PurePerformance - AI Is a Gift: Rethinking Software Engineering Education and Hiring

Episode Date: June 22, 2026

In this episode, we explore how AI is transforming education, from classrooms to corporate training. What changes are needed in schools and universities? How does AI affect both students and educators? And how should companies rethink internal training and hiring to stay competitive?To answer these questions, we’re joined by Rainer Stropek, CEO of Software Architects and Chairman of Coding Club Linz. With decades of experience teaching at high schools and universities—and helping organizations upskill their engineers—Rainer brings a unique perspective on how software engineering education is evolving.While many view AI as a threat, Rainer sees it as a “Christmas gift”—opening up endless opportunities to learn, adapt, and innovate.Tune in to hear why curiosity is more important than ever, how educational institutions can prepare future engineers, and why organizations must step up to ensure everyone has a fair chance to succeed in the age of AI.Links we discussedRainer's LinkedIn Profile: https://www.linkedin.com/in/rainerstropek/Rainer's Website: https://rainerstropek.me/CodeClub: https://codeclub.org/en/Coder DoJo Linz: https://linz.coderdojo.net/

It's time for pure performance. Get your stopwatches ready. It's time for Pure Performance with Andy Grabner and Brian Wilson. Hello everybody and welcome to another episode of Pure Performance. My name is Brian Wilson and as always I have with me, my co-host, Andy Grabner. Hello, Andy. How are you doing today? I'm good, I'm good. I just came back and I think I mentioned this. I just came back from the U.S.

And I was, we didn't make it all the way to Denver, but we did an amazing road trip with my brother and my two nephews. and a colleague, spent 1,600 miles on the road. What's the favorite thing he did? We saw Indy 500. Oh, wow. I've never seen a race like that. I've never seen any auto race, actually.

Yeah, we saw a lot of things. Indy 500, we went to a tractor-pulling event because my brother and his nephews or his sons are really into tractor-pulling. And, yeah, it was an interesting trip. But more stories for another time. I think I'm not sure if our guest is really interested in hearing what we're doing in our spare time. Now, talking about our guest and talking about the topic, AI, is still and will remain for a little while longer a predominant topic in our industry, how it changes the way we work, what type of jobs it may not eliminate but changes, right? And as a podcast host, Brian, I actually used AI to just ask it, you know, how would I introduce. What do I need to know about our guest?

How should I introduce it? And I will just start reading a little bit here, and then I will obviously let our guest introduce himself. But it says... I'm not AI. I'm not AI, but I just see what... It's a strong podcast introduction, ready to read. Yes, let's go.

Today I'm joined by... We are joined by Reina Stropeg, co-founder and CEO of Software Architects, based here in Linz, Austria. Reina has been leading the company since 2000. where he and his team build and operate the award-winning SaaS solution time cockpit and help organize organizations design and run modern cloud-based systems. He's a recognized expert in software architecture, cloud computing, and modern.net development

and has spent decades working hands-on as developer architect and consultant. And it goes on and on, but I want to now actually allow our guest to speak for himself. First, Raina, did the AI do a good initial job? in getting and explaining what you do and who you are now i'm feeling really old thank you for introducing me and the decades of experience yeah okay i have to admit i have a lot of white here because i still have here i still have a little bit yeah yeah but it's getting yeah i'm losing some of it maybe because of the i will see thank you for having me i'm really looking forward to this discussion today and rana you are sitting and we are brian and i are both

jealous. You're sitting actually on your patio enjoying the nice weather here in Austria right now. Absolutely. But Raina, besides, obviously, your deep technical expertise over many years, you're also very active in the educational system here in Austria, but I guess not only in Austria, maybe worldwide. Can you tell us a little bit about that part of your life on what you've been doing? Of course. In fact, this is what I started my career in. I finished a higher technical school here in Austria for computer science. And right after my final exam, I started as a freelancer doing trainings and coaching and consulting at the early days on an AS 400. I don't know if anybody can remember this mid-sized

computers. So this was, in fact, my first real job as a freelancer, as an entrepreneur, doing already some training and coaching and I've not stopped ever. Since I've done a lot of different technologies, but working with adults, training, coaching, but also working with young people. And even in the last 10 years, a lot of kids is what I really am passionate about. I love talking about technology. I love telling people or showing people the fascination of technology, especially software. I'm really a software guy. I'm not so much into hardware, but this is what I do.

and this is what I practiced over the last three decades. And to be honest, I had to reinvent myself, I think three times in my career. I started with all the C and C++ stuff, and then came the managed languages. I had to reinvent everything. The whole game changed. That was the first time that I had to reinvent myself.

The second time was the whole cloud computing stuff. As I mentioned, I'm a software guy. And suddenly I was able, doing infrastructure as code, to build infrastructures, something that I couldn't do before. It changed everything. It changed my entire business. And in the last four years, I would say, AI has again changed everything. And yeah, as somebody who loves teaching things and playing with things, this is not bad news.

On the contrary, I love it. For me, currently, every day is like Christmas. I'm not afraid of this technology. On the contrary, every day I can find out something. new and I'm fascinated with it. I'm really fascinated with it. I love AI. I love working with AI. I love showing people how to work with AI. And I love working with young people and children and young adults who are as fascinated by AI as I am. Really fascinating and really cool to hear your positive attitude towards change, right?

Because obviously change is something that some people may shy away from. because it means you need to change something that you have over many years, maybe learned how to master, and then you need to kind of start all over again, and I understand that many people might have a problem with that. But I really like it that you see it as a great opportunity. I like it that you call it even Christmas. Christmas happens every day for you.

Raina, you already mentioned that you like, that you work a lot with young adults. I think you're also a chairman of coding club Lins and you're supporting the Codotodotio. Can you explain a little bit about these two? Yeah, of course. Of course. Besides my work as an entrepreneur in my own company,

I have been the chairman of the Coder Dojo Links, of the Coding Club Links for nearly 10 years now. We run various community activities, and especially the Codododogilins, which is a programming club, a free volunteer-driven programming club for kids. We start with kids at age eight. Besides that, I also am a part-time teacher

at a vocational college for computer science here in Austria. I'm a regular teacher at the university for education here in Upper Austria. So I teach teachers about computer science and AI and things like that. So you see teaching and working directly or ring directly with kids and young adults is a crucial part of what I do. Yeah. I mean, obviously, I want to go later on to the topic of, you know, how the educational approach needs to change.

But I first want to ask you a question, because maybe some of our listeners now think, wow, this sounds really cool. Coda Dojo, but is the Coda Dojo something that only exists here? Do you know, is there like a global organization? Are there maybe people that can contact somebody in their own local region to find out if something like this is available, maybe do something similar? Because it sounds like a really interesting community. Yeah, absolutely.

Kota Dojo is something that exists worldwide or has been there worldwide in the last years. Things have changed a little bit. Kota Dojo was partly rebranded to what is known as Kodkp. So if you hear Koda Dojo and Kod Klopp, they are essentially the same. They joined forces a while ago. And they are also part of a larger organization, which is the Raspberry Pi Foundation. I'm pretty sure that everybody knows about the Raspberry Pi and the underlying foundation is the Raspberry Pi Foundation. So Koda Dojo, Kodoo Club and Raspberry Pi Foundation is essentially one large organization and it has a global footprint.

So you can either find Koda Dojos or Kodubs or Raspberry Pi Foundation events all across the globe. And yeah, the best place to find resources is to go to codeclub.org. and there you find a page with all the information about codeclubs and Kodogos and so on. We are Koda-dojolins, so if you Google for Koda-Dogulins, you will definitely find our website. Cool. And folks, as you're listening in, as always, we will add any links discussed also to the podcast description, so it's easier for you to find. Now, to one of the main topics, right, and for me, when we had a discussion a couple of weeks ago, and obviously we've known each other for many years

but the whole aspect of how does AI change our industry because I have a lot of people now that come up and say should I send my kids to a technical school should they learn coding isn't at all going to go away a lot of people are hesitant

because everybody's uncertain on where the industry is going I will just be interested in your perspective because you are both obviously very active in the AI field. You know a lot of stuff. You also educating. So my question to you is, how does AI change the game for junior software engineers? So what is the things that you can, that you see out there, right? Especially for junior engineers.

Well, there are multiple aspects to the story. The first one, let's not talk about how it changes for, the education for being a professional software developer. If you think about joining a technical school, doing that technical school, it's not, it doesn't need to be for becoming a professional software engineer. There are a lot of schools where you learn something and you are not absolutely going to be a professional in that area. There are schools for sports. Not everybody who is doing a school, particularly for sport, is going to be a professional athlete. There are schools with a very strong focus on music. Not everybody visiting the school is going to be a professional singer or

maybe musician or something like this. There are schools that focus a lot on languages and not everybody is going to be a professional translator. And people who study algebra do not need to be. become professional mathematicians. Getting a technical education, getting into software is not just about becoming a professional software engineer. It teaches you a certain kind of thinking. It teaches you logic. It teaches you systematic thinking. It teaches you certain skills, hard skills and soft skills, and they are very valuable, even if you are not becoming a professional software engineer. So I am a strong believer that it still is very valuable to learn coding even without AI. Let's take myself, I like playing a game of chess.

I'm not very good in chess, but I like playing chess. And I like playing chess with my niece, for instance. We are never going to be professional chess players, but it is fun. It is fulfilling. it is a training for the mind, and this is what I do. I will never stop coding without AI my entire life. And I think if I'm 80 years old, I will still write programs in, I don't know, C or assembler or rust or whatever without AI because it's fun,

because it's like a crossword puzzle. So why not simply accept the fact that something that can be done with a tool like AI is also worth doing without this tool simply because it's fun and fulfilling and it is a kind of education for general thinking and intelligence. So that's the first thing. So whenever parents ask me, should I send my young kid to the school where you teach, I can just say, yeah, of course, even if they are not going to be a professional programmer, they will come out with a lot of interesting skills that they were able to practice by learning how to code. Yeah, I think that's an interesting perspective.

I mean, both of us, I think we both went to the same school back in the day's Hotel Leonding. I think you graduated a couple of years before me. But I also remember we started with 36 students in the first class. I think 18 made it through Matura and only a handful really stayed in software engineering. But I think most of them really, leverage the ground knowledge and the basic knowledge, as you said earlier, analytical thinking, right?

It was a great thing. We also in that school, we also learned a lot of the business aspect of it as well. So I think some of my friends back then also went into that area. Cool. Now, thanks for that. And I really just enjoy your positive attitude towards all of this, which I think more and more of us, we need to echo that. with it. I mean, Brian, what are you

think? Yeah, I was thinking, you know, I think you made a great point with that, you know, the, if you're going to work in a field, having a, besides the fun factor that you talk about, right? As you were saying, doing a crossword, I was thinking about playing Sudoku, which I used to do

a lot, right? But when you're, if you're going to work in a field, right, if you want to work in information technology, let's say, having an understanding of all the components makes you stronger in what you focus in. Right?

And just to your point about everybody who goes to like a dance school, my daughter goes to a dance school, right? She might not be a professional dancer. But one of the things I've told her is like, have you looked into choreography? Have you looked into, you know, running a dance school? Right, if you don't have those basic foundations of all those components involved,

especially what's underneath at all, you're probably going to have a harder time doing what you want to do in it. Whereas if you're that well-rounded person that understands and push came to shove, could you fix a line of code? Well, you know, yeah, of course it's going to be helpful. And I think the other thing that it teaches you, what you're talking about there, is the curiosity. Right.

And when, and I don't know if this is going to go into a point that you make later. So hopefully I'm not stealing your thunder. But what I found, especially on my AI journey, has been the curiosity of what can I use it for and how can I employ it to better myself and better my work output is key. So in your case, if you're learning how to code, you're learning how to do all this stuff by hand, that's going to then open your mind to say, all right, now that I know what I'm doing, how can I use this AI to do what I want to do? Right? And learning that curiosity is very important.

You know? Yeah, I can fully second that and I would even add something. It's a lot about also learning vocabulary, vocabulary, learning a certain kind of language. Because nowadays, now that we have AI, we can implement whatever or we can ask the AI to implement whatever we can name. So if I have an idea, I have to be able to express this idea, otherwise the AI can't implement it for me. And if I've never learned, for instance, programming and software and IT from the ground up,

I can't name these concepts. I can't name the development patterns, the design patterns, the languages and things like that, and then it gets very hard to express your intent. And with AI, it's all about being able to express your intent in natural language. This is something, Andy, you asked me what has changed, and this is something that fundamentally has changed. A few years ago, if you would have asked me, what other subjects are very near or related to coding? I would have probably said mathematics, and physics and chemistry and logic and things like that.

Nowadays, suddenly, we are talking about languages. We are having huge synergy effects with German teaching, with English teaching, because our programming language is clear expression in the language of our choice. And that is the same with IT. I am a strong believer that good AI users in the future will always need to understand coding from the ground up because otherwise they can't instruct the AI to build the things that they imagine. Yeah, it comes back to garbage in garbage out, right? And I've seen that firsthand.

If you ask poorly or if you think the AI is going to be able to read your mind, it's not. you know and i i think on a previous show andy i brought it back to the idea of like when people started outsourcing or development right they'd give this generic instruction build this for us and something would come back and they'd be like what the hell is this this isn't what i asked for and then you look at the paper of what they asked for and what came back it's like oh this is what i asked for because i was not specific enough i didn't know how to ask for what i was asking for and to your point either with language or if you're not understand or if you understand at least what's behind it all. If you know how to ask for what you want out of the AI, then you'll get a lot

closer to what it is that you're looking for. I would also say, I think, besides language, I think it's something that developers and stuff might need is a concept of usability, right, and user usability. They might not be developing the front end, but usually I would say there's somebody, since they're not writing the code themselves, they have to take into account how their service or whatever they're being written is going to be used or can be used or how it can be interacted with a lot more

because they're not getting necessarily a full-on and picture of what needs to be get done especially if you're vibe coding something from the ground up right developers at that point can be the entire architect and if you're not thinking from that architect interface usability point of view you might get the AI to produce something for you,

but it might have a terrible interface. It might be hard to have other services interact with it. So now a developer not just has to think, how do I write this to fulfill what I've been tasked to do, how do I create this to make it good and usable, which is a whole different skill too, right? So it's almost like a developer, yes, or not a developer, but someone working in this area needs to know how to or understand the code behind it.

But having that knowledge of the big picture, you know, the well-rounded, the Renaissance person, right, is going to be more and more important, I think. So having education in all these different areas is going to help a lot. Raina, quick question, because I had a conversation last night with a friend who is also, you know, a colleague. And he said something really interesting and I wanted to get your opinion here. Because I asked him, you know, how has AI changed your day-to-day life? he has been also a developer for many, many years. And he said, AI now enables me to do things that I was technically able to do, but I would have never been able to achieve it in the time that I had allowed it to.

Or it's just like I didn't have the skills to do certain things, even though I knew technically how it works, but not in that particular language with that particular framework. So AI for him on the one side really allowed him to do. things that he would have not even started sometimes because you knew for that particular language, for that particular framework, it would be too hard of a learning curve. But now he can do it. But then he said he's getting into a stage now where there's so much code produced and in

the first time in his life, he no longer understands everything, every line of code that has been produced that he's responsible for and he feels like he's losing control. And I wonder if you've observed this yourself and if you have. any thoughts on that. On the one side, the excitement, the dopamine, like, wow, I can now finally do things like that I've never maybe would have even considered to start, even though maybe I understand the technical challenge, but I would have never started it. And on the other side, now I have all these great things, but I feel like I'm losing control.

Yeah, absolutely. AI has significantly reduced the distance between my visions and what I am able to. to practically do with my time. So if I can envision something and I can express what I want, AI makes it much more doable. That's the first thing. That's the positive thing. The second thing, well, yeah, of course.

I've put things into production that I don't fully understand. Yeah, I did that. And I think we have to get used to that. And here it is a certain advantage that I am a little bit older. You know, I can remember when I learned programming, I understood assembler, I understood C, and I was taught certain rules how a C compiler compiles the C4 loop into assembler X86 statements. I knew that. I was told SQL, and I was instructed how to write a select statement so that the theory compiler generates a certain kinds of index lookup in the database. I knew all that.

And over time, this went away. Nowadays, if I write a piece of software in Rust, it's compiled to LM. A lot of optimization is going on. LLVM is compiled to Assembler. A lot of optimization is going on. And to be very honest, in all my Rust courses, I always say you have no idea what will effectively be created from your Rust code.

It's so far away, you simply live with the fact that there is a lot of stuff going on. And many people tell me, yeah, but that is a deterministic logic and not a probabilistic logic. And that's not true. If you take a look at the C-sharp just-in-time compiler, it really optimizes the generated assembler code based on runtime statistics. So it collects runtime statistics, and based on the runtime statistics, it might recompile your intermediate language code. into a different kind of assembler language because certain parts of your software are hot. They are executed very frequently,

so it makes sense to optimize it more aggressively. So these things happen. And I personally think that AI is just the next level. What we need to learn, and this is something where I think that also the education has to change a lot for those young people who really want to become professional software developers, we have to get a feeling.

an intuition of where we can trust which AI model and where we need to micromanage these models. What I always tell my colleagues who are teaching is that currently we are teaching AI, like if we would teach piano like that, here is a white key, press it, you hear a sound. Here is a black key, press it, you hear a sound. Congratulations, you can now play the piano. No, you can. because understanding the mechanics is not the same as being able to work with this tool. And this is what we need to add to computer science education nowadays.

We have to get rid of some of the old detailed knowledge that we simply don't need anymore. I'm sorry, it is out of date to teach kids or to force kids to learn certain syntax constructs by heart of certain. languages, they don't need it anymore. Things like that can be rigged out of the curriculum. Instead, we have to, of course, teach a solid foundation, no question about that, but we also have to teach to practice with AI. And they have to come out of a school with a solid understanding what AI can do and where they need to closely manage AI. And then I have no problem. put in code into production that I have never read. Because I know that certain tasks can be reliably dedicated to AI, delegated to AI.

This is what I wanted to say. So, another task, I need to pay very close attention to it because probably AI will often screw up in that area. Which answers another question that I had for you, but I think it just answered it. because the question that I had for you is coming back to people that want to, you know, become software engineers and obviously out of school, junior software engineer, but who would hire a junior software engineer if the code that they produce is obviously not as good as the AI. But you are basically telling me that we need to rethink on how we are teaching software engineering.

It's no longer, as he said, about learning the syntax of an if statement, in five different languages. It's important to understand the basic concepts of how code gets compiled, but it's more important to instruct or kind of conduct the AI to really understand how to best leverage it

and to trust it. And I really like what you just said earlier. We need to teach people and to practice with the AI so that when they come out of school, they obviously know how to best use and then also trust that tool. I had a question

though Andy going back to sorry did you want to say something about that you know a few several episodes ago I forget who we were talking to it might have been Chris LaBrato I'm not sure about the idea right coding language is a human interface

to the machine right and we're training these agents to interface with humans to write human code right but what about the idea of, and I don't know if this has come across your thought process at all,

you know, do we need to have the languages that are human readable anymore? If it's machines writing code for machines, does it even have to be in Java, in Rust? Do you feel it might go to a level where the machine, you know, the AI just starts writing machine code for the machine, because if the humans don't have to write the code, it can just get right to the source. And we cut out all the translations of, you know,

if 10 equals five, you know, or 10 equals five, listen to me. Yeah, I'm a developer. Hi. But you know what I mean? Like, does that, if the machine's going to write the code, the reason we have the language we do is because no one's going to be able to write in that machine level code, right? But if the AI can do it, is that something that it's going to start moving towards?

and if it does move towards that, you know, besides giving the instructions of what we want, how do we verify that it was done well and efficiently or that it's operating, well, not operating as expected. Obviously, if you test the stuff out, right, but if you would look at, let's say, something written in any language and you have a code block, you know, really, really big that could have been done in five lines,

you can say, obviously, this is really poor coding. But if we get to that machine level, and it's something that humans can't really understand or read easily, how do we verify that stuff? I think we are very, very far away from that. AIs are so horribly slow. They need so much resources to produce a relatively little output that we are very, very far away from AIs being able to produce binary code.

I think the big value that higher-level languages bring to the table is the possibility to have an abstraction layer where with a few tokens, you can express a lot of things going on. It's a matter of abstraction layers. If you take away these obstruction layers, the AI would have to produce so many tokens that you would have to wait forever until you get what you need. And the second thing is, AIs are screwing up way too often so that we can do that. Nowadays, we need higher-level languages in order to combine probabilistic AI with deterministic harness engineering. We combine tools like Linters and compilers and unit tests and all this stuff. We need them as a feedback mechanism for the AI. The AI is rather slow in generating tokens.

And then the AI needs immediate feedback whether these tokens make sense. And the AI screws up a lot of times. It needs deterministic tools to immediately get this feedback. If we take out this intermediate layer of a higher level programming language, the AI would need to do so much more testing because it's much harder to deterministically prove the correctness of a certain piece of software without the higher level language that the AI will take forever. to even generate a medium-sized program.

So maybe in many years, when hardware has gotten better by many, many orders of magnitudes, maybe then things will change. But it will take quite a while. It will take quite a while. Nowadays, I think we even need higher programming languages. We even need programming languages which have more. powerful abstractions so that the AI can be more token efficient. Where I agree to you is that the higher level programming language does not necessarily be so easy to read.

So if we are talking, let's create a programming language that captures, let's say, the essence of rust, but with way less tokens. I'm fine with that. As a human, I would not be able to read it, but it's not a sample code. It's not binary code. It's not directly to the source. But it is more token efficient. I can live with that.

That makes a lot of sense. Yeah, that's that I can imagine, definitely. And I think on university level, I think there are already research projects going in exactly that direction that I've read it. Yeah, I think also, Brian, I remember the discussion too. This was kind of the direction we were going. would AIs come up with a different programming language if they would not be constrained with what the human expects a programming language to look like?

So maybe we will design some new languages that are optimized exactly for what you just said. It is something that is higher level and abstraction, but it might be optimized in a way that it can be easier understood and manipulated by an AI, but with the drawback that maybe humans have a hard, have a harder time understanding it. Maybe if the language is so efficient,

we can fit it on a floppy disk again. Yeah, maybe. Yeah. I guess the question is also, when is the right time for us as humans to really also give up that level of control? Because then it's really, if you cannot understand and read anymore

what the AI is creating, we are obviously losing control. On the other side, you said earlier, we are losing control anyway because we end up in a situation where we just trust the outcome anyway, even because we don't have the time anyway anymore to read every single line of code that is produced. Yeah.

Two things. The thing that I came across a large interview a few weeks ago about FFMPEC. And of course, I knew what FFMPEG is, but this was the first time that I really dealt with the structure of the FFMPEC project. and if I see the keyboards behind you, Brian, I'm pretty sure that you are very well familiar with FFMPEC. And I became aware how much assembler code is in FFMPEC. And I heard those stories that for a certain processor architecture,

there are, I don't know, one or two people on this entire planet who intimately know this processor architecture and who can write the assembler code to do certain things on a certain performance level, And no AI on Earth can ever create what they create because the AI has no idea. So I think this concept of, let's say, artesian programming will be a thing in the future. In the past, we had, I don't know, some Japanese makers of, I don't know, certain pencils where there are only two people in the world who can create a certain artesian. I think in software, we will see these things too.

They will not write code in a certain way, in a unique way, but they will have certain knowledge that no AI simply knows about. And FFMPack was a really interesting thing. And I was not that aware of this project. And yeah, it's interesting to see these niche projects and to see that AI is still struggling a lot in certain areas. Hey, Rainer, I have. one bigger topic to discuss with you,

or I would like to get your opinion. And I think it's actually two questions, but they end up in the same direction. The question that I initially had was, how does hiring change in organizations when it comes to getting the next high potential talent, right, that they need? So I'm kind of like, I'm looking at software organizations, organizations that need to produce software.

Does hiring change? but then I also wanted to pivot a little bit if existing software organizations that have a talent pool currently in the organization, now AI comes in, what does or can an organization too to also up-level them? Because I know you're working with many organizations,

teaching them on how to best use the AI. So kind of like, you know, how can organizations become fit for the AI future in terms of up-leveling their existing engineers or also what does it change, maybe even in the hiring process? Yeah, this is something that keeps me awake at night, not for myself, because I'm established in this field. It's fine.

Until I retire, I'm good. But if I take a look at the young people that I teach, I often think about how the environment in which they start their career has changed. Number one, to say it frankly, certain organizations are behaving extremely stupid now. They are reducing their workforce. They are optimizing to a very large extent. So they are giving AI to the most senior software developers. It's fine. You can do that. But then they take away the entire time that the organization has for educating juniors, for developing juniors. And I think this will work out nicely on the short term because they will save money,

no question about that, but over the medium to long term, it will hit them hard. Because in the next few years, we all know that a lot of senior developers will retire because they are too old on the one hand side, or I also know a lot of senior developers who do not want this change to artificial intelligence anymore. They will mentally retire early and simply say, no, I don't care about that. I will do whatever I did in the last 30 years. And yeah, I don't want to deal with AI anymore. So I think that many companies who are now saving, I think that they save money by not educating and developing young talents anymore,

they will have a lot of problems in a few years because the seniors will go away. Maybe they will be hired by somebody else and there will be no juniors coming coming back. That's the first thing. The second thing, I think that companies must be, it must be clear that in order to get an employee that can really provide meaningful value takes longer than it did in the past. In the past, you took somebody who immediately after college or after a university degree and they joined the team and they could provide value to the product. This is no longer the case. If you come directly from university or a vocational college or something like this,

you cannot write code better than the AI does. It's simply a matter of fact. You also don't have the experience in terms of enterprise complexity. you are not aware of all the problems that you need to look out for, for stable enterprise level systems and things like that. So it is just a matter of fact if you hire somebody after school, they just cost money. They cannot provide a lot of value.

You have to wait for another few years until they are experienced enough to really provide value. And the question is, are companies able and willing to invest, the time and the money to develop these young talents who then become the senior engineers in the future. And I think there is a cultural difference between the US and Europe here. I know a lot of people in Europe who started their career at a rather early age in a certain company and who stay with this company, who stick with this company for decades.

And in many companies, that's the regular case, not the exact. And I think this will be super important in the future. If you discover a young talent who has the potential to become a super powerful senior engineer in a rather short amount of time, then you have to grab this talent. You have to invest in this talent early on. You have to build loyalty with this person that goes in both directions. and then over many years you will get back at investment. This higher and fire policy, you are simply destroying value.

I think this is, remember Rainer, when we had our last call a couple of weeks ago, I mean, you brought this up with the high potential, right? And that's, I had, what I had in my mind was then flipping around. It feels almost like if you have coming back to the example earlier, when you go to a sports school, you don't necessarily become an athlete, only the high potentials. So the question is, can you really identify those high potentials and then really bind them to you? Like you are identifying a high potential football player in the early days and then bind them to your club, right? Because it will take a while in years until they then really produce the value and the amazing output.

So if that is the case, though, then playing devil's advocate, we know that there's only, a handful, a small percentage probably of high potentials that come out of university, of high schools. What does this mean for the average pupil, the average student in the age of AI? They will probably need to combine their technical knowledge with some additional talents that they have. they will probably not make a living from being a professional software engineer. They will need to find a job where they can provide maybe other services or products where software or their understanding their basic understanding of software becomes an additional skill that they have in their repertoire.

are. But I think the number of deep technical software engineers necessary to run our economy and companies will decrease. But there will be a lot of additional roles in our companies where having a certain skill in terms of technique and logical thinking and so on will be very valuable. I think we have some of those already, right? We have the SREs and platform engineers, right? A lot of that requires that logic, some understanding of coding, because a lot of the stuff is done is code, but you're not writing the hardcore code.

It also comes back to the analogy, like with the dance school, like, okay, we still need choreographers. We still need the, or the iPhone, right? There's the whole iPhone ecosystem in products, right? So it's out of AI, they'll grow probably even more of that kind of ecosystem, I think. When I think, Raina, it reminds me still a little bit going back to my own years in high school in Hattel, I don't think. I know because not everybody stayed in my class, at least, in pure software engineering. Many did other things, but many of them that went into different directions benefited from having a good basic foundational understanding of IT.

Because obviously IT is everywhere in our lives. And then they obviously have the benefit that they've learned the basics and even they don't need it on a day-to-day basis. But I like the way you explained. find a different talent and leverage everything you've learned for software engineering by combining it with other talents because you will provide value. Let me give you an example. Today I was at a customer workshop. I was hired as a coach and we were, I don't know, approximately 15 people and essentially

a very large company and they wanted to practice for two days in a workshop, how they use AI in different fields in their organization. I was hired as a coach and I was jumping between different groups. And one group was focusing on the UIUX part. So there were professional designers there. And they asked, for instance, one of them asked me a question. And he said, you see, I use this tool. In this case, it was called design and it spits out certain HTML pages and so on.

But what do I do if I want to have 10 different suggestions from AI? How can I work on them at, at the, at, the same page, at the same time. And I asked him, do you know what it is? Well, he has heard about it, but never really. Do you know what kit work trees are? No, never heard of that. And it became clear to me, this person would benefit a lot from understanding, let's

say, basic logistics of source code and folders and files and gift and work trees and branches, not because he's a professional software engineer. He is a master designer. He thinks about UI and UX, but suddenly he would benefit a lot from having certain foundational knowledge about how to deal with code. And this is exactly what I mean. If I go to a technical school and I learn to deal with things like Git and continuous integration and basic technical things, and then I go maybe to university and I study art. And you, you know, design and things like that, I still benefit a lot from all the things that I learned when learning the fundamentals of coding. And yeah, this person would probably not be a good software

engineer. This is not his passion. His passion is using the design, but to a certain degree, he now needs certain skills that come from the area of software engineering. So does this, if I'm not shift this is the right term, but I will use it anyway. Is this the return of the T-shaped individual? T-shaped meaning broad, broad understanding, but in a certain area, like in art, in U.S. design or something else, you have deeper knowledge, but you still need to have that power on the top,

that T-shape, so that you really understand what is going on, what are the concepts that really make you then fully leverage, that deep knowledge that you have because otherwise you're limited. I think that's the, is that what you're promoting? Or did I get it right or not? In fact, not. If I would have to take a letter, I would say it's the age of the M-shaped person. What I mean is you have a very solid, broad foundation.

And AI allows you to go deep into many different areas. whenever you need it. It allows you to go deeper, faster when you need it. I always tell the story of my assistant. I have a technical assistant. She's only 17 years old. I hired them when she was 15, and she's one of these high high potentials.

And a few weeks ago, I asked her to join me at a conference in London. She did a talk there, amazing, at 17 in London in English, about some match, match. And I told her, you know what? You should talk about this infrastructure as code with Bicep and OpenDConnect and federated identity and automate everything with GitHub actions. And she didn't know a single term of what I was mentioning. And you know what she said?

She said, okay, let me look into it. And she spent hours with chat GPT, learning about these concepts, applying them. and getting familiar with them. This is not T-shaped, as in I have a broad knowledge, this is my speciality, and here I will stay for the rest of my professional life. It's more, I know AI, I know what an HTTP request is,

I know what computing is, I know some programming languages, I know software design patterns, so yeah, give me an AI a little bit of time, and I will grow these, I will drill down these holes, to certain topics whenever I need them. And I will be able to do that in a very quick and very efficient way.

This is why I was saying not so much key, but there are a lot of things that you can get when you need them. Well, so if nobody ever used the word M-shaped engineer, then I want to attribute this to you. Yeah, thank you so much. Awesome, yeah. Love it. Raina, we are almost at the end of our time here.

It's unfortunate because you have a lot of experience and a lot of, obviously, exposure to the young generation, the next generation of people that will impact and drive our engineering communities forward. Any final thoughts, any final recommendations to somebody maybe that believes that they might be in the wrong field right now because the AI is taking over their job or because they're just uncertain. Any final thoughts? Yet two things. The first one, there is one thing I struggle a lot with, and that is making these technologies available to kids who do not come from a social economic background that allow them to spend 100 euro, 200 euro per month into AI. If you are established in your area of work,

if you are a company or an entrepreneur or even a private person and you have the money to become a mentor to those kids, organize some sponsorship for them. That's a real problem in schools nowadays. At least in Europe, we have free school books. We have free school. We have free university. But AI and the costs for AI is really is really dividing the people and the young people into two different parts, the ones who have the money, who have the socio-economic background, to spend 200 euro a month into a clause, max pro, whatever it's called subscription, and the ones who cannot afford 10 or 20 euro monthly subscription, and that's, that really makes it different. So what I try to do, I have close relationships.

to many companies and I want to especially thank your company and the Dinotrace. They are helping a lot with our programming club. Whenever I have asked sponsorships for diploma thesis and so on, this is exactly what I mean. So I would like, I would really ask everybody who has the possibility to help there to become a sponsor of a school, of a diploma thesis, of maybe one school kid, sponsor them something, help them something. it makes a difference. And as a society, we have to solve this problem. We really have to solve this problem. That's the first thing.

And the second thing is to everybody who is still struggling with AI, stay curious. I always compare myself with my mom when she works with a mobile phone. She's a very tech-savvy person. She is over 80 years old. But sometimes if something doesn't work the first time, she gives up and asks me. And I, as an established engineer, I don't want to become this person trying AI. AI doesn't do what I want.

I stop using it because it's not useful. And we have to wait for the next generation who is not giving up immediately. So I think it is a great thing. If the young people who are not afraid of AI at all and the established people who are a little bit hesitant, we can inspire us. We need this young energy in the teams, but we also need the experience of the established software developers to understand things like stability and enterprise complexity, as I thought. And I think this mixture is kind of the magic source where great things can happen in the future. I can only, so we'll definitely make sure that this word gets out, especially the first one that you mentioned.

we'll use our connections also in the industry to remind the leaders of tech giants right now to invest in the next generation, especially in kids that don't have the financial means. And I love it when you say, stay curious. I think that also marks a great engineer because if something doesn't work, you just try to make it work, right? And you're playing around with it. Yeah, that's great. Rina, thank you so much again for the time.

I love your positive attitude towards change. I know I'm repeating myself what I said earlier, but I think we need this positive attitude, especially in times of uncertainty and change. So thank you so much. And all the best with your future projects, with your continued education in the different schools and in the Coda Dojo.

And, yeah, Brian, anything else from you? Yeah, no, just thank you for this conversation. You know, I just started a new role on the enablement team and part of in the back of my mind was always the idea of bringing my knowledge

to the younger people on the team and I think this has given me a whole new context to think about how to leverage some of this you know

the new way of doing things right and how to hopefully help train them to to do I hate saying do more with or to do more with more, right? Now that they have AI, how can we use that enablement to allow them to do even more

and be much more efficient at it, right? So it's given me a lot to think about. Really appreciate you sharing your time and your knowledge with us all. Yeah, thank you for having me. It was a really interesting hour. Thank you so much. Thank you so much. And enjoy the rest of your beautiful evening there in the only thing.

Yep. Thanks, everyone. Bye-bye. Thank you. Bye. Thank you so much.