one of the goals of DevOps is to potentially create a system where that testing of the fence with, you know, in the software realm can happen more quickly or more reliably or, you know, with security and safety in mind or, you know, whatnot to be able to kind of find those holes in the fence or wherever that innovation lies. And I have a question for you both is if that's true, maybe I don't know if it's true or not, but let's assume it's true. Do you think we've, you know, lost something from doing software, like more of an old school way, by adopting these patterns, we've lost some of that creativity, lost some of that.

Viktor Farcic 11:49
Let me try to backtrack for a second. Okay for a second.

Nirmal Mehta 11:52
Let's say that this is an interesting question. I don't know if you're, if I'm leading you on the right path to something here, but

Viktor Farcic 11:59
I think To lead us to the answer you want to hear.

Nirmal Mehta 12:02
I don't know. I haven't planned to have no agenda here.

Viktor Farcic 12:04
But kind of going back to patterns. Let's say that I just figure out right now right how to light a fire. I just invented fire. Right? by myself.

Nirmal Mehta 12:18
You're the most amazing person on the planet.

Viktor Farcic 12:20
Yeah. Amazing. That's, that's absolutely amazing. It just so happened that actually others invented that before me. So is that innovation or not?

Nirmal Mehta 12:30
Yeah, I mean, not everyone's fire is the same.

Viktor Farcic 12:34
But let's, let's say the same

Nirmal Mehta 12:35
exact same for Yeah, I think it's still important because it's what it's what you do after that that's also important. It's not like a It doesn't just stop, right.

Viktor Farcic 12:45
Yeah. Because if what you say is right in the those patterns are not that important. Now, in my head, they are extremely important because understanding the patterns and then everything else. going on around in whichever segment of industry you work, allows you to get to the certain level from which you can innovate without actually reinventing the wheel. Right? So for me understanding patterns, it's true, it can easily put me in a box from which I cannot get out easily and cannot really figure out new stuff because you get so enveloped in what you know, that getting out of it. And getting out to the box is very hard. But on the other hand, if you if you don't really know those things, which might prevent you to innovate, but also, if you don't know those things, you might actually lead you to the path of reinventing the wheel, which is probably also not the goal. Because if I don't know what others know, then I might come up with something that others already did before me.

Nirmal Mehta 14:00
We see that over and over again in IT, right? Yeah. Like, same things reinvented in slightly different ways. Do you think that's useless?

Viktor Farcic 14:13
Same things reinvented over and over again, I have a theory about that.

Nirmal Mehta 14:17
Okay.

Viktor Farcic 14:18
I think that every once in a while we need to reinvent or I would actually rather say rename things. Because the objective so that things get lost in the meantime. And then we come up with the reinvention of those things, just to kind of put people back on the initial on the original track. Like to me, a good example would be SOA. I honestly don't see a difference between SOA and microservices. To me, that's the same thing conceptually.

Nirmal Mehta 14:52
Right? Not the implementation.

Viktor Farcic 14:55
Implementation went terribly wrong. So to me my description of microservices, is that. Okay, so you're misunderstood SOA. You went in the wrong direction. So let's call it microservices. Maybe reboot of that

Nirmal Mehta 15:08
And the API gateway is an ESB and

Viktor Farcic 15:11
Exactly, exactly.

Nirmal Mehta 15:13
Service meshes. You know, your message queues and

Viktor Farcic 15:18
I had a session today. Interesting. The main subject was GitOps, you know, what is GitOps and stuff like that. And, to me, that was a similar subject, because my explanation was, you know, we know for a long time that everything as code stored in Git or version control. That's the thing that we already know about for 10 years. But we misunderstood. So now it's called GitOps.

Nirmal Mehta 15:41
right. But also there's, there's changes though, there's, there's improvements, right?

Viktor Farcic 15:46
Absolutely.

Nirmal Mehta 15:47
And it's just like anything like Darin was saying, you know, you just, you sit in the chair, and you do it over and over again, and you do some there's some repetition, maybe 99% is the same each time, but that 1% matters. So, microservices today wouldn't be microservices without an attempt at SOA in the past, correct? Because there's still message buses, right? Which are actually from mainframe, right kind of concepts. Which, I mean, you know, conceptually, a microservices, if you're running them effectively or doing like lambda in the cloud doesn't look very different from a high level from a mainframe, like, conceptually, right? Where you just have functions and you have queues moving messages around, essentially, and I'm obviously oversimplifying. So there is value in going off the wrong path, that well, I don't want to say a wrong path but getting to where we are today, going down this divergent path is okay. It's okay to get in the cycle. I would you say it's a natural part. It's almost a natural instinctual part of the industry, right? Or maybe human nature. So, so at some point, we're gonna have this conversation and let's say five, six years again. Next time, I'm on your podcast in five years. And we're going to talk about something that is essentially just Kubernetes but over again, right? I mean, that's, that that's, that's the pattern, right? Like, that's what we're saying.

Viktor Farcic 17:32
That scenario that you're describing and that Darin is saying, oh my god, actually, I truly believe that's inevitable. We will be talking years for some time from now we will be talking first of all, how Kubernetes is horrifyingly bad. How something else is so much better. While essentially it is not much different either.

Nirmal Mehta 17:54
and I know I know we had part of the book we talked in our interview for that we talked a little bit about what's being called now low code, no code, right? This business logic abstraction. And and maybe we'll save this for another conversation another time. But it starts to venture in there because there's all these attempts at. I feel like there's also there's natural wall of where we can't get past in terms of, of abstraction. Right. Like, I feel like the reason why we do reinvention of the wheel on these certain levels and layers is because we can't get above a certain level of abstraction. And, and there's been many attempts at trying to decouple or abstract away all of the IT problems and only focus on business logic. But, you know, it never really pans out because there's that natural trade off, right if you abstract too high you lose your customizability. Right? And then vice versa. So maybe we'll save that for another time. Maybe tee up another conversation around that. But I think all these things are kind of linked together. And it's very interesting. I do agree with you Darin, like, there is a loop. There isn't like a reinvention loop or a just, you know, brute force almost kind of pattern to innovation, inventing creativity that I think a lot of people miss when they just shoot off their email saying, hey, you go innovate, you know, like, doesn't just appear out of nowhere.

Darin Pope 19:46
Well, it's how hackers work. Yeah, right. It's a hacker doesn't just like, Okay, let me crank up Metasploit and send it at one IP address. Oh, it didn't work. Okay, I'm gonna go to the beach today.

Nirmal Mehta 19:57
Yep

Darin Pope 19:57
No. It's like they're going after it left, right. And I think that gets back to it's either showing people and this goes back to a conversation that re-conversation that we had with Joost, the other week of, you know, sometimes people don't want to learn anymore. Sometimes people can't learn. Right, let's let's put aside the people, people haven't, that aren't in the situation to where they are in a... We'll use his example from the other time of kids that had a child with special needs. Right? And that's so there's special life or life circumstances that are causing you to where you can't do all the stuff you want to do from a career perspective. Completely okay. But that's a choice that that has either been forced upon you or a choice you've made, but for those of us that have chosen, yeah, I'm just gonna go off and learn. And I'm going to try this and I'm going to keep trying it until I either get good at it or I get bored, as Viktor does, gets bored and moves on to the next thing. I want to stay I want to stay back to the you were talking about the low code, no code, which is yet just another name for 4GL. For those of us that are old.

Nirmal Mehta 21:15
I unfortunately don't know what that is, but it's before my time. Darin is gonna kick me off this podcast, Viktor.

Viktor Farcic 21:25
Yeah, kind of you're a teenager

Darin Pope 21:27
we'll get back to what we're talking about. So So you were asking questions of, Alright, so we're when we first joined our hero, last left our hero, we're talking about low code, no code. Now, bring it back to where we're heading. Low code, no code and DevOps is like, okay, that's a whole nother conversation. In fact, we have a friend that works for one of the big no code places.

Nirmal Mehta 21:47
There's been so many attempts at that, and maybe maybe I'll come back on after that conversation and we can talk about it but that that's as a whole because even in the interview I did with Viktor, I kind of went both ways I, I can, I can see how it's very tempting to kind of go down that path of, you know, this nirvana state of, you know, easy to do things in IT. But the reality is a lot more complex always.

Viktor Farcic 22:19
One of the things that I'm, I believe I'm getting more and more aware of is that it also is greatly related with timing. So like, if you stick with no code, low code, or whatever we call it today. I do think that that's actually a good idea. However, all the past experience I've had with that was horrible. And that might be because I'm wrong. It's not a good idea at all, or simply because some ideas are premature, right, kind of, maybe it wasn't a good time 10 years ago when I used...no...okay..not 10. 20 years ago when I used Dreamweaver to do drag and drop for everything, I don't know if any of you remember that?

Nirmal Mehta 23:09
I remember that. Now we're in a conversation, like all that see the like the crazy HTML templates that came out of that

Viktor Farcic 23:15
exactly. But that does not necessarily mean that. I mean, that was innovative in a way if you think about it, but

Nirmal Mehta 23:25
still a very powerful tool at the time,

Viktor Farcic 23:27
oh, you can still download it. And I think that that that's one of the big subjects about innovation is not only about figuring out something new but figuring out something new at an exact, precise moment. When that something is, you know, you cannot be too early or too late. It needs to be exactly the correct moment.

Nirmal Mehta 23:52
I don't know if it has to be or it just becomes the right moment because of maybe the implementation or the ecosystem around it. Right. And what's in the Overton window in IT at that moment? Right. So, you know, maybe that was a failure of SOA was that not everyone was thinking in a way of in, in these units of work kind of paradigm, right. I don't know. Like, you get what I mean, there's there's like a what's in the Overton window in IT at that time matters. I bet you know, like you're saying about microservices just being SOA. Right. Microservices can be sold because containers because lambda, right?

Viktor Farcic 24:38
Yes, exactly.

Nirmal Mehta 24:41
And that's in the zeitgeist. Like people know what that means, roughly, so that you can you can hook the new the new old, the old new idea, the new old idea onto what's in what's relevant today. And maybe that's, that's the important piece that tips that that's a tipping point

Viktor Farcic 24:59
To me a good example of something that was maybe premature is my favorite Heroku. I still believe that, that it was in some ways, still ahead of what we have today. Right? I mean, Kubernetes or before then Docker and then cloud and all those. It was fantastic and we're still maybe not even there. But it died a horrible horrible death kind of mean, I don't know. Is there anybody using Heroku today? Probably not.

Nirmal Mehta 25:34
Well, I mean, I'm sure there are people using it and there's tons of you know, you know, if you want to mask shop right now, for for the pandemic, you can spin one up, you know, on Heroku or Squarespace or whatever, but they're also releasing more features obviously, as time goes on, but I understand what you mean like the simplicity, right, the the kind of just immediate ability to run these workloads,

Viktor Farcic 26:02
but also combined with that, which is just the right moment. I think that if Heroku appeared a year ago, it would be a much bigger success than when it did appear, or if somebody came up with Heroku maybe a few years ago, right, let's say but not like 10 or whenever how many years it appeared, it would be a huge company. It would dominate the world.

Nirmal Mehta 26:29
Right. But the thing that so it wasn't dotCloud, basically a Heroku kind of company. Yes, actually. Yes. Yes. That's interesting. I think another thing that that you were just about to touch upon, is that the user experience matters. I think we forget about that. And I think that's why Docker was very successful because the user experience is like phenomenal, especially for the developer. And I also think that's why Kubernetes is really tough because the user experience is a little rougher from a developer point of view. And obviously, that's, that's a big hot topic of discussion these days. But I want to take a step back just from like, the Docker and Kubernetes. But just user experience in general, for any kind of innovation, I think is super important. And maybe the failure of SOA, or the failure of XYZ thing in the past and the reinvention is just wrapping a better user experience around something old. What do you think of that?

Viktor Farcic 27:40
Here's the thing that just came to my mind. If you would measure success of something, some innovation by how adopted that becomes in companies, not necessarily individual users,

Nirmal Mehta 27:55
so not in the actual success of using that technology to achieve anything. But just by just adoption

Viktor Farcic 28:02
and adoption by companies. So I'm trying to distinguish companies and simply users. I have impression that actually, the solutions that are very user friendly, do not win.

Nirmal Mehta 28:18
Hmm. Why do you Why do you say that?

Viktor Farcic 28:20
I don't know. Like, let's let's compare Docker and Kubernetes. And focus only on user experience, right? Let's ignore everything else. Kubernetes wins. And Docker is definitely much better user experience than Kubernetes.

Nirmal Mehta 28:41
But if you talk to developers who are not making the decision to go with Docker or Kubernetes, then the answer to that is opposite. So that's driven, that's a selection biased on who's making the decision to purchase something in that company or to adopt something in that company.

Viktor Farcic 28:59
Exactly. So and I actually like that answer because that leads me to believe that decisions are more often not based on what will the majority benefit of, but rather the few minority making the decision. Yeah. Because if you look at if you look at Kubernetes, and Docker now, again, I'm ignoring everything else. A Docker, I do believe Docker or Docker Swarm benefit, higher number of people in the company would benefit from that, as opposed to Kubernetes directly as direct users. On the other hand, if you if you look at it from the point of view of let's say, a single department, then Kubernetes wins, because if you're only an exclusively sysadmin Kubernetes gives you much better experience higher return of investment and so on and so forth. Right. So in so that sysadmin department hypothetical the let's not stick with that right? Won over everybody type of stuff using only Docker Kubernetes, as example, right and ignoring the call histories, stuff. But I believe that there are many other examples where benefits of the few outweigh the benefits of the of many.

Nirmal Mehta 30:27
I can say one that's very controversial on this, I would say Jenkins. I mean, think about it, like the user experience for Jenkins is a really it's pretty rough, right? Like it's gotten better. But it's not the winner from a user experience perspective, correct?

Viktor Farcic 30:42
Yes.

Nirmal Mehta 30:44
But it's still a winner in terms of adoption, because of the people that were the people that had to select that tool for some reason, right? Yeah, maybe.

Viktor Farcic 30:54
Yes.

Nirmal Mehta 30:55
And then it morphs into this it grows right, like so that decision then instead of being done in a more design thinking way or more consensus, you know, proper, like systems engineering, you know, requirements gathering kind of way, you know, the tool is there by the small minority group. And then it because it's then selected and you can't change that, it morphs into this thing that's, like a monster of what it was initially adopted for. Right?

Darin Pope 31:32
You mean sort of like a spreadsheet becoming a production database?

Nirmal Mehta 31:37
I think I think Excel is probably one of the best user experiences ever. And I think it's gonna be hard to defeat that. Just just, I'm just gonna throw that out there because, I mean, it is almost a universe like, you know, it's up there with, you know, iPad or something.

Viktor Farcic 31:54
I must agree with that because I am yet to find I'm not a user of Excel, so I cannot really compare, but whenever I speak with some heavy user and ask him, why didn't you try Google, what is it, Sheets? They say, Are you kidding me? That's one of those unbeatable products.

Nirmal Mehta 32:13
I see. I saw it is it is it truly is. I saw last week or something. Two weeks ago, someone made a excel sheet that controls Kubernetes through Excel. I don't know if you saw this. It's amazing. But

Darin Pope 32:32
I wondered why when I did see it. Just because you can doesn't mean you should.

Nirmal Mehta 32:37
Yeah, it's it's just just another poke at the fence. Darin, it's just another poke at the fence.

Darin Pope 32:41
It is just another poke. Really? A spreadsheet? Hey, find something else, please. Anything but a spreadsheet.

Nirmal Mehta 32:48
I think that might be the future of low code right there. I'll put it I'll put a prediction there. It'll be Kubernetes to AWS Kubernetes. To our Excel to Kubernetes was the is the future of low code.

Darin Pope 33:05
I'm just sitting here stunned right now thinking he's right.

Nirmal Mehta 33:09
Just think about it. If, if Microsoft turned the switch and connected, like lambda, like Azure Functions directly into Excel, how many enterprise applications wouldn't be needed anymore?

Darin Pope 33:28
99%

Nirmal Mehta 33:31
that day is coming. I bet

Darin Pope 33:33
that extra 1% is just because it's on iOS. Yeah. But you're gonna run Excel on iOS. So

Nirmal Mehta 33:39
anywho I think I wanted to pick up another thread, which was about learning organizations. Do you feel like you're in a organization when you're talking to your clients and customers? are they learning organizations? Do you think that matters with DevOps? What's your opinion on that Viktor?

Viktor Farcic 33:57
No.

Nirmal Mehta 33:58
Why?

Viktor Farcic 33:59
Because they just because we just need to finish the current project, it is essential it is we are having a difficultly to reach the deadline.

Nirmal Mehta 34:12
So, taking the time to learn and to train the people around you is not important.

Viktor Farcic 34:18
No, it is but after this project,

Nirmal Mehta 34:20
but that never ends.

Viktor Farcic 34:22
Exactly, exactly. But it's always this project, right? It's not after this project. Yes.

Nirmal Mehta 34:30
So would an ideal situation be where it's a learning organization? And you're just saying that never, that's not real world.

Viktor Farcic 34:36
In a healthy organization, I think that learning should be in part of the job, right? It should be part of daily activities in a way. The same way as I believe that refactoring your code is just as important as writing the code, right?

Nirmal Mehta 34:56
Right or, or getting rid of some technical debt or re evaluating what you are doing

Viktor Farcic 35:00
so and but the real issue I think is that most of the world is driven by quarterly objectives or project based objectives. And that really prevents you to work on things that will give you benefits beyond short period of time. Right? So, refactoring or learning and technical debt, those are all the things that do not give you immediate benefits. And that is directly conflicting with how this how this world operates. Eventually, for a company and this is very sad to me, and I'm very disappointed but, but the thing that eventually companies would rather acquire another company than learn or improve and things like that. It's cheaper. We're gonna buy a new company where when. So if you're a bank right and we need this innovative FinTech stuff, we're gonna buy a company eventually. That's the type of reasoning. But you you just need, this is very important. You just need to finish this first. And it's always something.

Nirmal Mehta 36:14
I wonder if Is it a? Is it a hidden competitive advantage to be healthier then? Or is there just so much low hanging fruit so much? You know, fat there that there's it doesn't matter,

Viktor Farcic 36:32
I think is directly related with the position you have, in terms that if you're a leading company, then those things bring more obvious values right. Now so if, let's say if 10% of my application needs to be refactored the need to do that 10% refactoring in the benefits are more obvious than if you say, okay, so we look, we're in such a bad shape that actually refactoring obviously would mean three years of investment. So therefore, I cannot see a value in that. So, and similar for innovation, the closer you are to being on the top of the chain, you know, kind of the closer you are to really be at the top so that you can truly innovate, the more likely you are to see the benefits of such innovation, but if you're too far behind, then those things sound not so pleasing or beneficial.

Nirmal Mehta 37:48
Interesting. I tend to agree and it's, I think that is the paradox, right is when when all this stuff really hits the real world and in business motivations and and, you know, market, real, you know, realistic markets and all that and capitalism, you know, at a high level. It's a real challenge, right?

Viktor Farcic 38:11
If I'm right in what I just said, that would truly be a paradox because that would ultimately mean that the more there is a need to innovate or the more there is a need to remove technical debt, the less likely you are to do it.

Nirmal Mehta 38:26
Yeah. And that's why I asked the question in the beginning, that's what I was trying to get to is that there's this inherent weird conflict. It's a it's a it's like a weird and I know if you have any economist listeners, it's a bad analogy, but you know, it's the company's in a weird Nash equilibrium where they know they need to do some things to get to a certain point, but there's no incentive to do that. Right? Like there's no there's not enough of an incentive to do that. It goes against their short term rational behavior leads to long term irrational behavior.

Viktor Farcic 39:09
I think that that short term rational behavior, actually is very important. Because if you know, if I go back to, if if you need five 5% of our application to be refactored, then actually you can do it on a short term and see the benefits. I believe that many companies are locked in into that thing, like kind of we are in such a bad shape, that there is nothing anybody can do to make this better. So let's not even try.

Nirmal Mehta 39:41
and then that just makes it worse. But then you just acquire, you think you can acquire your way out of that, or you try some other scheme to get out of that position. Correct?

Viktor Farcic 39:51
Yeah, you acquire a company and then you put that company into your existing structure and then you destroy that company instead of improving yourself.

Nirmal Mehta 39:59
It's a sunk cost fallacy, unfortunately, because really at any moment during that, it is always worth it to just reset. Because it doesn't really matter that you've spent X amount of time doing something. It's always worthwhile resetting, but you can never, it's a that's a really tough sell, right? Like that's an impossible sell to a leadership. That that's short term driven.

Viktor Farcic 40:23
It's also partly not realizing the loss. There are some studies that say that if you go gambling, right, the more money you lose, the more you will believe that gambling more will actually make you win.

Nirmal Mehta 40:41
Right. The gamblers fallacy

Viktor Farcic 40:43
Exactly. And I think that that somehow applies to also what we're doing is that you will continue pouring money into something that you know that it is actually a loss.

Nirmal Mehta 40:56
right. It's a sunk cost fallacy.

Viktor Farcic 40:58
I didn't know it's called like that. Thought I just invented it. I'm not as innovative as I thought.

Nirmal Mehta 41:07
Yeah, so sunk cost fallacy is exactly what you're describing. And it is, I would say it's, you know, top five cognitive biases that humans have. Um, it's related to the gamblers fallacy, which is what you were just talking about where in gambling, each, each event is independent of the last events. But that, you know, emotionally it does not feel like it's independent. And that's where the, that's where the fallacy happens. It Same thing with sunk costs, like, you know, if you if you tie your identity or you tie some money into something and it goes the wrong direction, it's actually better just to change direction and just pretend you never spent any money on that. But emotionally it's very hard to detach a leader You know, a leader can't just detach themselves from that wrong direction, the engineers can't detach themselves from that wrong direction, money money was spent. But from an economic rational standpoint, you should change, you should just pretend like you never spent anything, and start spending the money on the new way. But it's very hard to do that. It's almost impossible for an organization to do that. And successful technical leaders are able to change that right, are able to politically navigate their organization to do that, but that's a very tough thing to do. Right, that's a that requires, you know, very, like, hardcore, you know, leadership, if you will. Right. So, that's, that's, uh, I think we've we've done a full circle around that paradox. Now. You know, the cornerstone, the cornerstone of this podcast, I think, I was trying to loop us around that a little bit, because I think it's fascinating. And what frustrates me and I think that's what are one of our initial frustrations were back in the day, Viktor is that, you know, we go to these conferences and we listen to these talks. I'm not talking about the technical talks, those are always great. And but, you know, kind of more of the, you know, these big tech companies, and this is how we did things. But it's almost like a different universe, right? It has nothing to do with, you know, my customer base, your customer base, the typical enterprise, like, there's such a huge disconnect between or a technical divide and, and philosophical divide and leadership divide between, you know, these tech companies that, you know, quote, unquote, do things very effectively. And then the rest of the, you know, the rest of the IT industry. And the paradox comes from translating all that, you know, to the real world. Right. What's interesting, though, is when you start to peel back the layers, you realize just how dysfunctional everyone is, you know, you know, to an extent, and I think one one message I would like to say to your listeners is, if you feel like you're behind other places, if someone is seems like they're doing really cool things, they probably are. But it's missing the context of the bigger picture, right? Maybe they are the market leader. Maybe they have a monopoly in that space. Maybe they have a different, you know, revenue model and etc. Onward and upward. And that's not put into the context of why they did the things that they did, right, Viktor?

Viktor Farcic 44:44
I would rather say that everybody becomes eventually dysfunctional.

Nirmal Mehta 44:50
Like, like entropy in the universe?

Viktor Farcic 44:52
No, it's that maybe the speed with which our industry is moving is too high. So that nobody can really keep up. And the advantage will almost always be with the newcomers. So if I just start a new company without the so the speed is so high that everybody generates baggage legacy, right kind of technical debt, everybody, absolutely no exceptions. And that that includes the uni... I mean, you know, kind of shiny beacons like like Google and stuff. So, if you ... when you account that speed of moving forward and generation, inevitable generation of technical debt, then maybe actually the world always belongs to new companies kind of start from scratch. But based on the existing knowledge, right? Maybe that's that's how our industry is destined to work in a way. Die and be replaced with somebody new, and that somebody new will die and be replaced by somebody else.

Nirmal Mehta 46:00
Interesting, very interesting. What do you think of that Darin?

Darin Pope 46:04
I think that death is good thing. No, I'm I'm awake. barely awake.

Nirmal Mehta 46:09
Oh, wow.

Darin Pope 46:11
What do I think about it? I think you're right. People think this goes back to the very beginning point very beginning question innovation. How does innovation work? People are born. People die. That's innovation. Right? The next person comes along and does the next thing

Nirmal Mehta 46:30
fighting against entropy.

Darin Pope 46:34
Right. Wow. I feel like we. I've said this past few episodes, it feels like we just got started.

Viktor Farcic 46:41
Yeah.

Darin Pope 46:42
So you have to come back.

Nirmal Mehta 46:44
Was this was this the kind of content you were looking for?

Viktor Farcic 46:47
This is awesome. I love it.

Nirmal Mehta 46:49
I feel like we're just warming up.

Darin Pope 46:52
I don't ever know what kind of content we're looking for. That's the thing.

Viktor Farcic 46:55
And let me tell you that Darin is very happy because every time we have an episode and I don't swear even once, he's happy.

Darin Pope 47:03
It makes my life so much easier.

Nirmal Mehta 47:04
Oh, oh, this is a family friendly family friendly.

Viktor Farcic 47:08
I didn't swear yet.

Darin Pope 47:10
yet. We're not done yet. Well, it's not okay. So yes, it's friendly family friendly. But if you're driving around in the family van, you've got your kids and you're listening to this podcast, you need help. Yeah, you need help. That is the paradox is don't your kids don't need to be listening to this. But we want it to be safe enough to free so you can take it into the boardroom or into the conference room right or share it with whoever and share it with other people and not have people freak out. We already say enough incendiary things anyway, so we don't need to add swearing on top of it. All right now if you are listening via Apple Podcasts, please subscribe leave a like leave a rating and review. All of our contact information including Nirmal's today is gonna be found at....his won't be on https://www.devopsparadox.com/contact but his information will be down below in the show notes. If you want to be notified by email when a new episode is released, you can sign up at https://www.devopsparadox.com That signup form is at the top of every page. There's links to the Slack workspace, the Voxer account and how to leave a review in the description of this episode. And I've already determined I suck recording late at night. I'm much better at five o'clock in the morning.

Viktor Farcic 48:24
That's that's what makes us work so well because I'm also very good at five o'clock in the morning.

Darin Pope 48:29
Yeah, but you're still awake then you haven't been to bed

Nirmal Mehta 48:32
you need to tell I feel so bad now. I feel like I'm not getting the full Darin/Viktor experience. Like do we need to reschedule this for five in the morning? I could do that.

Viktor Farcic 48:40
No, the difference is that I I'm good at five o'clock in the morning and then I go to bed.

Darin Pope 48:46
I will have just gotten up at three o'clock to be ready for the five. Oh, I see. So, so Nirmal. Let's let's figure that let's finish this up the real way. What is innovation? You spun it to us, I'm going to spin it back to you what is innovation? You have 30 seconds go.

Nirmal Mehta 49:07
I believe innovation is the combination of disparate ideas, disparate industries, different, you know, sources of information, new ideas, combined to create an additional new idea. And then sometimes, very luckily, a true invention or technological leap is added into the mix. And innovation is really doing that combination of things iteratively, at high velocity to make, you know, future progress for mankind, or humankind.

Darin Pope 49:45
man, you're good with words.

Viktor Farcic 49:46
I can already say that this is not that he's thinking aloud. He thought about this before.

Nirmal Mehta 49:52
Nope. I have no notes.

Viktor Farcic 49:55
Oh

Nirmal Mehta 49:58
that was off the top of my head.

Darin Pope 50:00
You're really good with words.

Nirmal Mehta 50:02
Thanks. That's why I'm a consultant.

Darin Pope 50:08
That's why you're a consultant because you're good with words. That's where I've gone wrong. I'm not good with words.

Nirmal Mehta 50:15
Do you want me to record that again?

Darin Pope 50:18
No, you're good.

Nirmal Mehta 50:19
Okay

Darin Pope 50:19
Unless you've got it word for if he if he if you say it again word for word. We know you're lying.

Nirmal Mehta 50:24
I don't even remember what I said. Now, since I feel like I should not have said mankind, I should have said humankind. That's the only thing.

Darin Pope 50:32
That's okay. Okay. You just said humankind. Yeah, we're not trying. Hey, look, we're offensive about a lot of stuff. In fact, if you listened to last week's episode with Tracy, you know that that was the very first female we got to talk to Tracy by the way and get some more. More people on of different

Nirmal Mehta 50:49
I think you need to get Laura on this.

Darin Pope 50:51
Which Laura?

Viktor Farcic 50:52
Oh, yeah. No, she's changed the last name now. She's Laura Tacho. Tacho. Yes, that one.

Darin Pope 51:01
Oh, that Laura. Okay. All right.

Viktor Farcic 51:04
But now, we can use this opportunity to kind of since this is being recorded, we can force Nirmal to join us. Friday 615 live

Nirmal Mehta 51:15
tomorrow?

Viktor Farcic 51:16
Yes, tomorrow. Yes.

Nirmal Mehta 51:17
Let's see.

Darin Pope 51:18
Well, so here's the deal. This episode comes out on June the 3rd.

Nirmal Mehta 51:23
Oh, the day after my birthday. How about that?

Darin Pope 51:29
This is your present. So then Okay, so maybe we can have you on this week, which is not this week, but

Nirmal Mehta 51:37
Oh, because there's enough time in between

Darin Pope 51:39
this time. Dude, I have to edit not that I edit a lot, but I still have to. The editing easy is easy. The transcription is hard.

Nirmal Mehta 51:48
You do it yourself?

Darin Pope 51:50
I do it myself.

Nirmal Mehta 51:51
Oh, man. Let me know if you need help with that. I'll help you.

Darin Pope 51:54
Tag, you are it.

Nirmal Mehta 51:56
Oh no. What do I do that? I'm your guest, Darin.

Darin Pope 52:02
Hey, you're part of the family now. All right, any

Viktor Farcic 52:06
How many hours we have of material, not only podcast but in general to be

Nirmal Mehta 52:11
just about this recording.

Viktor Farcic 52:15
You committed

Darin Pope 52:18
Yeah.

Viktor Farcic 52:22
And bear in mind that the main problem in in transcribing things is me. My broken English.

Nirmal Mehta 52:31
I know no one understands you. That's why no one's fired yet.

Viktor Farcic 52:35
Exactly. They don't know whether I what they said.

Nirmal Mehta 52:39
Is that 615 in the morning or 615 at night, tomorrow. And is that Eastern?

Darin Pope 52:46
I don't know. What are we doing to I don't know. It's we're doing it Friday. It's tomorrow. Let's Okay, we're recording on Thursday. I don't even know what today is.

Nirmal Mehta 52:54
Today's Thursday.

Darin Pope 52:56
Today's Thursday, May 14, right. This episode is releasing on June. The Third Wednesday, June the third.

Nirmal Mehta 53:01
Okay.

Darin Pope 53:02
And then we have a live stream. Of course you're listening in the past so you can go back and YouTube and watch this. When do you want to be on the live stream tomorrow? Its at 230 Eastern. Tomorrow to be on that? Oh, 230 Eastern. Yeah, that's the problem. You probably have a day job tomorrow. That's my lunch break.

Nirmal Mehta 53:18
How long?

Darin Pope 53:19
hour ish, you can just drop in if you want to. This is gonna be on the podcast anyway. So you're still listening to this. So we're, you know, we're breaking down the fourth wall, all sorts of ways. So if you're still listening to this right now, thank you, but just hang out for a sec more than we'll finish up. Can you do it? Can you do it?

Nirmal Mehta 53:36
230 Yeah, I can do it.

Darin Pope 53:38
Okay, I'll send you the information. Okay. That being said, we're gonna have a live

Nirmal Mehta 53:42
who else goes on there?

Darin Pope 53:43
It's, it's just me, Viktor and you. That's it. Okay, that's all that's all we can handle. That's all the internet can handle, I think. So if you're still listening to this, go back to YouTube. Watch the recording. In fact, by the time I should be able to get the show notes, you can go back and watch the recording of this. That's sort of happened live three weeks ago. That's it. Thanks again for listening. Hey, that's it. That's it. Everybody. Everybody quiet. Let me finish. Let me finish. Why am I being so mean? Because I'm hungry. I've had Pop Tarts today. That's all I've had to eat today. Deep breath. Thanks for listening to episode number 58 of DevOps... one more time. See even so much time. Here we go. Last time. Thanks for listening to DevOps. Okay, really? Can I really do this?

Viktor Farcic 54:37
Thank you. Goodbye. Thank you. Goodbye. Go away.

Darin Pope 54:44
Okay, that's the right way to end it. Thanks for listening. See you Bye.

DOP 58: Innovation and the Sunk Cost Fallacy

Show Notes

Guests

Nirmal Mehta

Hosts

Darin Pope

Viktor Farcic

Links

Rate, Review, & Subscribe on Apple Podcasts

Signup to receive an email when new content is released

Transcript