Curating for the Curious: How Tomáš Baránek Picks Books That Matter

Nigel Rawlins: [00:00:00] Thomas, welcome to the Wisepreneurs Podcast. This is the second time you've been a guest on the Wisepreneurs podcast. The first one was episode 41 last year in April. Welcome back, so Thomas, can you tell us something about yourself and where you're from?

Tomáš Baránek: Yeah. Thank you Nigel, for having me, the second time. I'm happy to be here. Yeah, I am, um, entrepreneur. Uh, I am a book publisher. I am father of three kids and husband. And, yeah, I am a life hacker, writing about ways how to improve your life and use technology during this process.

And that's. Mostly it.

Nigel Rawlins: Well, that's an awful lot because you are a rather young person, aren't you?

Tomáš Baránek: Oh, I'm not, I feel like but I'm, fifty-three in three [00:01:00] months. So, yeah. I, I hope I can tell it's still young

Nigel Rawlins: enough.

I think that's young. 53 is a good age, I think. Okay, so let's talk about publishing. Now, we did talk about it in episode 41, but I want to talk to you again about it because it seems quite fascinating. You haven't got a big population in the Czech Republic. So tell us a little bit about publishing again in, in your area, and then we'll talk about what you've actually published, which I think is very fascinating.

Tomáš Baránek: Yeah, thanks. It's a great question, because population of the Czech Republic is actually, uh, 10 million people. Maybe a bit more, plus a lot of Ukrainians now because we help them to settle, uh, temporarily. And, uh, yeah. So our population is not so huge, but we have [00:02:00] huge book market because the book culture here is very, very deeply and rooted in in our history. And, we have, I don't know, 16,000 of new books per year. Or several thousands of publishers. Including the one book ones. So it means there is a huge competition, and a lot of, um, cooperation.

Uh, by the way, our publishing company is, is like a middle, uh, sized, we have 25 people as a inside staff. But, it's not the biggest, but we publish books from nonfiction area that are published in circulation around in hundreds of thousands. Yeah, for example, End of Procrastination, from Petr Ludwig.

Uh, it had nearly 2000 copies sold from the [00:03:00] publishing date. So, uh, we can deliver books to our readers in a big circulation because we can explain what's what's important. And that is the, the main point. We publish just, uh, I don't know, from, it depends, from 12 to

20 books per year. But we really try and we try hard to choose or to create books that are really important that, that prepare people for the future. So, uh, for example, translations of books that we, buy licence for are from the the most interesting thinkers or writers from the whole world. So for example, Harare or Kahneman or, um, and many others. So, if you choose, if you choose well, uh, and if you don't [00:04:00] push too much. Yeah, in a sense of, of many titles that you publish per year , you can deliver a great experience, great knowledge to a big number of readers that can pay attention to you. When you publish hundreds of books per year, which is, it's common.

Yeah. You know, big publishers do this. You can't rely on their affinity to you. They, they will not wait for the next book that you publish. They, they just see you are flooding the market, uh, and waiting what will grow bit more than other. So we don't do that. So we really do our research.

We check for authors and experts, as soon as possible, uh, proactively, uh, I don't like the word, but we search for authors that could even write a [00:05:00] book because they, for example, publish great things on social media. Yeah. So, we ask them, we persuade them to, to publish books, or book and sometimes it works.

But maybe I talk more about our company than about the, the Czech market.

Nigel Rawlins: That's okay. I think it's fabulous. 'cause you know, we, we don't always think about the books that we see, but from the sounds of, if you are only publishing 15 a year or up to 15 or 20 maybe, and you've got a reputation for the type of books that you are choosing. Do you find, um, we, we will talk about some of those books in a moment, are, do you find that you can almost have people subscribing to every book that you send out because you're fitting their particular niche?

Tomáš Baránek: Yeah. That, that's something we even consider creating some subscription. Yeah. Entire. But it wouldn't work probably because some people, some readers want it. They, [00:06:00] they tell us, okay, Melville, we read all your books. I buy all your books because, you are like cur curating uh book choice for me. So create some subscription and I will just pay for it once a year and then deliver. But the, the problem is that books are not magazines. We can't predict easily what we would publish in the future, what price it would have, and so on.

So, it could end not very well predicting our revenue and the budget of this project. So not sure if it could work, but yes uh, people can rely on us as in this society where everyone is pushing information to you in every mean, in all means, uh, uh, it's hard to find re reliable sources. And I think [00:07:00] publishers should go back to what their job is about. Actually, um, and, uh, start being reliable again as they were when choosing titles and when being the point where you can ask for a topic or problem you, you want to solve or you are coping with and you got the answer. It's like any other service. Your doctor should work like that. Or your therapist. You just ask him her and you got answers that are good for you that don't damage you.

But the market says something else. Many creators, not just book creators, but content creators just flood the market. So we want to be like the old fashioned curator of information in information age [00:08:00] in AI era for readers who want deeper information, more concise, and still very informative in this world.

Nigel Rawlins: So I heard a couple of things there. Curation was one word. You know, you are cur curating almost an education for your people, uh, for your readers. Um, and you're looking for to give them depth of knowledge. But what you're doing is you are licensing the book and then you are translating it into the Czech language.

Don't the Czech people read English or do they prefer to read Czech?

Tomáš Baránek: Yeah. They, they do, they do, but still, there is a slight, I don't know how to call it, but there is slight barrier that you have, nearly always when you don't read in your native language that you talk mostly through the day. So book as a medium itself is, is, [00:09:00] uh, not easy to consume.

Even I I if you read it in your language. So, the slight barrier still makes a huge difference. When you are tired and you want to just sit into your couch and relax with a book, it's not so relaxing when you read it in, in your second language. So that's, that's the point.

People are really understanding English on a basic level. Really great here, very well. Czech language is Slavic language. It's not similar to English, so it's not easy to learn. It would be easier to learn Russian, but sorry, we will not again.

But, uh, English is okay for our readers. They usually say, okay, my favorite author, which I read newsletters from, who I follow on social [00:10:00] media is publishing her or his book in English now. But I will wait, for your translation. And by the way, it can take half a year or even Yeah, several months, or more than half a year. But still they wait for it because, and that's the reason why we have to do our work well. We are nearing the topic of AI and automated translation now because if you want to enjoy reading it in Czech, the translation has to be really perfect. It has to be even better than if you read it in English. It's not about changing the content. Absolutely not. But it's about fluidity of reading, or it's, no, not the word, but it's about how how flawless or how well it is compatible with reading habits and so on. So we have [00:11:00] to push really hard to make it easily readable in Czech and still keeping all the emotion of original text. And that is the point. We haven't found a way to use AI translation, uh, at least not directly or fully at this point, because, I don't know if it is only in Czech, but Czech is very rich language. It has a lot of synonyms for every word. You use a bit different word, it has a bit different meaning too. Even you could consider it as it is the same, but it is not. So, you always have to choose what word is better in this context, in this book, in the context of other books that we publish, in context of the terminology of the field.

So, there are many variables that go into this process, and yes, it's about context. As you know, we know [00:12:00] that AI's need context to work properly in some useful way. So, you still can't, can't pour into whole cultural context.

Yeah. It's impossible. So we are not yet there and we are happy about it actually. Uh, we like how people translate what they put into translation. So, that's it. Readers in Czech read in  Czech , and it's still not changing too much. But we publish Czech audio books too, which are narrated by some Czech, um, actors and it sounds really real. They are real people. There are some small mistakes in their speech and so on, and it's very similar to reading human made translation. And we still think that there is [00:13:00] something that attracts readers to paper books, to to human voices, that actually shows that the paper book market and the audio market is growing. And, the paper book market is not diminishing, not just in the Czech Republic, but in the world. Audio books are growing a bit, but mostly, they eat ebook market a bit.

Yeah. Not the paper market. So I think that's interesting. It means that there is something in, in the book itself, as in the medium. And then in the format, in the paper and the way how you consume it. Because when you take paper book, you usually sit somewhere else than you sit when you have your phone or tablet opened.

It's not computer. It's different mindset and that is what people, in my [00:14:00] opinion, what people. strive for. They need this in their lives, so that's hope for our business too. Yeah,

Nigel Rawlins: I'm a big reader of Kindle,

Tomáš Baránek: I know.

Nigel Rawlins: and the reason is that I, I do take notes and feed them in through, Readwise, into my note taking app. And I've been with Roam for five years and I've just shifted to Logseq.

Tomáš Baránek: Mm-hmm.

Nigel Rawlins: is very simple. I'm, I'm not too sure how I'm going to go with it.

It, it is very simple, but I like to grab those notes and then use them. What you've just said is quite fascinating. 'cause publishing is a business. I mean, a book that's purchased by somebody is bought for a reason. And you mentioned earlier that it solves a problem.

It might be the problem they wanna think more about a particular topic or, or get some depth of knowledge. Now, I saw on LinkedIn, now you are always writing in Czech on LinkedIn, but there's a, a button there that says [00:15:00] translate. And one of the books you mentioned was the Wager.

This was a while back, and I read that book, I got it on my Kindle to read. I thought it was the most amazing book that I'd ever read. Maybe you'd like to say something about why you chose that one for your Czech readers. Tell us something

Tomáš Baránek: Yeah.

Nigel Rawlins: particular book and then I want to talk to you about, Yuval's book..

Book that, and then we'll talk about AI and stuff like that.

Tomáš Baránek: Sure. It's book from David Grann, purely a historical book, about, big and, yeah, somehow defining voyage.

Nigel Rawlins: 1741. So we're talking about old sailing ships.

Tomáš Baránek: Yeah. It's, very old sailing ships. For me the book is maybe more about other aspects. Even the book is really terrifying or, or maybe it's story that [00:16:00] you can't just put off and you you have to, you have to read it until the end without stopping because the conditions in which sailors lived and what things they had to do to survive. And it's, it's really very strong book. But for me it's very interesting that even in these, centuries, during history that we consider, I don't know, maybe I'm, I'm, I'm a bit stupid, but when I look at the history, I sometimes see or think that people were not so clever like we are.

I'm sorry for that, but I know that's not true. But somehow, some bias that we have looking at, I don't know, maybe black and white pictures or drawings, I don't know, the world was colorful. And was maybe even more mindful in some way [00:17:00] than than ours is. So when you read this book, and I haven't read it, I, I listened to it in the audio version, narrated by a great actor, and it sound sounded really like I can't stop.

So, I saw that we haven't made some big progress yet, as a society in many ways. Uh, it seems like, for example, the media manipulation, and the way how sailors that returned from these voy voyages after really excruciating, terrible, experiences.

How they tried to improve the way, how were perceived their trip and what they did there. Then not just society, but media and governments tried to use this information. And, it's so similar to what we do [00:18:00] now. The kind of propaganda or media manipulation, and they had no internet or social media. They just used books. So, so sailors returned from the trip and to improve their image, they just wrote a book, like diary or, or journal. Yeah, they, they rewrote or republished their journal, but sure, it was a bit biased or, or even, uh, deliberatively, uh, improved, to serve purpose of, of improving their image, even save their lives, in some process later. So, uh, it's something that we see now and we saw it probably then. So David Grann made a great work showing us that this is something that was always with us. Then we have to be critically thinking every day

Nigel Rawlins: I was very fascinated by it. The [00:19:00] hardship that those sailors went through, but also the intelligence of those sailors. That they actually could write. We, we assumed that because they didn't go to university or they weren't educated. None of them would've probably gone to school and they would've probably gone to see at a young age.

And yet they were still able to write detailed things. But the, the reason we know that they had to write this stuff cause it was potentially a mutiny. And in those days, because they were part of the British Navy, they could be hung, hanged by the neck. So it was pretty horrific.

But what I don't get is why would the Czech people be interested in this?

Tomáš Baránek: Okay, sorry I forgot to answer this question. Most of our books are like non-fiction books. We started with productivity books and economics and, uh, psychology and how to make your life better. But then we found that you can't learn, and you can't improve your future if [00:20:00] you don't understand the past. It's nothing new. But from the perspective of publisher, we saw the gap here, and because Vit Sebor, likes history and he, even was a editor in chief of some magazine on history.

We hired an historian, Marek Vlha who is a really great editor too. And he's now editor of in Chief of all books that we publish. But his view on history is very compatible with the view of how we want to help our readers shape their future. So, we started the set of books on historical topics. The main rule for choosing books is interesting and well written and sourced to [00:21:00] entertain our readers, so we want to bring history and historical topics to our readers that could be a bit like I was, uh, you know, uh, blindsided or see the world just, uh, through lens of, yeah, I don't know, productivity, effectiveness and these all buzzwords that you can listen everywhere.

And it's enlightening in some way because when you look into the past you really find other prisms, how you can consider what you are doing, where you are going in, in your life. So it works. So we chose not just Wager and Wager is a very special book because it's about sea and or ocean, and we don't have ocean.

We have just lakes here. We are in the center of Europe. So it's [00:22:00] something distant enough and something we want. There is many jokes and memes and feelings about we don't have sea. I understand that, uh, yeah, it was your intention to bring me here so we, uh, somehow can materialize, ocean and see through our imagination. So, we guessed and we were right that it would work. So Wager is booked like that. And by the way, there will be the, I hope there will be a movie, Wager with Leonardo DiCaprio and I hope it will be in cinemas and theaters, next year. I don't remember the last information about this. So, it could help the book itself too, but I still think the book is much better because, uh, or not better, but it [00:23:00] helps to use your imagination, which is very nice process itself because, it helps you to, to thrive, uh, not just consume.

And we publish other books on history, by the way, The Dawn of Everything. There are many books that are very influential and we published a book from Yuval Harari, Nexus, but we already published book in Czech years ago from Harari, Sapiens, which is probably his most known book, and it is huge bestseller. Many million of copies sold. And he's probably the most influential living historian, uh, at this moment. And even there are some critics that are sometimes substantiated. We have chosen to publish the book again, uh, because the [00:24:00] previous publisher, uh, which published in Czech, lost their rights, lost their license. So we bought it, we had the highest bid here. And, bought both licenses for Nexus and Sapians too. And it'll be published again in three months. What's interesting that, for example, this book about history, Sapiens, is not published republished, it's not printed for, I dunno, several years already. But you can still buy it for twice or three times higher price than, than is normal book price. Because people want it. You can go to library. We have a great network of libraries here too. But if you want to own it, you have to pay a lot. So that's another signal that people still want to read popular books on history here.

And we want to deliver the [00:25:00] book or these books to our readers too, because now, as you probably know, the history is somehow repeating, or it at least seems like that. In our region, Russians are closing and we already lived through these nightmares. So I think history is even more important now after the peaceful period in Europe that has ended. So that's it.

Nigel Rawlins: Okay, so Nexus, Yuval, Noah Harri. Have you published that one or you're about to publish that one?

Tomáš Baránek: Yeah. It has been published. We have the same cover and it's really beautiful. It has been published the last week. It seems like it's one of the biggest book launches in our country. And the book is about history of Information networks from Stone Age until the [00:26:00] AI. I have thousands of notes from the book, yeah, I read it for two months, but it is not so heavy book to read. It's, it's much easier. But if you are a heavy note taker, as I am, you have a problem with this book because there is so many disturbing or interesting, or just funny ideas, uh, that I found or, or connected with my ideas.

So that I couldn't help myself and I had to do these notes. So this is something that I read and we published it uh just in time, I think there is still time to think about how to implement AIs in our lives and being aware of its drawbacks and even risks.

Nigel Rawlins: Well, it's 500 pages nearly, isn't it? So your editor or your translators had to do quite a bit of work on that. Obviously editing it [00:27:00] and, and keeping the gist of it. But obviously you knew that 'cause of sapiens there was the interest. But did, did you choose this particular book because of the issue that you wanted to have AI discussed in your country?

Tomáš Baránek: It's not just about our country. I think somehow even this is translation from English, we know that if we publish book in Czech, we influence mostly Czech readers. But, uh, if you help the particular book and author to be very known, and if you translate it well, you can have bigger influence even on, on the world events. Because, there are some books that are so terribly translated in Czech, but I think it's not just Czech, uh, situation. I heard about translations from English into Polish or other other languages, German, and [00:28:00] if it is not well translated, it loses the momentum and the book is like buried. And, the author is kind of, yeah, dead for the country, not always, but it can happen. So we consider our responsibility to publish it in the best shape, in the best way we can to support the idea.

And the idea behind this book is some things that worries us and me too, uh, personally, and it is if we don't understand how AIs work, or not just  AIs , but information networks. In general, we can have big problem with humanity as a whole. We can even not survive, because we can live in a much worse world. So, why is that? I studied informatics [00:29:00] and I love computers. I'm geek, so I have a lot of things here that, that I could show, gadgets and some e-readers and some, some electronics and which I test, review, write about, and I like it. I'm skeptical, but still, uh, I like it. But as for AI, I'm, I'm probably the most skeptical from all these areas that I talk about, because first there is something that we think that it is,

but it is not. There is some magical thinking in our view of AI, especially chats, Chat, GPT and other. I think it's mostly about the interface, how one blogger wrote, until OpenAI published the chat interface, there were LLMs, there were models and they were [00:30:00] really good. Like the first one, which we knew 3.5, it was nearly there, almost, uh, yeah, on the level that we know now, but, no one knew this and you couldn't use it, and was missing something. And now we have the way to communicate with it, how to talk with it and how to be yeah, bosses. We can say something and it tries to fulfill our needs. And this is the problem. These algorithms are based on fulfilling our needs and they do it instead of being truthful. And it's not something that creators of these algorithms can easily change, because first, machine learning is based on huge data and the data is biased.

That's the first problem. The second problem is that you can't take data that are in these models as [00:31:00] something that is put there as a database of facts. It's just a soup, like when you cut vegetables for your soup, and until it's very, very small pieces.

And then it's recreated again when you ask for something, the problem is, it doesn't know anything. It just recombines at some level, on some probability principles, what it already had, but sometimes the randomness seems like very creative and impressive, and, it's still just a probability model. And I, I am afraid, and here I'm back to the book, Nexus, that if we don't understand what happens inside and we consider these things, uh, like they are thinking, feeling, or just simply truthful, or reliable. We can [00:32:00] give them more power than they deserve, and it means, for example, the agents, yeah, and now we are in the hype of agents used everywhere. Agent is something that you put into, for example, your phone or your company backend software or CRM or something. And it should do something for you automatically or semi automatically. The problem is that yes, you can ask this agent to do something on a, on a regular level, but these agents are still just following these random probability models.

So If they find something that is similar to what you wanted, they answer right? It can work. But then some situation that they couldn't predict that was not in training data, occurs. Then they just create some answer or action. And that's a problem, [00:33:00] that was not predicted.

And, you can just miss it, if you don't see what's happening inside, it can start doing some things that can hurt your company. I'm talking about hallucination on agents level, and you can't frame it easily. It, it's a huge problem.

It's still not solved, but we all are using it everywhere and this is just one of the problems. When you imagine that this kind of agent will be uncontrolled using some framework, it can do harm to company, but then later it can do harm to big company. For example, the company that insures for example, that electricity flows into your home, and so on. And we saw politicians asking GPTs to create some, some outlines of their politics. And, then we don't see politicians asking this. It's even worse. They consult [00:34:00] decisions, but then they decide, okay, it's great.

So let's make it automatic. Yeah. Let's make these decisions without we being there and so on. So it's still like we are in a stone age, not in the AI age, uh, in our thinking about AIs. So we have to learn about, we have to go inside and see what happens. But Harari is more about something like what I said, but in a fantastic way.

It creates the context, the understanding of historical use and understanding of information networks. He says that information has no relevance. What's relevant is the connection. And even the truth is not so important because many, many societies were kept together just by lie or, but by story that was not true in a way that some of us want to understand what is true.

So, he really concentrates on how [00:35:00] we can be less naive in in understanding that people don't need to know true. They just want to align with some stories and be part of some information networks. And these information networks can be created by AIs or in very short time, it can happen on a huge scale.

So we can face much bigger problems we can imagine. And still we are not in an AGI era where these AIs are somehow self-aware or self-conscious, or conscious. It's another step closer to some dystopia, but it's not here. And still we can destroy everything just by not understanding what we use and what it can do to us.

Nigel Rawlins: Well what I'm seeing now is the use of Chat GPT by people who, who don't have a lot of knowledge say, kids at school, university students who haven't got a great deal of depth of [00:36:00] knowledge and they're using it, I'm even hearing cheating by using it. So you could go through university using Chat GPT and not really know anything. You've used Chat GPT to get yourself through if there's no way to check that, which means you're gonna have some people graduating who don't know how to do anything, potentially.

I was a primary and elementary school teacher many, many years ago. When we went to teacher's college, we had three teaching rounds a year where you'd spent probably 12 weeks out in the schools per year and then you had in classes. So, you know, it was pretty obvious at the end of three years whether we were any good at teaching.

And then you took a few more years to get experienced. In Australia they've changed nursing to where you go to university to become a nurse. Whereas in the past you learnt in the hospital. So the nature of nursing in Australia has changed quite significantly.

So what we're talking about there is relying on Chat GPT to [00:37:00] do the thinking for us and not doing the hard work of actually having to learn and think and write for ourselves. We are losing something. Now obviously the group I work with are my age and older, and you're in your fifties, so you've had a, a lot of learning through your life and you've read lots of books and you think about things.

I guess if you are using AI, you've got a sense of whether it's, um, giving you useful knowledge that you can use. But I'm also hearing because I, I work in marketing and SEO and all that sort of stuff, people just churning out articles using ChatGPT, now they might be getting really good at the way they prompt to get some reasonable articles out there. But if you are flooding the market with masses of this information, you're losing it. Plus most of us are gonna give up reading it anyway. And that comes back to it. If you've got a book, a real book in your hand, you know you've got [00:38:00] something that you can't be swamped with.

That's the thing I worried about, is we are giving up thinking and thinking for ourselves.

Tomáš Baránek: That's great question and huge question . I introduce myself as a geek and technology enthusiast. So I'm really not against new technology or against AI in general. I actually study machine learning on university again, maybe we'll talk about later, but, and by the way, why I studied this, because I want to look inside the black box and I already can see something.

I'm really bad student and I haven't made the last exam because I would need two days, not just 50 minutes for it. Yeah. But I would do it, but slowly.

Nigel Rawlins: We should, we should explain that you've actually gone back to university at 53.

Tomáš Baránek: Yeah.

Nigel Rawlins: Tell us about that.

Tomáš Baránek: I suggest this to everyone. It's [00:39:00] like, yeah, going back in time, it's my alma mater.

So, I can compare how it was then and how it looks now there, and my students, which are, yeah, 18 to 22. How they changed. And that's fun because we have some project to work on. I'm going just for a lecture per semester. I don't have more time to, to go there. And even this is very hard.

I don't have to go there. You just pay for being there as a regular student. If they have spots for you then you can go there. If not, you can't go there. But there is still some capacity. So I'm the only, uh, 50 plus there. But, I don't have to finish exams.

I'm just going there for understanding the terminology, the math behind the algorithms, [00:40:00] but not into each detail, because I know where to find it if I need it. Yeah. And because I code, I, I'm a, a hobby programmer. I can even do some simple machine learning, algorithms and try it on our publishing data, by the way.

It's interesting and impressive, but still it's like you see, wow, that's so stupid. The feeling that I had when the GPT started, I was impressed too. I was like, it's magic, and they really found something. And they deliberately spread information about not understanding what happens inside.

And, we are afraid it can have consciousness. And so they should be more careful, but not with the consciousness, more with leaking data and so on, which is by the way, not solvable this moment.

So, back to the question. I use Chat GPTs, I pay for [00:41:00] Claude and Chat GPT and some specialized tools to help me code. But it always depends what level content you want to generate comparing to what's the output needs?

If I don't use it for writing, for example, my articles on my blog, I love having bad, bad wording in my articles. I like writing them. I like creating, yeah, terrible stuff. Because I can look at it and then I read it and I can improve it, but I don't want anyone tell me how to write it.

But sometimes I use it for checking. So I just ask, okay, have I forgot something important here? And yes, this is the task these generators can do pretty well, because it's machine of generating following response from the context you put into. So it can really [00:42:00] find, for example, you missed the URL for what you promised on the beginning of the article.

So I use it for such a technical thing, but not for writing, because my writing has to be my, even with the flaws it has. So if you are bad writer, but still you have something you want to say, it can help you create the content to be readable for general reader.

And if you just want to say something, for example you are expert in some area and you want to share it, you can try or you can use it sometimes, or every time, for for generating the content that you curate, that the ideas are yours.

But still, the wording, in the way it can be read can be improved by these machines. So, It's probably okay, but I think I'm better than the machine in writing, including my flaws, so it can't surprise me. Then there are some [00:43:00] other areas that I use it.

I talked about coding. When I use it too much, for example, you just press the tap key and it finishes the whole function that you started to write. I don't enjoy coding then. I want to think about, I don't know why, but, I like it because probably it's my hobbies. I don't want to just finish the work. Sometimes I use it because I write in Closure script, but I don't know too much JavaScript, but sometimes I need to use it. So, because I don't want to learn JavaScript, I just ask, okay, write it for me.

And then I check it if it is okay, if it doesn't use any harmful techniques or so, because by the way, it can be a problem, for example, Deep Seek can inject some code into your code. And if you don't understand what the code is, it can work until it doesn't. Until someone decides your code stop [00:44:00] working and so on.

There are some threats that this can be misused this way. So, I try always to understand what's inside the black box in many areas. So, life is hard for me sometimes, but yes, when you use it often it feels like, I don't know, I just start feeling dumb.

Maybe my senses of dumbness are somehow anxiety based. I don't know, maybe I just am afraid because I'm anxious. I'm just afraid of being dumb here. So I, I see the triggers too soon. Maybe it's normal that sometimes you ask people and they answer and you are not dumb. You just don't know. But this is another problem with AI. You don't have to think it has some information for you.

The problem is that you have to check it always. And the problem of hallucinations is not solved. And it seems like it doesn't change at all from the 3.5. Even they promised, they need [00:45:00] money to invest so they say, okay, it's much better, and there is several steps thinking and so on. And that's funny because it, it is still not in, in the version of 4.5, it's still not able to give you answer that's on Wikipedia. It still hallucinates something that's not true. Even if you ask something that it was trained on.

But the algorithm can't have the same data. It's not Google search. It has to create new connections and that's great. It can be creative, but, this is a drawback too. It can't keep the data in the input form, in the training data form. It's not the same. So you can't consider this something that's reliable. When you want to ask about facts, sometimes it is, maybe mostly it is, but [00:46:00] sometimes it is not. And the less you understand the topic you are asking about, the area you are asking about, the bigger chance you will be fooled that you will get wrong answer and without noticing, that's the problem. And the more you are expert in some area, the more you will see how full of flaws are the answers.

And so you, you can check it for yourself. Yeah. Just ask about some things that you understand well and you can see how it is close to the true or the consensus. And then imagine this is the same in other areas that you don't understand and that you ask the most. So really this is something to be careful about.

Nigel Rawlins: Yeah, I totally agree with that. 'cause what you're saying is if you haven't got that depth of knowledge, you have no idea whether it's realistic or not. [00:47:00] And, and especially if you are an expert in area and you go outside your expert area, you can easily be fooled as well. And this is the biggest danger. I mean, the scariest thing I think you're talking about too is because it is so fast, it can be used to spread misinformation everywhere straight away because it can feed into the social media, it can send stuff to journalists to dominate their thinking in a certain way.

I think we're seeing quite a lot of ideology across the world, especially if you spend too much time on, X or Twitter as it used to be known. And you can imagine, AI feeding into that all the time as bots, which are there now.

So it could dominate your feed, it can manipulate the algorithms there and show you certain things. Yeah, look, I agree. That's why I think reading is really important. Taking your own notes, maybe [00:48:00] using your own notes to feed your writing. I like to work with AI. I use a lot, and I've gotta be honest, I, I use Grok as well, and I'm finding it quite smart. I'm finding it at a higher level

Tomáš Baránek: Yeah.

Nigel Rawlins: than chatGPT even, but I haven't been using Claude for quite a while either. But yeah, I'm very mindful that you can get overwhelmed with this, but I do aim to work with an older audience my age and probably your age and on too.

They do have what we call crystallized intelligence. They've got a lifetime of working in a particular area. I'm talking about professionals in some sort of area. They have a good sense of knowledge, but what I think Chat GPT or some of the AI can do is help them do a bit more, especially if they do need to spread the word about what they're doing.

So there is some benefits, I think, but you know, my concern is for is for the younger ones.

Tomáš Baránek: Yeah, I think it's like when we had [00:49:00] our first calculator. I remember, I just had some very old calculator with these red wired numbers in eighties or something. My father brought it from somewhere and the battery, nine volts, lasted, I don't know, one hour of use, it was terrible. It was Texas instrument, it was beautiful, and I remember having it in my hands as a, I don't know, eighth grade, seventh grade. I thought, okay, so now I win everything. I can do my homework faster and so on.

I don't want to be like the person that says don't use something because you will be stupid. My parents told me, okay, don't look at TV. And I hope I'm not so stupid as maybe I could be better, but still not sure if it was because of TV that I'm sometimes stupid. So, it's just tool and we underestimate kids, [00:50:00] in many ways.

For example, smartphone use, there is very interesting study from people from Maastricht University, that goes against results of some studies that say everything got worse when kids started started to use smartphones. Yeah, their mental health issues started with this and so on.

This study I'm talking about, they try to do it really properly and to forget about biases that you have because as a parent you want some explanation and you don't want to feel like you made this mistake with your kids. So, it's easy to just show the culprit is smartphone here and it's not you, and just to switch it off or something. It doesn't solve mental issues, health issue problems in young generation. There is more to it. There is more about their control of their lives that they [00:51:00] lost probably in last decades, because we are completely controlling their free time and everything. So all these gadgets and, uh, smartphones are just another way how they spend their time mostly with their friends. And the study shown that there is no difference between kids that used the smartphones and they which did not. And by the way, there was very interesting moment, that kids can stop using it by themselves, especially when they feel they want to connect in real life. Yeah. And they do it. They actually do it. So it's not like they are just always watching these smartphones.

They are connecting even. It is just something we don't, like, we want them to run or to do other things, but they do it, maybe not so often. [00:52:00] So I would like to say, it's not about it's all crap and it will make world worse. If you feel you are over using it and you feel like you are cheating your mind, cheating your efforts, you can just stop doing it and do it the old way, but then you have to do some work which the system created to be very tedious. And it's not something human-like, because even before AI there were not human-like bureaucracy things, which, Harari talks about too. Bureaucrats not always bad because it can connect people with the government. But, when it grows too much and loses its spirit, it doesn't work.

And it's the same, it's always about the level of implementing these tools and being, truthful to yourself in, in this. So I really [00:53:00] try not to use Chat GPT for things that I can think myself. For example, I try always to make it a bit harder for me, sometimes I know I need it quickly.

It's not about publishing something that I want to be authentic. So yes, I ask for, for some kind of help. And there is, for example, one very funny use, uh, one of my colleagues has some business partner of us which is really terrible in communication. She's like really toxic.

Yeah. But we have to cooperate with the different company and we rely on the service they do. So my colleague found a way not to be triggered by her constant rude words. She just asks GPT to rewrite her reaction to be more acceptable. [00:54:00] And uh then she found that even the person I'm talking about, she started probably to use Chat GPT too to answer more politely because maybe someone told her or she's just lazy to think about her rudeness. So she showed me the exchange was like between two very polite people that tell each other very not polite things and it was much less stressful than it was before. Yeah. So it can help. Yeah. But you have to contain it in some area and say, okay, here, it's okay.

But I warn you, be careful not, for example, using this with your wife. So that's it.

Nigel Rawlins: Yes. Be real, I think. Fantastic. We're probably coming to the end here. I was just gonna say with, with your calculator. We used to use slide rules when I was in high school doing mathematics and things [00:55:00] like that. So, you know, it's just the different generations and the technology. I was analog. I'm digital now, but we've got people my age who are still very analog and a frightened of this sort of stuff. But I think our conversation today is explaining, yes, you've got a brain, it's important to still use it, you know, don't fall for all the technology. You've still gotta think critically about these things.

Now, Thomas, how do people find out more about you?

Tomáš Baránek: Oh, they can find me on Blue Sky now. I migrated from X Platform, Twitter, or via my website, or my blog

Nigel Rawlins: I'll put them in the show notes.

Tomáš Baránek: Yeah, because I don't know how to spell them.

Nigel Rawlins: You're on LinkedIn as well? That's right. 'cause I, I see you, you all the time on LinkedIn. I keep an eye out for you, and the, the beaut thing about all of this, even though you write in the Czech language you're still able to press translate and it will tell [00:56:00] us what you're writing about.

And that's, that's how I came across your book, The wager. I thought, oh, what's that book? And that was just a fascinating story. So Thomas, thank you for talking about publishing. I'm hoping people can hear into what you are saying. There's a whole lot of things we're talking about business in there and marketing and thinking critically as well. So thank you Thomas.

Tomáš Baránek: Thank you, Nigel. It was great.