Ep 165: Kaspar Korjus

On building an AI-powered product, outsourcing negotiations to AI, selling to Fortune 500 companies & do brothers make good co-founders

Kaspar Korjus is the Co-Founder and Chief Product Officer of Pactum AI - a tech startup that is revolutionizing the world of business negotiations with chat and AI. Walmart, Maersk and other Fortune 500 companies are among its clients! Prior to Pactum, Kaspar was the founding Managing Director of Estonia’s e-Residency.

On this episode we talk about:

• How AI can help in business negotiations

• The effect of openAI on the AI industry

• What unique challenges are builders in AI facing

• How to tackle adoption and trust issues for AI tools

• Selling to Fortune 500 companies

• Personal and organisational use of AI
  

We are on YouTube and Linkedin as well

 Watch select full-length episodes on our YouTube channel > https://www.youtube.com/channel/UCP6ueaLnjS-CQfrMCm2EoTA 

Connect with us on Linkedin > https://www.linkedin.com/company/pursuit-of-scrappiness/

---

Read the full episode transcript below

All right. Hello, hello, hello, dear listeners. Welcome to another episode of the Pursuit of Scrappiness podcast. Whether you're building a business, running a team, or just starting out in your career, we are here to bring you scrappy and actionable insights to help you become more productive. My name is Uldis Teraudkalns, and today I am without my co -host Janis Zeps, sadly. Just me, you get today.

We are here in sunny Tallinn and when I wrote sunny Tallinn I didn't realize that there's going to be actual direct sunlight shining on us. It's a beautiful day out here. We are live at the Latitude 59 conference in the sub -podcasting area with a live audience. Audience still showing up. Glad to see they don't hear me, but I hear them and I see them. So before we start, quick reminder, follow us on Spotify and Apple podcasts.

It helps us more than you know to bring more and more great founders. And in exchange to that, you will find more than 150 episodes. They're covering all topics you need to become a scrappier and better version of yourself in life and in business. So open Spotify, Apple Podcasts, click the follow button and be the first to know when we come out every Tuesday. So about today's topic. AI has taken world by a storm. Yet only two years ago, if you ask the average person,

what can AI do? They would say, I guess there's some kind of voice assistant on my phone that answers some questions in a very organized way. There's somebody building driverless cars and maybe when I play video games, there's some kind of AI fighting us. But that has changed. And however, the AI industry goes back much further and there have been exciting products being built prior to OpenAI and the AI revolution as we know it today.

So today I wanted to talk to one of such companies that is building practical business applications for AI. And please welcome Kaspar Korjus from Pactum AI. Hi Kaspar. Hi everyone. So for those of you who don't know Kaspar is the co -founder and chief product officer of Pactum AI. A startup that is revolutionizing the world of business negotiations with chat and AI. Walmart, Mask and other Fortune 500 companies call themselves as its customers.

and prior to Pactum, Kasper was also the founding managing director of Estonia's wildly successful e -residency program. By the way, I'm an e -resident. Welcome. So today we want to talk about cracking negotiations, applying AI, and also a bit about selling to Fortune 500 companies, which is something that not many companies in the region have cracked so far. So let's jump in. You started an AI -enabled product.

way before the OpenAI story came out. How did you get this idea at all to apply AI for negotiations, such a unique use case? Yes, it's actually quite an interesting story. As you mentioned, before the OpenAI revolution, there were self -delivery robots. And actually, one of the co -founders of Baxium, my brother Christian, was the AI lead at

Starship Technologies, which is doing the self -delivery robots. And another co -founder, Martin, sold his previous company and was doing negotiations at large scale and experienced the pain of negotiations. So he brought the topic of negotiations and when we looked into it, then we realized that, hey, this is the most important aspect of commercial companies' business activities because every deal that you do with your clients, with your suppliers, with your employees,

There should be some negotiations part of that. If you don't have time, you just sign a contract, but otherwise you have negotiations. And it turned out through science that people are very bad in negotiations. They really suck at it. We have these biases everywhere and we can't get rid of them. And then when we put these things together, that there is world most important value creation aspect of negotiations where people suck at. And then there is AI that can take over.

many industries, even like before OpenAI era, it was clear to us that AI will revolutionize all the industries. And you just merge and you get AI negotiations. Okay, that's the starter. And then you start figuring out how to do negotiations via AI. And for us, the chat was a way to do it. Okay. But, okay. We mentioned, of course, that the OpenAI revolution has...

created almost an industry, probably industry experts and more experienced people like you will not agree that it has created that industry, but how has it maybe changed the industry for you and how has it affected your product roadmap if at all? It definitely has created an industry because like what the investors...

Investors tell us that every second company comes with another LLM application, you know, and so it's quite annoying for them to hear exactly the same startup. So we were lucky to start that few years before it and we used large language models where they are most useful, but the cool thing for our business is that when it comes to negotiations,

That means that you can't reveal your secrets to other party, but if you would reveal something to large language models, it's like a child, right? You can get that data out of that child somehow. You see how they're lying and how to manipulate them. So you can't trust large language models with the data that you can't want to reveal to other side, which means that we can't put clients' most important data through LLMs. We need to do that in a structured negotiation manners.

and use large language models in different ways. So for us, it has been very helpful because it has put all the founders globally to the wrong path of using large language models to compete with us. On one hand, on the second hand, we are selling, like you said, to the largest companies in the world. Our first customer was the largest company in the world, Walmart. And we need to speak to their C -level, literally president CEOs, chief procurement officers. And...

And it has helped us because they themselves have used ChatGPT personally, their children have used it personally. So they have become friends with these bots all of a sudden. And now introducing back to AIBot to do deals with their suppliers helps us because we don't need to convince that these AI bots can be useful. Okay, so now we know who masterminded this whole thing.

divert all your competition and educate all your customers. That sounds like a pretty good master plan right there. And when building in AI, you're actually, you know, a lot of these companies are plugging into OpenAI and probably not doing much of AI technology development themselves. But when building this AI enabled product, what unique challenges?

have you encountered that maybe you haven't seen before in your business experience? It's a good question. Most of the challenges, I would say, is normal commercial business challenges, how to grow. But the uniqueness, one unique challenge is that it's crazy how well the product is working, meaning that you have an old contract, let's say you'd pay 10 millions for that, you put that through Back to AI.

and now you pay nine million for that contract. So you save the million. And you didn't have to do anything yourself because POT reached out to the client, renegotiated the deal, came back, updated the data in ERP, everything is done. So if this works, you would think why it's not scaling, like why they don't put like 50 billion through Bactium to save two billion every year. And that's struggling because...

Because it works in different use cases, but the change management in big corporates, the client success upsell, the thing what the Lloyds and KPMGs are doing to kind of transform organizations to new place and invite bots to be part of their colleagues. This is a process that is very difficult for startup, right? Because we don't want to be a services company, right? So we need to partner with others.

And so the process is very long not to do pilots because maybe people in Europe don't know but American corporates are very pro innovation like they're inviting you to do pilots. And if you fail, they will throw you out. But basically as a founder, it's super easier to be in the US to kind of go into these large companies no matter what's the size, even like the Walmart. Like we called them, we said that what we have, we had a meeting basically and we got the contract to try it out.

And it's unbelievable in Europe, like you need to sell three years before someone recognizes you as not as a salesperson, right? So there you can get into the door, you can try it out, but the usage increase and change management process in large enterprises is our this year's, let's say new challenge that we are dealing with. Okay, but that actually brings us to a very good question about adoption and about trust, about acceptance because...

Obviously it's very easy to play around ChudGPT for your, I don't know, annoyingly long email to your colleague or some kind of birthday greeting or I don't know what, but to put your 10 million contract entrusted to AI, I think it's a very big leap, right? But also, from the other side, how does it feel for the other side that is in the negotiation? Do they understand that they're talking to an AI and are they okay with it? Is it not disrespectful that you're...

for example, biggest supplier comes to you and actually, you know, it's not their VP of sales talking to you, but it's back to AI. It was pretty crazy how many very risky bets we did when we started five years ago. One of them was that we can create a mathematical formula of a value function of our client so that AI would know what trade -offs to do, what is better contract than the other. Then...

But for the supplier, as you said, the other party, to speak to AI about actual business. If you fail there, you can't feed your family. It's that important business for them. And to trust the bot in the chat version to do that, and then the medium to do that. And what was the biggest win of Bactom? Maybe it happened two months before establishing Bactom was when we did the first year as a...

Autonomous Negotiation was a e -Resident company by the way, a friend of us. And there was a limousine company and we sent the negotiation link to a 75 year old Italian guy who said that he hates computers but the negotiation was about weekend rates of the limousine service or whatever. So we saved some money for the limousine driver and we saved, I don't know, 60 euros for our client. We didn't get paid for that. But we saved 60 euros. And then...

The client, the E -Resident company called to the limousine owner, 75 year old guy, he said that he hates computers, he even doesn't use smartphone, but he loved the engagement with bots. And that was a striking, like what? Like you hate computers and you made a deal that you made a better deal and you liked the engagement. And then it turned out that, hey, we are not against bots, we are not against AI as itself, we don't expect spirit.

We don't expect that emotions if you get the better deal. And there are many things that AI has proven us wrong, like that the AI can't be creative, which it is at the moment, that AI can't do negotiation deals because you need human, you need to play golf with each other, you need to touch each other, you need to build trust. And this is all wrong. Eventually it's about the value that you create.

And it turns out negotiations is quite stressful environment where to be like imagine going to your boss and asking salary increase. Like do you want to do that? It's stressful environment to ask more salary because if it doesn't work out then kind of it's weird tensions in there, right? So negotiations is stressful environment and when you can do that by a bot you feel more relaxed and you feel more open because there is not like 50 years of experience of guy who tries to get very good deal for himself.

It's an AI that is neutral and is really trying to listen you out, really has time for you to come back the next day to ask your questions to get you a better deal. And so, maybe I forgot your question, but it turns out that it really worked already five years ago when we tried it out and later the LLMs just made us a bit simpler way how do we can actually sell it. Is it possible and have you ever...

detected that your AI is talking to another AI or packed them on both sides of the negotiation? One day definitely. At the moment you need human on the other side, but we have had use cases with logistics, for example, where the other side is checking their Excel tables on what rates they need to answer, what trucks they are available, and that's also nonsense, right? And then the question is, let's do it API to API negotiations, bot to bot.

But it takes some more time. It's easier to automate one side first and then automate the other side and automate both sides at the same time. Okay, and this is probably for anyone building in AI that can replace some part of a company's team or staff. What are the concerns and maybe the ways how to get around them when selling this product and saying basically, well, you don't...

no longer need this procurement specialist, there's going to be AI doing it, or is going to help you as a specialist, but obviously there's the fear of replacement immediately, and also maybe some kind of vanity questions like pride, like how is it better than I am with my 30 years of experience? And this goes beyond negotiations, if you're going to be replacing any kind of part of a business, you're going to face that. And this is definitely one mistake we have done, is that we are very...

egocentric, meaning that back to me AI created billions of savings or millions of savings and back to me AI dashboard shows how much it saves. While now we are saying that how buyers, how the users created savings through AI and they are, AI is the extension of these users, not that AI is like individual agent who competes with them. And I think this positioning needs to get right on very early days to avoid fights against AI bots, fight that they will take your job.

and embrace the AI bots that they will enhance yourself, that all of a sudden you have an army of thousands of bots, or thousands of people if you may will, if bots are like people, thousands of people working for you, waiting your directions to lead them and you can release them and they work 24 -7 day out to produce results for yourself. And to understand that these army of bots are for you, not...

to compete and get your job but for you to lead them I think this is the key difference to make. Yeah and for example increasing your budget by 10 % just through better negotiation deals and also I think it goes along together with this also trend of

of number of team maybe no longer being metric that's that's because some years ago it was like okay how big is your company we're like 200 people and and that was an achievement in itself something to brag about and I'm sensing more and more that now it is completely the opposite right yeah when people ask we have 20 people or how much you grow we grow from 60 to 200 people exactly good luck with that

And now, yeah, but the same with us, like we have had the same headcount, we have 90 people, same headcount for now over two years. And with AI, it's possible to kind of keep it more or less the same while our revenue has grown every year three plus times. And AI also not helps only the product, but productivity, like AI helps our engineers to become much efficient, more efficient.

While first it was thought that only junior engineers can get AI help, but now very senior as well. And also the processes. Like you may have a question about HR or job profile that you want to do. Traditionally you will reach out to HR to help you do that, but we are not used to that. We are not used to actually write to GPT and within five minutes you have a job profile ready. Like in so many tasks internally, every employee can do with these AI agents that we are still not leveraging actually.

How much are you actually applying internally in AI? And do you have like a specific person responsible for bringing as much as productivity as possible using AI? I think I worked before in government and then there was this mistake what many other governments did is that they had one person responsible for all the solutions. Like technology is here for every person, every industry, every department. So you can't give it to that one person like everyone needs to.

embrace them the same way and lead these changes. So in that sense now that every engineer fights for their position and for their salary and every person needs to find ways how AI bots is exactly useful for them, for their own sake. So I guess it starts with the founders, right? It takes time, obviously, but I'm surprised how few people still use. Like I'm using ChatGPD like every day, like 10 plus times for very different roles.

Super helpful. Like you just click an app and then you have the smartest person next to you to speak to. Like why you shouldn't speak to the smartest person, why you should speak to us who are not as smart. Like obviously for other reasons, to socialize and meet and stuff like that. But you know, we have a smarter person in our pocket, let's use that. Have you tried the voice? Yeah. I'm doing that with my children and it knows Estonian also, so basically we're speaking to the...

to the AI in Estonian, which is quite funny. And you can tell the AI to draw pictures for you, so children are asking to draw Spider -Mans and cartoons, and then do cartoon in Tartu where we live at the moment, and Spider -Man in Tartu with me, and then it draws this picture and has conversation like super fun. That does sound fun. Well...

For the second part, you kind of answered already. You just called them, try and it works. Selling to Fortune 500 companies. But how did you actually define, because on the top of your website it says that Pactom AI is the negotiation AI solution for five billion dollar plus worth companies. How did you arrive at this quite precise definition and super ambitious definition that you as a startup are going to be hitting?

these kind of target customers? Yeah, it did help that the worst was Walmart and then we managed to come down and the question is what's the bottom for us? And there are just business reasons because if it's five billion revenue company it means that the spend is approximately three billion unmanaged spend approximately whatever 500 million out of 500 million 200 million is applicable through Bactium. Out of 200 million we can create

like 20 million savings, out of 20 million savings we take 10 % cut, we get 2 million. And this is the maximum per account then with 5 billion revenue, like 2 million. But in reality you can't get to maximum, you get 1 million. And so you get the limit quite soon, kind of what is today applicable to autonomous negotiation and hence the limit of 5 billion. One day most likely it will be the AI negotiations, back to May I will be on your pocket for a daily negotiation with your children and with other people, right?

But it's... Negotiate the allowance, yeah? Yeah, yeah. And divorce negotiations is definitely a very good use case because there are lots of emotions there and emotions are stopping rational good decisions for both parties. So definitely, if you're going to get divorced, Google back to AI before, see whether it's helpful. But we are not there yet. We need to focus in the place where we are, in procurement, supply chain, large enterprises. And there, indeed, you need to be large in order to give enough spend to back them.

This is going to be an interesting case study as often times software providers start with some kind of either premium or premium or mid -market and then always, you know, should we try to aim for enterprise or should we not? And you have started from the other way. Yeah, and it's quite funny how few people realize even from investment side that

They are asking us, when do you go to mass market? And we're like, 500 companies in US contribute 66 % of the US economy. Like majority of economies only by 500 companies. This is the mass market. Like why should I care about rest 1 .5 million companies if they give 10 % of the total spend they can put through? So the mass market definition is so wrongly for many, for us at least, because it's not number of companies, it's dollars managed basically.

And that's why we don't need to go to the medium market unless we want for branding reasons and later in future revenue sake. But yeah, it is just the problem is it's difficult to land with them because when you go to Walmart and we had only office assistant and three co -founders and then Walmart asks you how many negotiations you have done before. We said under 10 ,000. And then you have six months before the actual contract is signed.

to build a product. Because the sales process is slow, right? So the challenge is that you're less mature, obviously, in the product when you're starting from large accounts compared to normal progress. And hence you kind of need to be more creative there. But I do suggest that working with large enterprises is not more difficult than SMB market where you're competing with everyone else because it's so easy.

to work with SMBs because you know some friends who are there and you give it to them and you feel your pain yourself as an individual and then you start building SMB companies in Estonia at least and then you compete with everyone else and I wouldn't say that life is easier then. Yeah but okay you mentioned a couple things so one is the long sales cycle, second is this perception of maturity.

And how do you go around that with three co -founders and an office administrator? Like, does it depend on what you're selling? Because I have encountered myself in my sales when companies much larger than mine have kind of looked at...

at us and said that you know guys you need to a bit mature for you to be able to work with us. For us it worked only if we are raising capital early on like with the slide deck I see there's ten time give our angel investor we had Tavet Indrikus, we had Jaan Tallinn and we raised from the first million and what you do then is instead of hiring team you outsource every unit.

The outsource engineering, legal operations, sales agents, basically you start managing like you would have 100 people in your company, but through contracting, through partners, PR, HR partners, right? And then you scale yourself from the first week, because we knew already who are the best partners globally for each unit. Within the first week, basically, although we had only one office manager as an employee, we were able to create massive impact.

through that and of course you burn money quickly so it's a high -risk kind of business through that hence we have had to raise money every year and these days it's a bit more difficult to raise so I don't know how sustainable it is but for us it worked because if something works then usually you lose through the competition because American startups get more funded grow faster are there

are there in media and everything. So in order for us to succeed, we have to double down very quickly and raise capital often and much. And of course, dilute yourself as founder through each race, which is the side benefit of that.

So, with Walmart, you called, they liked it, great. That's, I would say, quite lucky. You're lucky later when you reflect, but if you call and meet 50 companies, and then individually you're lucky, but 50 companies and you close one, then it's not luck, it's hard work.

Obviously. But you have added quite a few serious names to that list. So what kind of strategies have helped you with the next? Is it we are working with Walmart and that helps? For us? Others? My previous background was e -residency. So we did lots of crazy stuff in government in order to get media attention. And when the next media bus arrived, like blockchain or Brexit,

or crypto or whatever, then we ride with that. And this helps us to be in a picture, be relevant when the next wave came, right? And also my other co -founder sold this company through PR. So PR was our bet from early day. And it's very dangerous bet, of course, because you don't know whether you can get any media stories out of that. So out of first million we raised, we did a contract with 8 ,000 euro monthly fee.

for the PR agency for one and a half year runway that basically meant over 10 % of our budget went to PR. Basically just paying to media agency to find stories. And usually you would put that to engineers, you would put that to like other stuff, sales, direct outreach. And you do that and you do that and sometimes for 12 months you can't get any stories, right? And then 13 months, 14, you still keep putting money in, money in.

And one day it starts building up your brand and then enterprises find you. Basically out of Fortune 500, we have 400 companies who have put their emails to us on VP level at least, to our website. Like OpenAI was last week, for example. I can't say that, but basically these examples are there and then sales becomes different. Then they reach out to you and you don't need to sell theoretically.

then you explain what you do basically and you have options and you're in power position. So I do recommend to invest more in PR in the US media of course, maybe London but mostly US because UK is following anyway US. So US media, New York, go there, find an agency, our agency is for example what Trump, President Trump had, hence we have been to Fox News many times and pay.

and don't negotiate the cost down. If they say 10 ,000 monthly and you think that let's do 3 ,000 months, then you can't get it. Like have a budget and just be ready to pay for that. And the art of media relations is that you exchange the media partners after every 12 to 18 months because they use all your resource, they want to impress you and then they forget you because of other clients. So change the media partner after 12 to 18 months to new one, they will try to impress you, use their network again.

and you get your tier one stories and after tier one stories, results won't happen straight away. The brand builds and many people find these stories month after that story has been published and then they find you. Well, this is quite a flammable approach. The Eastern European in me says, do not try this at home. It's such a go bigger, go home approach and I love it. It's bold, it's great. And it has clearly worked for you.

But yeah, it's scary. I have been also encountering these kind of situations, these contracts, and I have went the other way and was like, my god, these guys in the UK or New York, they're just gonna rob me blind and I'm not gonna get anything for it. And you need to work yourself to be great at finding these stories, of course. It's not only paying, it's like...

Finding angles how you are relevant every day, responding to something that happens in a matter of hours to get your story and response out. So it needs collaboration time. But if you consider this not as a brand building, but actually sales machine, then the investment is very clear there. But whether you hire 250k account executive in the US, or you put 10k monthly and you hire a PR person internally and 10k monthly for that media pocket.

And then you compare the results, whether one account executive 250 can create more sales than PR altogether and you do the evaluation yourself. Did you figure out some good tracking of that PR ROI? As it's always a very painful question, it's like, okay, what is actually this PR bringing? It is not that direct, so yes, but you can get website traffic.

And you can always ask where do you find us? And where did you find us in this case? And quite often in our case has been that someone read something from somewhere and that's it. You already mentioned this application for kids. But what other main applications of AI do you use for your own personal productivity? For me, it's mostly just JetJupe .de app. That simple is that.

But what is the most frequent use case that you have it for? I can check out what I have asked today. But mostly, there quite often are work -related things.

What is a podcast and how I should behave on one? What I asked today morning was that I'm creating a new title, a new position and I don't know what title the position should be so chat is helping me to pick a title when I describe the role of who I want to hire next.

And then it's important to understand that it's conversation. And before that, I compared two titles and I asked the differences of those two titles. Then I wrote my own, then I put my own data there about what the person has done before or what to expect. So it's like to understand it's not like FAQ. It's conversation and true conversation, you become better. Mostly work related stuff, but in different roles.

legal HR, product, everything else. But are you somehow training those GPTs to some specific topics or just use the general application? You train yourself how to ask better questions, like in Google search engine game or before other search engines, you had to train how to ask questions. And now you need to train how to ask prompts, basically. Okay, okay. All right. What they did before that was not with ChatGPD, I created a song,

there is another app that you can create song. You just say that do a song about autonomous negotiations and you can describe what's the song about. And basically it finds data from web about your company and then it creates its song and we show that to one of our client during the first kickoff presentation. So there are many applications, most of them are just fun and cool. Speaking of fun and cool, you mentioned that you use also Chaggpt for entertaining your kids.

Can you tell us about what exactly you do in the sense of also maybe there are some concerns about security and what they might do with it or is it completely supervised or do you let them also go a bit crazy? My oldest kid is six so they don't have mobile phones yet so basically it's through my mobile phone. But obviously we need to take care of the internet hygiene and privacy hygiene.

Like you wash your hands, they need to understand what you can and shouldn't do with the internet and with chat. But chat itself is not that dangerous in my opinion than putting pictures to Instagram or Facebook. It's still a private conversation you have with the agent so I'm not too worried about that. And it's pretty reasonable, like when you put something on the internet, public search, then you can get...

quite bad results in some matters but these large language models prevent kind of very dangerous answers. I guess they have been trained and baked in by thousands of people. Obviously you can get something out if you're very smart but for a child it's quite difficult to get something ugly out from the answer so I actually trust GPT having the conversation with my kids.

I guess the tricky part is if they master it very well, then they might overuse it for school and outsource their brain. It's happened anyway, so we need to just train them how to collaborate. Calculators came and we started using calculators in math class. This is here to stay and we need to change the way how we teach and study things, so there is nothing to be afraid. We are just, again, more capable, more powerful as children already, especially in emerging nations where...

education is not that freely available. Okay, well we have this unique opportunity here that we have people in the audience unlike the other podcasts. So I wonder if there's anybody who has a question to Kaspar that we have a microphone that we can pass. So guys, you have something to ask to Kaspar? AI, fun, entertainment, kids? Yeah.

there's there's there's a person there

Should I ask this question in English or in Estonian? English please. Okay, sure. So how do you make sure that children who have this unrestricted access to any AI consultants or let's say for example your children using ChatGPT, do you think it might be essentially...

a way they get too overstimulated as we can see what's going on in America currently. And what do you mean by overstimulated? Overstimulation, for example, they can no longer have these higher attention spans. For example, the average attention span has gone from 30 seconds to I think eight in the US, mainly due to applications like TikTok, but also including, well,

For example, one main thing with Chudge GPT was the fact that children got too invested in it in school, made them get distracted and also impacted their learning capabilities. They started using these AI chatbots to essentially do their homework, essays, stuff like that. Yeah, okay, good question. I have my perspective on that and maybe a bit higher level. My perspective is that the world has experienced different technological disruptions.

Quite often, but it used to be like per 40 years time, which is now per like five years time or per two years time we have some new disruption. And I think what we have learned from the previous disruptions is that when new technology comes, we realize that it is as good as it is bad. From day one it comes. When social media came, then it took many, many years to understand that this is not actually connecting people. It was the Nokia, Nokia.

It was actually destroying relationships and making people miserable. And now I claim that we are more aware of the negative consequences also what new technologies bring. I don't claim that we are better in fighting them and preventing them, but I don't think it would come as a surprise to anyone when there are negative consequences of the latest AI trends.

So it is on a very high level about the sustainability of human as a race and it goes down to the topics of personal topics like children, right? So in that sense on high level, on meta level, I think we are in better position dealing with new technological disruptions than we have ever been in terms of awareness. But practically, I think...

It's just, I don't know, up to parents to deal with their children, not to expect anyone else to get over and support their lives. Chat is one thing and Chipie Tea, but all of our lives are becoming much more cozy. We don't need to really go out for fight because the average...

support system is much stronger for children than it used to be when I was a child already or my grandchildren or etc. So which means that we need to proactively create situations of stress and fight to learn and to come out from a comfort zone basically. And this is I think parents thing.

In simple sense, if you have much money, don't give all the money to children, otherwise they don't learn themselves to fight and get that money. But the same context without money, if all the tools and applications are very easily accessible and solvable, then as a parent we need to just see how we use and leverage them without becoming too comfortable with our lives.

And it's tricky, but I don't think that the latest revolution of technology is more difficult than the previous ones. I think it just takes 10 years to adapt, to learn, to do mistakes and then we get better. Thank you. Anyone else? Yes, I was hoping for you, Erik.

Thank you.

Good question. Do you know some family businesses of startups in Estonia before Bactum?

One comes to mind is something about mobility, I think.

So the latest most successful is Bolt, Marcus and Martin filling. Six billion valuation. So we have to be working for them. We have quite good case studies of brothers working together before. And obviously it's an interesting case for us. It just works out and we didn't even have to test it because we have always wanted to do startup together. And it's like,

In some sense you have this ability to speak to each other without speaking in words and you are much more productive, kind of you understand each other and as brother you're always ready to fight, right? So the fighting aspect is still there but in a constructive way. So yeah, for us it has worked out.

really well and trust is there and our kids still meet each other in private lives. That's good.

interesting Independence Day or whatever dinners you have for Estonian family traditions. All right, thank you very much Kaspar. Thank you to the listeners, to the listeners here and to the listeners who are gonna be listening to this a bit later. Yeah, we see you next week as usual and hopefully we're gonna do this more often.

This live conversation is much different than our usual online connection. So thanks to Jukas, but thanks to the listeners also here in the audience. Thank you very much. Thank you, everyone.

Please note that the transcript text is AI-generated. We apologize for any potential errors or inaccuracies. Thank you for your understanding.