Can AI Replace CEOs? | Innovantage Podcast #50 | Powered by BMI Executive Institute

Well, hello there and welcome to another episode of the InnoVantage podcast where business meets tech to bring you your competitive edge. As always, I'm your host Max, and it is my ongoing job and duty to explore that edge between business and tech and to find out something interesting for you to listen to. And today's topic is a very special one. It's a contrasting one, one that I rarely ever do. We're going to be having more than one guest and we're going to be having more than one take on innovation and very different takes on innovation within the public sector and the private sector within different areas of, well, business and tech. To help me on that journey today, one of my guests is Jonas Valuta, who is the Head of Asset Management and Sales at LTG Competence Center. And Vydas Vyšidauskas, who is the CEO and product creator at Teamhood. Welcome, welcome guys to the show. Hey. Hey. So Vydas, since you're across from me, let's start with you. What's your background? How did you become the CEO? What does product creator even mean? Yeah, so to make it short, probably product creation or in general, the word creator is the essence. That's my background. I like to create things. I'm an engineer by my education, software engineer, but in general, it doesn't matter for me whether I solve problems by building something, painting something, or by organizing and leading people to build something. So, in essence, that's the same. And that's why I prefer that product creator title, and I love to create digital products that are used by various peoples and just, you know, make their life easier. Makes sense. And what does Teamhood do? So my company Teamhood, we're operating in project management software space and we're building a product that bridges the gap between simple software where there are many tools in the market and there are very complex tools like people know, like Jira, SAP, all that stuff. But there are a plethora of companies that require something of both, like this golden middle, still simple, yet already a bit powerful. And this is where I want to plug the whole bridge the gap of the market with Teamhood. Makes sense. Makes sense. A startup through and through with a revolutionary idea to streamline things that people are tired of. Jonas, what about you? Hey, hello everyone. I'm Jonas. I work as a head of asset management and sales, and my experience in LTG company comes from 7 years of work. I started as a project manager and I saw all asset transformation, how the company split into subsidiaries, then those subsidiaries implemented the SAP system. So I've seen the journey and I see the hard journey ahead that is awaiting us. So a lot of challenges. So what does SAP What's that management within LTG do? Like, for our listeners who don't know, LTG is the national railways of Lithuania, right? Yes, it's a Lithuanian railways company. It operates as a holding. It has 5 subsidiary companies that are responsible for different business models. For example, like railway infrastructure, freight, passenger transportation. So what asset management is, it's like the whole legacy, because LTG is a company that's over 100 years old and it started in 1861 when the first railway tracks were introduced from St. Petersburg to Lithuania. So we have a huge compiled asset portfolio in our backpack and we have to manage it. It's everything from the railways, from the buildings, from the rolling stock and so on. So you name it, it's a very asset-heavy company. Oh yeah, I can imagine. I can only imagine the amounts of assets that exist there. So then coming to the discussion of our episode here today, innovation. What is your take on innovation? So Jonas, let's start with you this time. What do you think? How does innovation fit in into asset management with LTG? Well, since it's a, we have already established it's a very asset-heavy company, so unlike tech innovation, you have to do it incrementally. You have to analyze all the base that you have, and then you have to take all the legacy systems that you have, and you have to master all the data from them. And then you have nowadays built transitional mechanisms to integrate into modern datasets. So that is like one kind of innovation from data point of view. Another part of innovation, for example, is like a corporate innovation strategy. You have different business models. You take those business models and you see what those business models encompass, which asset basis, which asset structures, and then you try to update them or make them work differently. So you cannot shift the switch and in one month build an MVP, minimum viable product. Launch something new. So lean startup methodology is different in a huge corporations. Maybe you can also allocate some different funding, like separate funding for innovation. That will— we do also at LTG to do like pilot projects. So it's different kind of ways. But if you imagine you take a huge asset base, you do some optimization on some innovation, you change like 2% of impact there, and it translates immediately to hundreds of thousands of euros. Oh yeah, that whole process does not sound like it's too fast, right? It depends. Okay, yeah, it depends where you look at it. So we have different teams, it's a huge corporation, so over 6,000 people, and I'm only responsible for one part, and that part exactly is like real estate, real estate data management in the SAP RFX module. And also, so you see, when you want to like do a fast innovation, if we take like AI, we can do like this. We have a team from HR that made an inside pilot project and they made open AI-based tool for HR queries. Yeah. So they did that, they developed the proof of concept in one month, and they then went into the corporate machine, got all the approvals, and they started, they launched it. So that is one way putting it. When you take asset innovation, for example, you have a train, train has a camera, The train is moving through Lithuania and he films everything. You can integrate the visual data and you can calculate how many passengers are getting into train, what's happening around the station. So you can translate those previously datasets that were unused and you can use them in a different ways. So also another example, like for example, those acoustic walls when you have trains going through the city. And have acoustic walls to reduce the sound. We put solar panels on the side, so we generate electricity. So it's like a side benefit also. So very— who would have guessed having a lot of assets helps in building new innovation? But Vytas, what about you? You're a startup, you're trying to disrupt an existing marketplace of existing solutions. Asset poor. Asset light. Or digital asset heavy. My assets are in digital world. So sorry, I jammed. So the same question about what it means. Same question, yeah, like how do you face innovation where you don't have as many assets? I would say it's universal and I think there is no big divide in what type of company you are. It's just different areas that you can probably attack with the innovation. In essence, everywhere you are, it's, I think there are like 5 things, I don't remember that well, but Either you can make something faster, cheaper, or you can take something from one area and apply it in an area where it was never applied. And in technology space, it's mainly like different ways of working. So for example, like in my specific case, it's we want to be way faster to market because competition is crazy in digital space. I'm not competing with a few companies in Lithuania. I'm competing with hundreds of companies or thousands of companies across the world. Of course, we can discuss what is competing and geographically, culturally, but in general, that's the reality of digital world. So for me, being faster than my competition is like I always say to my team, in the triangle of projects, like cost, quality, and speed, for us, speed is always on. There is no discussion. It doesn't mean quality needs to be crap, but Still, speed is the essence. And it doesn't matter then what is the underlying innovation itself. If we have goals like we want to ship more value to our customers or we want to reach more customers with the same resources or we want to be efficient on how we like do marketing maybe to reduce the budget and invest it in our R&D products. So that's just, you know, again, something's faster or cheaper or just simpler in different way. That's a good point because, um, Jonas also mentioned speed and it can happen in LTG, but I have a feeling that with a startup, the level of speed that you would be used to is still much higher than a larger organization can make it. So like, what is a fast implementation for you? That's a good question. I would like still to place a remark. I think we should not treat that company size defines the relative speed. And I worked in a bank, in Danske Bank in Lithuania with thousands of employees the same way. And I can confirm that I was able to achieve not like exactly the same speed as a startup, but very close. And why is that? Because there was freedom to act. Yeah, the resources. To take decisions. And the less stakeholders you have, the faster you move. That's the simple reality. So of course, small company has less stakeholders unless your product is, you know, used by millions of users, then those users are stakeholders and then you're a bit slower. And my competition, like famous Jira, they are slow. They cannot punch in a way we do. Though, so establishing that part, right? So speed is not an issue if you're willing to do that. But in terms of what is fast, Shipping things in hours which are impacting users. That's the speed. Oh, wow. So you can do that and it's not an issue. And in highly competitive market, people do recognize it by examples. Like a user comes in, says, it doesn't matter to one of my employees or me, I still speak to users and they say, okay, Vidas, there is an issue in the system due to time zone saving time. We have this calculation mismatching. And what do you know, I passed to the engineers and they already know what's the problem and we can ship it with a few button clicks and that's it. And user says, wow, I've never seen that. That's awesome. So that's what a small company can do. Yeah. So Jonas, I'm guessing even the fastest of your projects cannot react within hours to user feedback still. I don't want to say anything bad about LTG, but that would seem out of the realm of possibility. Well, let's not jump to the conclusions. It's a giant but still has some moves in it. So we also have modern systems for our clients where they can get direct feedback about our services. Of course, if we need to change something, we already immediately go into the asset world, into the physical world where you have rolling stock, you have railways on the ground, buildings. And to move things there fastly. So it's a different kind of game. So we have also IT systems we can change faster there. So yeah, so digital is easier. Yeah, let's take it for granted. I mean, digital is easier. It's very easy to push the bits, you know, instead of pushing real bolts and bricks outside. True, true. But from my conversations previously, I've noticed that for physical businesses, it also kind of instills a certain tempo, a certain, I don't know, culture, I guess, of working. You know, you are limited by the physical speed, so you adjust the digital delivery culture to that physical speed. And the way that you develop the business becomes kind of dictated by that limitation where you needs to spend more time thinking about the next feature or, you know, collecting more user feedback than just reacting like this. And I'm not saying that this is inherently a good thing, but it has its positives and downsides, I'm sure. So again, we are tied on the speed. Huge corporations can be fast if they need to. Startups are fast by definition. They must do that to survive. What is then the next thing in innovation? If speed is out, we have still the quality and the price. So what do you think about, well, you mentioned quality, right, is still a factor with this for you. So how do you maintain that quality and what is kind of the measuring level that you use for quality? Yes, very good question because you cannot just have everything right at the same time. That's the magical triangle of projects, right? So you choose two usually. And that's the crazy game of balance. And it depends on the time where you are at. If you're building a proof of concept, and usually that's what I would go anytime, and I'm doing it right now as well. I go full speed and then lower price because I need to go to push something through the door, get feedback, understand whether this is going to stick or it's a complete nonsense that we figured out and nobody wants it. So I need to go back to the drawing board. So having fast and cheaper is better in there because if the idea is that good, we have plenty of examples. People can line up, queue up and ask for your product even if it's, you know, breaking or cracking. But if it's the lone or sole product that offers this groundbreaking or different approach that's valuable, people don't care. Same can be evident in like AI tools that all of them are treated like they built on the AI. AI means fast, but also that means lower quality. They're shipped very fast because competition and whatnot. And all of them has evidence of poorer quality than usually crafted product, which took few years to polish and spit out in the market. True, true. I just today I watched like a wouldn't call it a documentary, but like a series of videos on Open Claw, you know, the whole journey of how it was created and how it was implemented and the initial user feedback. Insane, right? That speeds and the levels of like how you can just speedrun the whole startup lifecycle. It is just insane. Like, like that's, you're just fidgeting around with your own stuff and then you're bought by Mark Zuckerberg because Well, because he likes you. But in any case, why I remember this story is because the quality in and of itself is kind of there, but the security is not. Like the whole, the main thing that it is being criticized for is how vulnerable it is to outside influence. And I feel that since you mentioned AI, I think that's one of the things that are underreported or underconsidered, especially in the investors. Point of view of how secure it is. Those could be divided, like quality and security as like separate characteristics, non-functional characteristics. But in essence, I think it's very close. It's the same thing because if you, even in the physical world, like if you buy a phone and it explodes, is that a security testing quality issue? Where is, what is the, was it hacked because it exploded? Nobody knows, but the quality is crap, right? Yeah. So it ends up in the same, you know, junk box of perceiving how people perceive the product. Indeed, indeed. I would agree with that. So Jonas, then a question to you, on the same level, do you still struggle with a larger organization between those three portions of the triangle? Do you still have to choose between something that is quality, fast, and— We live on the same planet, I guess, so like everyone else, we adopt the same laws of physics. What a joke I can tell you that I use LLM and they made the circle into the triangle into a circle and they did all the lines in the circle around. So you do everything together, but talking more seriously, yes, of course. And when you mentioned the security aspect, LTG, it's a number one thing. Because you operate as one of the biggest companies, state-owned company, and the security is of utmost importance for us because we also military, military, huge interaction with them, with the logistics. So our data must be secure. So it's the top, utmost priority. LTG was one of the first companies in Lithuania that adopted AI Governance Act and released it. So for all the employees and all the outside partners to just to make sure and validate that we comply and we bring quality by security. That's also our selling point. So everyone can be like, learn from us because EU AI Act is going to be implemented and going to enforce in this year, August. So everyone's going to have a very interesting time who's now living by different rule set. Unless it's postponed. Yeah, well, already preparing for it. So yeah, that triangle also works for us. And I guess what we do, we search for the utmost value for our customer. Maybe we can shift like one pole for triangle, but our customer doesn't need it, but he needs those two. He needs time and he needs a price. Then we can maybe reduce the quality and so forth. So what's the best for our customer that to do. Yeah, that makes sense. Instead of taking the train, you can just walk along the lines, you know, you'll get there eventually. It's going to be very cheap. Yes, something like that. Also, our trains have like different value proposition. You can buy the first class, the second class tickets, so there's options to choose. True, true, true. And I guess again, like the wealth of assets gives you a bit of a flexibility when you approach— Yes, and huge responsibility to manage them efficiently. If we manage those assets efficiently, we increase the value proposition for our customers by reducing our costs. So it's a huge strain and a huge responsibility for our company. Then, what comes to mind to me is the question of governance, you know, enterprise governance specifically. You mentioned the implementation of the AI Act as one example. Like external pressure, regulatory pressure, it's felt much more on larger organizations than it is on startups. And again, like you just mentioned, it's much easier to adapt faster to anything changing in the market than to be prepared for it in advance. While LTG or a company of the same type will probably need to be planning in advance in order to be ready for something. Like this. But we've all touched a little bit on AI, and I want to ask you, what's your general feeling and opinion in your spheres on AI? Is it the next big thing that will transform your industries, or is it overhyped, or is it something else? Like, Vidas, what do you think? It's done. I mean, it's already done, the transformation. There is no going back. I would say it is overhyped as well, but not in a bad sense. Maybe it's usually, you know, when some business area becomes hot, it attracts a lot of smart people. And I think AI transformation attracts all the builders. I'm not saying people like me, but people who like to solve problems. Yet they always didn't have full capability and power or time to solve those problems. But they're passionate about it and they are everywhere, probably from someone working in a factory to someone working at the workstation in the office. Yeah, it's just their passion. And now the tooling allows to do way more. And I was always fascinated, like we have customers in the US how different thinking they are compared to customers in Europe or in even more conservative countries. And there are many examples what is conservative. Like somebody can start by like, oh, I'm afraid of some technology because it has a risk. While in US it's being cherished and taken without thinking about the risks maybe upfront. Because it's so cool to build something. And that means people get more time or more power to do what they love and they will work differently and everybody is going to get exposed because other people will see, okay, so, you know, now there is this guy or gal who is doing this and I'm not doing this. So what's happened? What happens with me? Business will see it the same way. This tool can output amount of X people and not 2X's way, like 5 times more. If you're really serious about it, if you're just ChatGPT prompting, that's not AI. That's like, you know, scratching surface. Yeah, exactly. It's just a faster horse. But if you want a car, you need to go deeper where actually AI does the work for you, like really. But still humans are in the loop. That's where the quality comes in. So for us as well, we're now just discovering, like my company now creating a new product version of Teamhood, major version from scratch. Because we said we need to be AI first, like not AI on top. Our tool needs to be like a database that is transparent to the AI and they can, you know, do a lot of things with tooling. And I think what we're now discovering is How much do we interfere? Because I think as a lot of people talking, coding is solved, design is solved, product management is solved. I don't think humans are that necessary anymore. But where the human comes in, like who ensures the quality, who ensures the AI understands the big picture and builds towards the right direction? How much do you need to tell? Like how much do you need to change the way you work? And I think my reality, you need to change a lot and it's either all in or just a bit will be a very painful and slow process. So what's the major difference then? Do you need to follow the example that you've given, just rebuild the whole solution from the ground up if you have that capability to be AI first, or is there like a slower transformational process that people can go through? There are many playbooks probably in this case. You don't need to invent a new one. Like when I was working in a bank, there were always multiple options, like rebuilding the system, cutting system in pieces and building one piece at a time, changing facades and putting new stuff under the facade once it's working, you know, many options, layering like a cake, you change one layer, another layer, but it's up to the organization, right? How much money they want to invest. Sure. Do they have the skill set to go big boom or they need slow, you know, less painful process? Do the people, will people accept the change? And in large organization, that can be very challenging. True. It's not an issue by given, but I mean, come on, you come in and from one day, one day you say, hey, this engineering department now needs to marry to project management because actually these are same person. It's the same person now with AI. It can be the same person with both angles if your approach to things is on the line with it. That's a big if. So then, Jonas, what do you think? Are you on the same page here? I was very anxious to step in, so I'll just start by giving two examples first. General AI, like no SOE. Now it's best to be in the shovel business. So you sell the shovels and give them to dig for AI gold and you then become rich. Talking about SOEs, LTG, well, there's like organizational levels of AI adoption. There are 5 of them and first is like exploration. People know about AI, they're skeptical about AI. I bet you heard about it. Second is like you have some most valuable players, MVPs, they do some proof of concepts, some pilots. The third layer is like you integrate, you try to build your like AI departments and so on and so forth. So where are we at? We at, in my opinion, at around somewhere second and above stage level. We already passed the skepticism. I'm talking from my like small circle. Sure. And 6,000 people, I cannot speak for everyone. And from what I see, you're not working with AI. You're working in a change management position and implementing a new software for people to work with. That's how I look at it. And from the basis psychological point of view, you have to roll with it because imagine before you have like a fax machine, you have the calculator on the table and suddenly bring a computer. Huge keyboard, you will work with this now. It's better. And we still have friends who work like that, right? Of course we do, but those are epic guys, old-school businessmen. They need to do that machine calculation for them. It's another topic. So you bring the whole change management curve. Maybe you are an MVP and as a manager and you go to a guy, okay, we'll do it like this then. Come on, man. We like to do it the other way. It's the best way as we always done. Then you have to do all the calculations, the usual thing, and it takes time. It takes time as with any other software to adapt. What's different this time, in my opinion, the slower you do this, the less energy you put into it, the you won't get those gains because when you learn something nowadays, after one month the technology might be changed and there's the learning curve must is a lot steeper. It goes faster like with other systems before. So you have to force yourself as a manager, as an organization, like Microsoft CEO Satya Nadella said, you don't be the know-it-all guy. You be a learn-it-all guy. And every day you go with a fresh perspective and you try to adapt somehow, think about it. For example, in LTG, we have around 200 systems. Let's call them systems, IT systems, right? It's huge, huge from signaling to internal to whatever. Some of them are new, some of them are old. Some of them running analog code. So you have to build up some transfer systems that would recode those things. Then AI can work with them. And it takes time, not only with people, but with the asset-heavy stuff that we have. So it's a huge process. It's going to be a huge undertaking. And I see a lot of work, especially with data management. And I have some good examples of how to do it. Like what? What's your good example? Yeah, for example, that HR tool. That's an interesting story. For example, they uploaded like internal processes about HR, like hundreds of documents. And imagine those documents were written in a different time and they had like some kind of pictures inside it, some kind of schemes that were hard for AI to read. And those documents were like bad. It wasn't usable. You couldn't get any output from it, but you had them. So what happened then from that reverse engineer? The team went and fixed all the documents. That was like a benefit from adopting AI. You streamlined, you smoothed your processes, so it's better for AI and better for human. So right now, in my opinion, when you're working anywhere, SOE, different corporation there, when you try to adopt a process, a new workflow, you not only have to think about human interaction, you have to think about AI interaction. How would AI read this process and how can you like implement into to it. So it's a huge undertaking. So is that— does that mean that you're going to change all of those processes within the HR department, or— They already changed. They already changed. Team is working it, iterating it. So those old ones are changed already. Okay, okay. Talking about speed. Yeah, okay, okay. So then at the end of the day, we all accept that, you know, AI is here and it is already impacting innovation and transforming how we work then. What do you think, we talked a little bit about the AI Act and generally speaking, Europe is a much more regulated space than the US or honestly any other market like China or India or anywhere else. Does that mean that we are going to be losing on speed here in Europe versus everybody else or? Do we still have a chance to compensate with the other two parts of the triangle? Didas, what do you think? I think that's not gonna, like, obstruct the speed in general with these, you know, safeguards or guardrails. I think Europe has its own thing. Like, in Europe, we value quality of life, quality of things. Being more, uh, how do you say, like, we want to give our civilization, uh, a better world by placing some safeguards. And there are very good examples. Of course, there are very bad examples in that. So, for example, so one, one on the good side could be like, if, if you look about like, uh, what kind of laws govern what type of food can we import, what kind of, you know, things that should go into the food. We are pretty strict compared to US. Is that a bad thing? I don't think so. I would like to be healthy, actually. I don't want to eat stuff which is not really cool in 20 years for me. I get cancer for it. But then each day I open a browser and see a GDPR cookie notice, I cry a bit inside because it's Europe's delivery of innovation same way like a bottle cap. It just annoys you. It's nice and it had a good idea, but the final outcome, we can just delete all those cookie notices. Everybody knows it, who cares? So a balance. And with AI, if we are ambitious in business and ambitious to compete, the governance doesn't obstruct that. It's more about How many ambitious people can we gather in one room and how much freedom they will get from financing something and getting the right people to work with them. I think that's where it comes from. And we do have like Lovable, for example, that's a Swedish startup. Come on, it's coming from Europe, right? Of course, it's already in San Francisco because that's where you get skills and money. Yes, yes. And they work differently. That's the sad reality. I go to kids and go to ride a bike in a park. In the US, they keep working overnight. They don't have vacation. It's our choice. What do we want out of life? Yeah, yeah. I have a lot of things to comment on, but, Jādas, let's hear your take first and foremost. Well, EU AI Act. Thank you, Vydas, for mentioning the bottle cap meme. Also, it's my favorite. What I get from this— Europe, it's a heart of our Western civilization, as we see from this. So we work from a huge cultural background, and this background gives us deep historic knowledge to make things sometimes complicated, but in other times, these complications might benefit us all in the future, like Ritas mentioned. Where I'm getting at. AI Act, we're innovating in EU legal acts. We're innovators in legal in the whole world. Stick with me. So we, because everyone else is not focusing on it. It's just wing it, see if it sticks. We are innovating in legal and we're building the foundation the foundation that would be acceptable for all the businesses, the clients, the people for a sustainable AI adoption. And I guess you have to start somewhere because from some examples, things are getting maybe out of hand with data leakages, private stuff that is out there, some deepfakes that are also affecting public opinion. Affecting elections, and we have to counter it somehow because some people are spending like millions into this stuff that is, it would be illegal, like in normal legal world, it would be illegal. So I think this huge foundation brings us a huge security and a framework that is understandable for all. We are humans, we need rules. That's it, I guess. Yeah, that's, that's a good point. That's a good point. So talking to both of you, which you both mentioned in one way or another. So like, I personally like the structure that regulation gives you, you know, first and foremost, it gives you a guide on what to focus on. So you mentioned cookies, for example, at the very least, it raises awareness of the general public that such a thing exists. And what this thing can be used for. Personally, I mentioned it several times on the show, I think. I like it, not for the user experience, of course. Nobody likes opening it up every single time or clicking and clicking. But I like that it's there because, you know, I have the option to opt out. Because every time I'm outside of Europe, that option is suddenly gone. And that, you know, I remember the pop-ups and that's still somewhere in the back of my mind where, okay, oh my God, I just need to imagine how much stuff are you currently collecting about me? And since I am a little bit in the tech field as well, so I see implementations that are irrespective of cookies, and then you just have like just complete screen capture of every single user that goes into your website. Do the users understand that they consent to that every single time that they go to your website? No, of course they don't. Of course not. But with this, they at least have the option. And when I look at the AI Act, I also see those levels that you just mentioned. And I see that, okay, yeah, organizations can make at least a choice, a conscious choice for themselves of which level they want to address and then which level theoretically they will exist within. For example, just what I see ahead, what will happen maybe in the procurement contracts and tenders will be like, some notes that you must comply with this act. And that's how maybe Europe will protect its market somehow by having those rules and make those products special. Yeah, yeah. But that's absolutely like straight to the point, good comments on the end value. That one I totally agree. UX side, I— That's the problem of the governance. It's budget-driven. When you get budget to develop new laws, regulations, there is no budget for marketing usually, like in business it would be. And the point where people get to understand why it is the way it is, why we're doing it, is hard. Because for me as well, when I first time heard about the AI Act, I thought, okay, okay, so yeah, let's put the brakes on. But then when I get to read it, and I saw a very nice presentation. What are the benefits of it? If everybody could see it, then it's nice because it might not be for me or maybe for me, but it can be for a certain group of people. For example, these scammers that call you. Oh, yeah. Our parents are at risk. And now people can mimic your voice, your video. Everything can be mimicked or faked. And if there is no regulation, there is no way to prosecute that, first of all. Like, who told you that we cannot do fakes? Yeah. Common sense. Where is that in the legal books? Yeah. So the act would establish that basis where if you're not marking content as AI-generated, you're exposed to law where you can get charged for that. And that's actually the end value. Protecting, like, as you both said, protecting people. Yeah. Yeah. It also speaks to that point you mentioned about the cultural differences of, you know, Europeans like to leave work at like 5 o'clock and spend time with their families. Americans, on the other hand, maybe would like to do something like that, but they don't have the option to. So the same goes to AI implementation and innovation. Like, what even is innovation, right? Is innovation replacing every single job with an AI tool that does its work probably worse, but faster and cheaper? Is it truly innovative or is it just profitable for the shareholder? Exactly. Hard to say at this point in time, right? The market will balance itself out over time. True innovation will probably stay on top while fake innovation will just die out with time because people will not use it. Charlatans are everywhere when it's hot. Yes. That's easy. But with the speed that things are happening, like when I mentioned Open Claw, like so fast, like what a regulatory body would not be able to do anything in that sense. 3 months ago, nothing existed. Awesome. And then it was bought by, again, Facebook, one of the giants in the world. And you can imagine. And it might not even exist in half a year. Yeah. So regulation would be pointless. Absolutely. Absolutely. So yeah, it's a choice. And I personally mentioned several times that I prefer the kind of, not the slower innovation, but the thoughtful innovation that Europe is able to do. But at the same time, sometimes we are lacking in some sorts of innovation that other countries and other regions— You win some, you lose some. Yeah, you cannot be the best at everything, right? You have to make some compromises, and even with the bottle cap example, right? Is it annoying? Yeah. Does it reduce waste? I hope so. I haven't checked the numbers, honestly. I don't know the numbers as well. I know when it sticks in the eye. I hope so. That one I notice. You rip it off, usually. Oh, really? Yeah, I don't do that. Because then it's even more defeated purpose, right? Yeah. Anyway, we mentioned people that are inside of innovation a little bit as well, and they are the driving force behind innovation most of the time. You know, most of the interesting ideas come from people just throwing things at the wall and seeing what sticks. For example, in your HR department example, like if people do not raise that capability, nothing gets accomplished. And with this, you mentioned that basically most parts of the software development problem, software development process are now solved as a problem. And now basically there's talk of, you know, one-man founders who can make the next unicorn on their own. Recently, two men made over $1 billion company in the United States. Great. Which one is that? It was some kind of, it was a man and his brother. They did everything with AI. So it's like a, telemedicine, some kind of stuff. It was crazy. Two man, one man billionaire. Yeah, so we are three behind this table, we should already have like a billion and a half at the very least. We should do that. So what's your take then on the role of humans moving forward? So Vitas, you mentioned that it's basically like you need to do basically more, right, with the tools that you're given, but how do you see it moving forward? What do people need to prepare for? I'm not sure because it will be way different in, I mean, 3 months ago it was different. In a year time it will be crazy different. And whether humans will be still, you know, looking at terminals and prompting AI or whether they will be, you know, lying on the couch and just saying, oh, that I like, that I don't. And then AI goes and does everything. We don't know yet, but that's changing. I think the biggest problem now is some people who are really excited about it have this problem of fear of missing out. So they have anxiety about missing out things. And that's actually a very sad part. Then there is a group of people who fear the change. As you mentioned, some people can be like against the change because that's how they work. For change champions, tens of years. So they have anxiety of job security maybe. And then people who still cherish the change and they are diving deep, they have a problem of anxiety of context switching because at least for me, I see this is human's limits. I mean, AI makes us faster. Extremely faster, more stuff can be outputted, but who's going to consume it? Who's going to accept it, verify that it's still on the same level as a human being would make it? That's still not solved that well. And the human in the loop is now becoming more and more expensive, I would say. So I feel like if we code, we can code multiple things at the same time, but can I review those multiple things at the same time? No. 2 max. If I go 3, I will start doing sloppy job on one of them because that's my speed. That's my— So you're the bottleneck. Exactly. Exactly. So should AI replace CEOs then? In Terminator movie, I think they did. And it doesn't matter if it's CEO or not. It's like all the management who are, you know, accepting decisions, verifying, confirming something, because it's not that much different. Like before you had like, so you have 100 people working in the department, right? Like Jonas has a department and he has middle management. He would not survive without middle management. He would die from like all the problems coming to him every day. Da da da, there is this and this, there is that. You go nuts. And I know I almost did that myself, went nuts because of that. So the same with AI, like how many things you can run alone? Will you have AI for AI? Maybe. I don't know. Theoretically. Theoretically, that's how it should work. So it's just, you know, same structure, just maybe some layer will be replaced. Human is still somewhere and quite a bottleneck. True. True. I do remember a quote, like vaguely remember, I don't remember who said it, but basically a computer cannot be liable for the decisions it makes, and as such, it cannot be the one making the decisions, basically. So it cannot be in the decision-making roles, at least now, and it's an older quote, it's like from the '70s or something. I would bet. Yeah, so like— And how is it working like that right now? Yeah, now we are getting much, much closer to that. Going to the border. And so that's where regulation comes in. That's where we both mentioned, right? Like, if something happens, who is liable? If the AI can be liable, as an example, probably a company that supplies that particular AI or the people who manage it, then that problem is basically solved. Yeah, it can make a decision because it is responsible for that decision. And sometimes it can make a wrong decision and it can bear the consequences. Of that wrong decision. I really fear one thing to happen with all this innovation and technology where humanity loses ability to think. That's like this example, you know, we like to joke about the innovation in finance where you can buy something now, pay later, and in US people already have debt on Starbucks because they take coffee now but pay in 5 installments, which is like 50% more the price in in few months. Insane, right? But that's not driving nice things to humans. It's driving the opposite. Same way like we hear about advertisements about gambling and all that stuff. That's BS, right? It's driving wrong things. So with AI, do we lose ability to make decisions? Oh, you know, ChatGPT, should I get a Starbucks flat white or large cappuccino right now? Then you wait. Ah, it's cappuccino. That's what I fear. I'm afraid it's already happening, unfortunately. But, Jonas, what do you think? Like, is the same type of thinking applicable to a much larger organization where, like, is a future where the Lithuanian railways are governed mostly by artificial intelligence feasible and possible, or is that just too far away? Coming from Lithuanian railways, I have to emphasize that people is the biggest treasure of all because we work in such a field that we need, we are in need right now of experts of railways because this like profession is in a decline and we try to proactively recruit people to work here because it's not like a high fashion bliss profession if everyone wants to. And you don't mean programmers, right? No, no, not programmers, but all the people, other, because of different business models. We need people. Without people, Revoist wouldn't exist. So we treasure our people. And where do I see like how AI comes into this in the future ahead? I see all the data, like the mundane work that you have to like do in Excel, do that, do that, do that. It becomes so integrated into all the systems and automatized. So you have people who are experts in their field in railways. For example, they like manage a yard, railway yard, and they see that some things, AI gives them suggestion that you need to improve something and they do it. I see input like this in the future. We just need to wait a little bit. How much? How much do you think we need to wait? Well, it depends. I think on some cases things are moving fast. For example, we have huge data assets already. Few years to get to the level where we can do everything with AI, to do all the suggestions, calculations. And we have other legacy systems that we still need to work on. So it's like this, so from a couple of years to like more. But I'm guessing AI won't be able to cover those people that work in railways on the ground, you know, the ones that build the rails, maintain the trains. Yeah, I wasn't talking about that, but we also have some innovations in those parts where we have like huge railway yards and we use like Maybe you would call it like operation optimization or something like that, but it's also an innovation. We use like drones with cameras and we have to manage our yard somewhere. Also do— did a pilot project and use a Boston Dynamics LIDAR robot to inspect the rolling stock. So some of the parts, yeah, but you still need people, need experts, and you will always need them because you work in a physical field. The land might be different. The robot could not go there. Sure. Yeah. Maybe there's bad weather, drone cannot fly and you still need to inspect the yard. So different business. Yeah. So I'm guessing all of the product managers and project managers and mid-level engineers who are not the best in their fields are just going to go and requalify into, you know, railway maintenance. Doubtful. There's a lot of places to work. Like, for now, let's see in a couple of years. Here's what happens. Yeah, but, well, from my personal view, I guess I get sometimes anxiety from the AI stuff because I want to learn it all. I have a special prompt that updates every day at 8 PM to get me all the AI news, what's new out there. So I'm in that kind of mindset. I want to like— You have a FOMO. Yeah, I have FOMO. That's the attention economy that's driven by Silicon Valley and everybody else. Everybody's fighting for a second of your attention. Even now, like even we are, that's what we are doing basically. We're producing content that we hope is going to be consumed by people and we are fighting. We're contributors of our own anxiety. Yes. Yes. And it's probably going to be a short somewhere where I'm shouting like, hey, you stop looking at your screen. Listen to us talk about innovation. Stop scrolling. Stop scrolling. Yeah. Should we stop? No, let's do it. There's enough space. The good thing is that there's enough space for everybody's content. Maybe not at the top, but somewhere along the line, we will survive. But yeah, so then what it comes down to at the end of the day is the balance, I guess, right? That's kind of the through line of our discussion where There is innovation, both in larger corporations and in startups, and that innovation can be done cheaply and fastly and even in a good quality, but probably not all three at the same time, right? That is still kind of the driving factor behind it. AI cannot solve that. And people in the middle, they are still there, both in startups and in larger corporations, but those people just need to do things differently. More open to AI, considering it as an existing force. And regulations, of course, also affecting us from all times and all sides. Quick question, by the way, Jonas, why did you decide to build a tool internally and not go outside and buy something? Because I'm guessing there is definitely a startup somewhere. There's a lot of tools startups, but look at this at a point where, at this point of view, But as I mentioned before, those stages of companies were like at the stage 2 and we had like a team that wanted to do it, proof of concept. And based on that, they did it based on the security reasons, because if you hire some kind of third party, you're going to feed them all your documents the same way and you're going to need to fix them to do it. But you'll have an intermediary. That you have to pay a fixed cost. Maybe it somehow balances out, but it puts a third party inside your system. And since it was like a pilot, a first-time thing, the management wanted to try it and it worked great. So basically then it becomes more expensive, I would say, than to buy a startup, but at the same time more reliable. We did a calculation of business case and and it's making a profit. Okay. Yeah. I'm trying to get us to downsize, but I cannot. No, we do IRAs, paybacks, everything. And it's profitable. Oh my God. That's why it's very hard as a startup to compete with existing businesses regardless of what industry they're in. Just too many assets. Yeah, we're not selling this product outside. Yes. Yet. But yeah, let's try and wrap our discussion by talking a little bit about the future. And we've talked a little bit about the future. So Vytas, what do you think from the startup perspective? There are many more startups that are being founded and both funded and founded. And the people that are inside of those startups are also changing and more people are able to do startups. Scale. So what's the next big thing? Will everybody be a product creator like yourself, or will everybody have their own micro startup that solves their own problem, or will just startups cease to exist because everything will be done by AI and there's just no innovation left? That's an interesting idea. Could be. Still, someone will still need to do the AI stuff. But so micro startups, that's already been there. I would say for 10 years almost. And now it's nano startups, maybe. I don't know what that means in reality, but maybe there are already startups where there is zero humans. We need to check, right? But that, I wouldn't be surprised by that. Yeah, me neither. Like you ask an agent, hey, you know, just figure the shit out. I don't want to run a startup. You run the startup. And by the way, there is one tool that does, but still, you're a human being who uses it. You need where there is no human being. But aside from that, I think it's changing the landscape. What can you venture in? I think, like, for me especially, it's also a very challenging yet interesting time. Like, how do people, how will people use products? So me, I see that, for example, so we're project management system. Yeah. What is the main issue there? Like Jonas said, there is monkey job filling in the data, getting the data, but without good data, you cannot make good decisions. Simple as that. And it doesn't matter whether it's a human or AI who makes the decision. It will be, you know, crap in, crap out, like with audio. So So that's where a lot can come in. And people like, if I'm a designer, I don't want to fill in boring docs about what I need to do, how I did this and that, and it's specified in this documentation. No, I want to go do design. So let somebody do the data part for me. And systems like project management systems like Teamhood, they need to go to the place where they are like databases. They are accessible, consumable, AI-first databases. And then there is no UI. The user experience now will be a chat window. That's it. And you collect updates and statuses not by asking humans to go and click button. Oh, it's done. Oh, I spent 2 hours here. AI does for that. Inspects like what the human did on the machine, on the computer, what has been written in Teams or Slack. What has been sent to the email, what transcript from the meeting, what are the keywords mentioned, projects, everything is like, you know, ingested. And then human doesn't need to do that boring job. Yeah. I think that's a very, very near future. And what are what's left then? Hopefully, you know, the fun things. Yeah, we go ride a bike, run, walk a dog, you know, whatever, play with kids. Subsidized by the government. Government giving like money for you, no work, just like Elon Musk said. But then you need to go and spin the wheel to make the electricity because who makes that? You get a monthly check just for— renewables. There's so many solar panels in Lithuania for a country with basically no sun. There are so many solar panels. I'm always fascinated by that. Oh yeah, and the subsidies has been lowered. Yeah. We'll see about that. Not as profitable anymore. So, Jonas, what do you think? What's going to happen in 5 years in bigger corporations and in public-facing companies? That's a huge ask, because like talking from a huge company point of view, like LTG, talking to all the shareholders who are like Lithuania Republic, My opinion is that we will do incremental innovation because of asset-heavy things. We will linearly improve some things. Some things will be improved faster, some things will be slower because of CapEx-heavy industrial projects that need time to build up. The cement needs to dry because you cannot build something like this. On the other parts, the progress will be faster. And in the final conjunction, by reaching the peak efficiency or value proposition for the customer, since we have a huge obligation, we are the, like, only the railway tracks are there, rolling stock is there, the customers are here. They do not have an option to choose for now from someone else. And we have to bring them the biggest value by seeking that innovation. AI is the only path forward. And we have to do it not only by AI, we have to do it by lean methodology, adopting that, adopting other efficiency strategies that we can get from the market. That's how we are looking at things right now. It's a future ahead. AI won't gonna challenge you by doing railways, right? AI will be our good partner. Maybe you can analyze some kind of datasets, give us suggestions, but not only AI. There's a lot of space where you can do another like innovations. Yeah, yeah, makes sense. Makes sense. So yeah, thank you. Thank you, guys. I think it creates a picture that is both interesting, you know, innovative and grounded. At the same time. So we kind of, I hope, managed to cut through the hype a little bit and see that at the end of the day, people are still there. People are still needed to both startups and existing larger corporations, and new work is being created due to just innovation happening, not just by AI or through AI, but, you know, a lot of things then thanks to it as well. And we will just see where it takes us. The hope is there. The hope is there. A new hope for all of us. It's going to be great. Thank you. Thank you both. So where do people find you? Where do people read more of your thoughts? Vidas, where do people see you? I do a few videos here and there on YouTube. You can find that. I'm more active on LinkedIn. I tend to give examples how we work, what's happening at the time. So if somebody wants to check it out, just Widas Wasylauskas on LinkedIn. I think there is few of them, but they will recognize. We'll put a link in the description. Hopefully we'll find the correct one. Dionysios, what about you? Just sure, reach out on LinkedIn and we'll connect. Sounds good. Sounds good. All comes down to LinkedIn. Soon there was going to be a LinkedIn just for AIs. We already had a Reddit for AIs, even though that was a fake, I think. I saw a LinkedIn plugin that scans your profile and gives you suggestions how you should communicate with the person. That's useful for sales stuff. Does not help, I think, but still, because people don't like talking to people they don't know. Anyway, thank you both and thank you all as well for joining and listening to our episode here today. Did you like it? Leave a comment. Did you not like it? Leave a comment because the algorithm thrives on comments, and the more inflammatory they are, the better it is for the algorithm as well. So if you are very angry and you didn't like a single thing that we have said today, do let us know. We value your feedback and we will share it. And subscribe to our channel and see the next video when it is out. Thank you very much and see ya.