Radical Techno-Optimism with Prof. Dr. Ivan Yamshchikov

00:00:02

Like everything in this world, openness is apparently a spectrum. Technology is good. It was initially created for military. No, no, no. It's fine. Then the king of France, you might even have more sexual partners throughout your life executing and they were exploring. So if you're a designer, design the stuff for me, please design sneakers. Look, look, look. That's the thing. That's the thing. That's the thing. Have you heard of this thing called the internet? Right. Because um

00:00:28

I love it. This is your first goddamn episode. What did you expect? [Applause] [Music] Hello everyone. Welcome to another episode of Highi High Podcast. We're extremely happy and proud today to have Ivan Yamchikov as our guest. Uh Ivan is a research professor at technical university at Birdsburg if I'm pronouncing it correctly, but he's very well known in the uh in aim ML sphere having worked many years at different companies and leading educational institutions and so on. So we're very

00:01:19

very happy to have you with us. Thanks for having me guys. I'm also a co-founder of players which is a French startup that builds open source LLMs. We released the first family of fully open source LLMs where not only the large language model is open source but also the data on which it's trained is actually open sourced which is probably the only uh model of this type so far uh available to to general public. The real open AI what what is supposed to be. Yeah. Yeah. Right. You know uh the there

00:01:50

are like everything in this world openness is apparently a spectrum and uh it starts with like having word open in the name and then yeah and then things get more open. Um and then the probably the most open you can be when you talk about AI models would be everything starting from data finishing with weights. not only the weights of the model themselves but also the training loop and the data on which you trained and potentially it's quite important for high-risky things right so if uh something's dangerous could be done with

00:02:24

your model then having data out in the open could allow you to potentially prevent certain safety breaches because the data is available for audit and so we believe yeah so sorry would it be correct to say that deepseek is also truly open no because they don't publish the data on which they trained Ah okay interesting. Yeah, so they published the training loop, they published the weights, but we don't know what data they used um and to which extent, for example, it seems that it it used a lot of synthetic data generated

00:02:54

by CH GPT uh which in itself potentially could be uh so I'm not a lawyer but it could be potentially um having some legal implications in US because the license of OpenAI specifically said that you cannot use the data generated by OpenAI models to train Albany a competitor model. Uh so you know um the the data plays a crucial role without the data the model is useless and uh most of the time most of the models that are big enough can actually cannot actually publish the data on which they train because that would

00:03:31

probably make them criminally liable in a variety of jurisdictions u including I'm assuming like the most prominent models like open including open and gemini and all of those yeah yeah yeah very interesting times. Okay, we're going to speak a lot about AI hopefully today and about other topics as well. Um but one topic where we would like to start is like maybe um boring. Yeah, we had a preparatory meeting with you IA and you shared a lot of optimism techno optimism. I know you are tired of

00:04:07

convincing people that everything is fine. Stop complaining. Stop your fears. Uh can you just please dedicate like a few minutes to our audience like coming down sharing this optimism you already shared with us? No, I mean so let's let's be very clear. Um technology is good. Uh okay, it's a great if if you had doubts that technology is not good. I have um you know I I think like my my my selfident identity that I proclaim in Lincoln is radical technoptimist and technoptimist is clear especially with Mark Andre's

00:04:51

technoptimist manifesto. So it became kind of a clear term. Um I was selfidentifying as technoptimist long before but it was a very let's say narrow movement of people who believe that generally technology brings more net benefit to humanity than less. Uh after technoptimist manifesto became it became much easier to explain to people what it is. Now I personally think that the like my personal interaction with a lot of people uh made me think that I I'm not only technoptimist I'm radical

00:05:23

technoptimist in a sense that I suggest that the fact that technology brings more net positive than net b than net negative um is actually an axiom like I don't want to prove it to people I I I think it's just very tiresome experience and no matter how many um how many statistics you show uh no matter how much data you provide it never works because it seems that it's actually a core belief. There are some people that in deep deep down believe that technology is good and there are some

00:05:56

people that deep down don't share this assumption and so since different aiomatic gives you completely different uh worldview. Yeah. It's it's like believing in God, right? Some people believe in God and like no matter how much you how much time you waste explaining them that God doesn't exist, they still believe in God, right? And vice versa. Like there's one part of it that's controversial to me. Which one? Um um and that a lot of this technology is was initially created for military.

00:06:26

Yeah. A lot of and a lot of religions killed a lot of people. That's also true. No, but seriously there was an interesting talk by the guy who called I think George Freriedman like he wrote a very interesting book where he at the beginning of 21st century he tried to predict the geopolitics of the entire 21st century and predicted many things at the beginning at least um rather accurate and like he he he was like in one of the lectures and he's just saying like hey guys look at look at your

00:06:52

iPhone it's like complete military technology everything that's there like from GPS um to LCD to um re like u wireless networking capabilities, little battery, uh internet itself. So all of these technologies were initially like little cameras. It was all created for military and it's like maybe a coincidence that they're used now for whatever like is it though? Is it though? Right? Because um I'm sorry. Yeah. So is it though? Right. Because because for for a thing that's

00:07:29

uh inherently evil and developed solely for killing other people. It seems that humanity is surprisingly good in uh finding applications for this technology that actually do good things, right? I I think nuclear power is is an ultimate example, right? nuclear bomb was developed specifically to kill people on mass and the only country that ever used it to do that was United States of America. And by the way, this was of course a war crime that nobody answered for so far. Um because it was in the end

00:08:04

of the World War II and a lot of people died uh a lot of civilians. Um but but um then we found out that we can use this thing um for uh for energy. Guess what? Um and we still do. Uh so um even the technologies that are uh developed for military use first and foremost somehow inevitably and I would like to underscore it inevitably uh convert into some other use cases where they're used for human good. Yeah. Maybe maybe like I just I think no no let me let me finish the point. I'm sorry. Um that's an

00:08:44

important point. I I I I coined this comparison several years ago because this question keeps on rising which is like you know wars uh people use technology to kill each other bad. Yeah. No, no Of course it's bad. Um now the whole situation is a bit obscure to me because wars are not started by by technology. Neither wars are started by scientists. Wars are started by politicians mostly by certain political decisions of certain people um and they are ended by them as well, right? The whole the whole thing with this

00:09:24

technology is dangerous because it could be used in war to me feels a little bit weird because it's like you have an elephant in the room which is humans are inherently aggressive and tend to kill each other as a species, right? So like if you compare us with chimps, chimps are like very territorial. There's fascinating documentary where you can see for example Netflix there is a fascinating it's called chimp king kingdom. I think it's like four episodes of like just following out chimps and

00:09:52

showing how territorial they are. If a tribe of chimps finds another chimp from a different tribe on their territory, they rip them apart and eat them where where they find them. Right? It's very fast and brutal experience. So we as a species are aggressive. True. True. statement. How about we think what can we do about it? Well, there are multiple things we can do about it. Starting with culture, education, finishing with, you know, levels of testosterone uh in the in the population. Um, moreover, it seems that

00:10:25

those are actually happening, right? So, as we progress as a species, we become less aggressive. uh statistically at least there is a very good convincing book by Steven Pinker, the good angels of our nature that tries to show that in a long run humanity seems to be less aggressive and less people die through violent conflict. But that's a species problem. That's biology. That's not technology. Now, since we are as a species, we can also take a stone and hit each other with a stone. We get more effective in

00:10:59

killing each other than if we claw each other. That's true. But the ultimate decision to to use the stone to kill another human being is is is the biological issue um that is like deeply rooted in the nature of human beings. So I don't think that it's a valid argument to claim that technology that that that like technology is inherently evil because a lot of it is developed for military reasons. I would say a valid argument would be humans are inherently evil because they always think about how

00:11:28

to kill other humans and they're so good at it that they invent technology. Yeah. So if you stating that if you believe that humans are inherently evil, well I I don't think we can have productive discussion because I personally don't. I don't share this belief. And the proof against the statement is that despite the fact that some humans are inherently evil and they come up with all these ridiculous ways to kill other humans, uh other humans tend to actually find how these abhorent, ugly, dangerous things

00:11:57

could actually be used to improve our lives despite the fact that initially they could be developed for very nefarious reasons. You score. I don't know. You definitely score. But please, let me um just lower a little bit the degree of our arguments. I'm here responsible for um shameful questions and so imagine uh I'm a designer and a lot of generative AI uh is produced every day maybe while we are speaking a few new products came out will I lose my job what to do hey is there if you're a

00:12:38

designer design better use the generative tools to design more think of how they can be used in a new way. By the way, the amount of designers in human human population was zero 300 years ago. Yeah. So, so there there was certain like it was negligible, right? They were architects were artists. Yeah. The architects are still here. The architect No, the architects are still here. The the the artists are still here. Designers uh not so much. Designer is a thing that is actually actually very tightly connected with mass production

00:13:14

like once we started to produce things on mass then this thing starts to emerge. So no technology does like like technology created this thing in the first place and yes it will transform it as it transforms all the other things as we go along but I mean once again ultimately statistically 500 years ago most of the people on the globe were actually involved in one activity only which was growing food to sustain their hunger right so like 99.9% of the global population of humans 500 years ago were

00:13:49

actually really worried about dying of hunger. That was the major worry. Worry number two was dying of disease and only third one was probably dying of violence. Right? A and now um we don't have the hunger problem at this scale anymore. We have certain percentage of population which is actually less than 10% of the population that leave um in in um like under the threshold of global poverty. this this um percentage is actually stable at reducing and we basically produce so much food and so

00:14:22

much medicine that we leave twice as much as people 300 years ago. Uh and and and we live far better than than the king of France, right? So like king of France uh 200 years ago probably had the life quality lower than the average IT specialist living in in France today, right? So like if if if you're just a web designer living in in nonse I don't know or or uh or in in even in Paris you probably have higher level of life higher life expectancy than the king of France you might even have more sexual partners

00:14:58

throughout your life um because of Tinder and things like that if if if you wish to than the king of France 200 years before despite the fact that he could have all the concubines he could ever wish for because he was an absolute monarch. America. Uh so even on this very weird and shameful aspects of our life, statistically we we we progressed a lot. I'm not I'm not arguing it's a good thing. Uh but but I'm just saying that die from a tiger or people from other tribes. Oh, come on. Come on.

00:15:28

Tiger tiger. No, no, no, no. Like once again your your comparison is like dying from tiger. Not a lot of people died from tiger claws as well, but like just dying from an infested wound like dying just like you know cutting like you you're cleaning the fish and then you make a cut on your finger and then whoopsie doopsy you're dead in two in a month. This doesn't happen that much in the developed world anymore. Yeah, but there's still competition for resources in our like this civilization. Okay, we

00:15:57

have warm water everything. Yeah. Have you guys heard about Simon Simon Erliff wager? Not sure. So, so there was a famous very f so this discussion is not new and humans discussed technology for a while and famously Paul Simon uh Paul Erly and and and J what what's the name of Simon uh he recently passed away but I forgot his name. Um Paul Erley and one second let me Google we'll we'll find it and edit later in post-prouction. Yeah. So, um, there was a bet between between, uh, Paul Erley and Julian Simon. And the bet

00:16:33

was actually made in 1980, right? And so, Paul Paul Erley is this apocal like, you know, doom and gloom type of apocalyptic type of cult leader who claims he he is a scientist, but whatever. U, like, so he wrote this famous book, Population Bomb, about how all the resources will be depleted and we're going to die. Uh, nothing turned out to be true. The saddest thing you can ever see online is is the video of an old lady who read this book some somewhere in the 70s and decided never have children because why should I have

00:17:04

children? So they live in this gloom apocalypse and now she lives in the middle of like a bustling city where you can order food. It comes to you in a minute online. You can have all the entertainment you want on a buck. And like she she's lonely and she's incredibly miserable because of this um apocalypse cult leader. But that's not the point. The point is that they had a wager in 1980 and the wager was that basically uh let's make a bet that over the decade so so basically Simon who was Julian

00:17:33

Simon who who is who was an economist um he offered a a wager he said look choose the combination of natural resources that you believe will be more expensive in 10 years than now and let's see in 1990 if it's cheaper or if it's more expensive and if you lose, you send me certain amount of money and if I lose, I send you certain amount of money. It's very simple. This is how economists resolve any question, right? And so they made a bet and then Erlick sends Simon the money because in 1990 everything

00:18:05

everything that he put in this basket became cheaper than it was in 1980. And the reason for it is very simple, right? Technology. Because technology makes things more affordable. Technology makes things more affordable. Techn technology enables recycling. technology enables finding new resources we didn't we didn't even know exist I think fracking would be one of the examples right so a lot of people who are saying we won't have enough carbohydrates uh you know this uh carbon energy type of

00:18:34

fuels uh they just never thought that fracking isn't is an option and then fracking as a technology comes on board once again I don't discuss the external that it has for environment or whatever the prices but it turns out that no it turns out that like a lot of places have a lot of gas that we didn't even know existed before and and so in the end of the day ultimately even if we talk about the planetary level well I mean we have multiple planets only in this in this uh in this solar system right

00:19:06

and I mean like there is moon is very close we can mine radical radical no but like moon is no come on just like moon is minable right so if something that we needs so badly is not available on Earth, then the prices for this thing go up and they go up till the moment it becomes viable to mine the moon for this And once it's viable to mine the moon for this people start mining the moon for this and they develop a bunch of technology to make mining moon easier and cheaper for this So the price of this thing goes down and it

00:19:43

becomes more affordable. This is how marketing works. ironic like even though we have all this abundance of food and everything and sex as you said uh but still most of the food that's in the supermarket is arguably not really that healthy still much better than have no food right I'm not saying we're perfect safe yeah it is safe I mean we are improving my point is that like you know I guess the the core idea here is not that we are perfect and by no means the core idea is that we are improving and

00:20:12

we're doing it through technology like things get better not worse and they get better logically because we are using technology to make them better. That's my main point and no there is no end progress and just just one last thing regarding this I mean whenever like when I mentioned solar system your reaction was like oof that escalated quickly. Well I mean just to be clear this attitude is also not new. People had this attitude in the 60s uh in the 20th century. This was basically a mainstream

00:20:46

thing. It was not controversial to say that humans should colonize space. It was not believed to be a wild crazy idea. It was a mainstream thing, you know. Um and and then due to a very very strong consolidated effort of people who basically in my view look at humanity as as as something evil fundamentally like because this kind of type of apocalypse scenarios they basically always assume the worst about humans. Yeah. They they assume that um the they make certain suppositions about humans and then they project those

00:21:26

oppositions and they predict the very very ugly scenarios and they managed to shift this discussion from how do we do it to we shouldn't do it ever because we're so inherently evil and bad and like that's not the way to go. In the meantime like I think in the 60s or in the 70s there was this cosmologist Kartev in Soviet Union who proposed the Cardartese of scale which I which I love because it gives a very different point of view about like technological thinking because now whenever you talk

00:21:55

about the person who's technopessimist the people tend to speak about the planet or even like habitat of certain species within the planet and so on and so forth. Um and they claimed that we are so far into this space of technology that like there's no point we pass the point of no return or it's very close and blah blah blah. In the meantime, Carter scale was saying that like there are three types of civilizations. First type on the cardiff scale is the civilizations that fully controls its

00:22:19

own planet. And by full control, I mean you control weather on your planet. You control climate on your planet. You control all the resources of your planet. Your planet is available for transportation. You can go anywhere you want on your planet. And we are not even there. We are not even like scale one on cartes of scale. Like there are places in Amazonia where we can literally not go cuz cuz there's no way to to to get there. We we cannot control climate. Uh but we like we we don't but but we we we

00:22:49

can like physically we can because in the end of the day the planet is a closed system um with with certain energy inflow from from the sun. That's it. So we can control it. Physically it's possible but we're not even close. And then the second scale would be fully controlling your solar system. And the third one would be fully controlling your galaxy. Now we are very very early and we are not even close to like you know grade one civilization on this scale and and the people are just

00:23:18

like let's not do that. Well, I see zero arguments why we should not do that because I think it's awesome. I want to reduce the degree of our conversation. Let's talk back about So coming back to the designers. So coming back to the designers. Yeah. If you are a designer, design stuff because by the way, uh I recently tried to find futuristic clothes on Etsy, right? Very practical thing. It was 14th of February. I wanted to buy something cool for my wife and I don't want to buy this shitty steampunk

00:23:53

that the Etsy is full of which is aesthetic of 100 years old and I didn't want to buy this kind of minimalist of the 60s from like this kind of you know bow house type of minimal dresses which is aesthetic of also like a little bit less than 100 100 years old and I was just googling something that would look new something that would look like a clothes from the future not from the damn past. I couldn't find a single thing on the whole Etsy. The best thing I could find looks like, you know, this kind of very

00:24:26

ugly Star Warsy type of which was designed in the 70s. So, if you're a designer, design the stuff for me, please. I thought it's your job. You're very bad at this apparently. Where where is new stuff? I mean, like, have you ever tried to find a sneaker that looks nothing like other sneakers that you wear? Have like come on, what's up? I mean, sneakers look very different from the shoes people weared in the 19th century. Our shoes look very similar to the stuff I wore in the '9s. Where's the

00:24:56

new stuff? Hey designers, you're now you have so much workations of fashion which I think a lot of it is like it needs to be like the signaling that's in you are belonging to the group and not everyone likes to be like you know like like uh rapper genius rapper took the word of shoes by storm just because he thought hey maybe I need to design something that doesn't look like the stuff people wear for 100 years old, you know, like maybe maybe cool kids doesn't want to wear stuff that their grandpa wears, you know,

00:25:33

that's what a crazy assumption. People who created Crocs also. Yeah. Yeah. Yeah. But like those are like two things in the world of shoe wear that actually stand a little bit aside from everything else. So, hey designers, how about you design stuff? I mean, this windows look very much like the windows I had, I don't know, since time in memoria. this table, this chair, this lamp. How about you designers design No, I mean, I heard you're good at this. I think it's a completely different type of frustration. Of

00:26:06

course, we are like very um very confident about having food in our fridges, having clothing even if it's not like too new and futuristic, but still we don't know how to spend our time. I think this is our greatest resource. That's a great and that's a great followup. You know most of the time like what did people envision in this AI will do it better and no no like should I try look look that's the thing that's the thing that's the thing so how people envision this idea the people

00:26:40

envision this idea of like we solve some basic stuff basic needs and then humans can do what humans are best at coming up with new stuff both technologically and aesthetically and express themselves in a new way right so okay we're here. So, let's do that. A and and you ask me like, oh, but like AI is taking the job of designers. Well, AI is not very good in finding new things, by the way. It's it's it's making things much easier and cheaper to reproduce and to produce on scale, but new things, you

00:27:17

know, the thing that designers used to do uh by definition. So, maybe that's that's what we should do. That's create that's what's creating value, isn't it? Like I think that yeah, that's a very interesting segue into maybe the next topic. But let's talk about then um how people can how someone can become a designer because over time like it was like through apprentice ship through working with masters and learning from someone who's more experienced. But what

00:27:42

we see today is that like geni things are actually taking like eating these jobs from the bottom in my opinion. So they're already doing what the job of a junior designer would be like simple things like transforming existing styles and whatever and then likely they will move into a mid-level senior like maybe even principles we don't know where it's going to stop if it's going to stop anywhere. So how can someone become a designer if junior designers are no longer needed? Well, have you heard of

00:28:10

the thing called the internet? Like you can learn a lot from this stuff. Like you can and like it costs you like next to zero and like you can literally go and learn from the best of the best in their craft directly. Like you can watch the works, you can take tutorials, you can do stuff and test it. And yeah, by the way, there is another thing. It's called Genai. And like if if you're puzzled or if you're stuck with certain thing, it can help you. And even if you prompt it correctly, like for example,

00:28:38

you say, "Hey, I want to do this front end." and I'm learning this particular framework and this is my code and I don't like how it works. Can you optimize it? But can you also generate commands to explain what you did and why you did the way it you did it? You can also get tutorship from this thing for like 20 bucks a month. And you know, isn't it great? It's amazing. Uh but like in order to to become like senior designer, you have to arguably learn for years. Yeah, sure. I

00:29:07

think it's very difficult um for someone just with access to internet especially by the way because on internet there are so many um things that are much more attractive and funny and interesting than I know learning about some craft. Uh well that's the thing right so uh craftsmanship is is not is not easy and by the way if you read like the standard narrative around craftsmanship in like the most traditional cultures like if you if you read Buddh Buddhist um stories about like how people became

00:29:37

masters of their craft or if you if you especially I think in Southeast Asia the philosophy of crafts craftsmanship evolved and stayed with us because I think Europe was the first industry iz and so this industrialization kind of destroyed to a large extent the philosophy of craftsmanship that for example European master guilds had right so we had guilds in Europe which had this philosophy of craftsmanship and now we kind of learn about this philosophy more from Southeast Asia than from the ancient European roots but this

00:30:08

philosophy is always based on two things is that you know practice makes perfect that you you need to actually have a diligence and and and um internal motivation if you want to become a master and that's the only thing that master cannot give you. Like most of these craftsmanship tales they're always about this idea that on one hand side you need to be dedicated yourself if you want to become the master because it's philosophically prerequisite for one. Yeah. You have to convince the master to

00:30:40

even pay attention to you. Exactly. Exactly. Right. Right. Right. Exactly. And like I don't know one of the the epitome of this idea would be this fantastic documentary that's called uh Jirro dreams about sushi. You know this this sushi master that that's explaining that like for the first 20 years in in my shop you just cook rice and then once you're good at cooking rice well then you you're allowed to start cutting fish 20 years into the into the craft. Right? So so this concept of masterership is

00:31:10

there and it doesn't change. Moreover, I would actually argue that we are back to the way it was before we had this strange phase in our development when we had mass production and revol like a series of social revolution and technical revolutions that brought us the the institute of public education because public education to a large extent is copied on the proian system of public education and Prussian system of public education was a big innovation in the in the uh in the time when it was

00:31:44

developed mostly aiming for two things military. Yeah. So the state wanted to to to have people who are competent to serve in the military and the ability to run factories and and to have uh skilled labor in them. And this inflation of skills that you mentioned is also not not new. For example, my uh university here, THWS, it's um it became it was it's one of the youngest universities, like one of the young youngest places for higher education in Bavaria because before the 60s, it was kind of a

00:32:17

midlevel professional education for people who are doing certain production works. But as production becomes more and more um complex, technologically heavy, you need to make people more qualified, right? So it's sort of this progression of like at first we we just wanted people to read and we believed that there would be cool thing and particularly Protestants did a lot of it because Protestants religious beliefs were grounded on the idea that you need to read the Bible on your own to be in

00:32:47

touch with a god and then turned out that once you can read you can also learn a bunch of other things and by the way there's a lot of this demand in the economy so we build this school education and then oh yeah by the way once you have school education there's a demand for even more skilled labor and now we can build that and and and so we kind went through this full cycle of having this very established system that kind of takes care of you since you were born till till till till I don't know

00:33:11

your 30s sometimes and arguably uh it's um it's very well suited for industrial societies but in the post-industrial society I would say that it's very inefficient and it's a very inefficient use of taxpayer money bringing not that much to the taxpayers's back uh because we are back in the situation where what matters is this direct craftsmanship transfer from the master to the student. But we also have this amazing toolbox of of of various technologies that allow to simplify this transfer that allow to

00:33:49

have this multiple uh learning experiences and we we can kind of scale certain mastership on some levels on one hand side and on the other hand side we built more and more tools that allow people particularly in digital area. Right. So we we are not not near enough to build physical things the way we can build websites now. Yeah. But like in certain areas of economy, specifically in digital economy, you can prototype very fast and you can learn by building very fast. I would argue that like a good designer education would be an

00:34:21

education structured around the creation of certain digital assets that pay off the education. Yeah. So ultimately like you build stuff that finances what you learn as you build stuff. that that that would be the most efficient way to learn, right? wouldn't it? And we are actually quite close to this bifurcation point like in some areas we we're quite close to actually uh having a a possibility at least of a learning trajectory whereas the amount of stuff the student builds through their studies

00:35:00

creates so much value for the society that it pays back on the costs of the education because on one hand side the costs of education are drastically lower due to AI mostly and the internet and on the other hand side the productivity of the student is drastically higher because of the same things. A and and and and the current education system is very far from this trajectory. I would say it has a very huge overhead of things that I don't think exactly fit in the 21st century. That's interesting to

00:35:29

hear from someone who is like almost embedded in the system or like plays in not almost I am a part of it. Yes, of course. No, but what do you describe this flywheel almost? That's super interesting. Like when so personally first like on the moral side I actually pay very close attention that that my total amount of taxes that I pay to the government actually exceeds the amount of money that I get the subsidy from the government for the job I do because I I'm not only a professor I'm also a

00:35:55

founder on the side. Um so like I actually try to be moral about it in a sense that I don't embezzle taxpayer money for doing things that I don't believe in. On the other hand side, I um also try to experiment as much as I can within uh the reach because Bavaria has the freedom of research and and teaching that's actually in Bavarian constitution. And so I joined THWS partly because I actually see the same spirit from many of our colleagues. Um we are university of applied sciences which means that to be a professor here

00:36:26

you have to have at least three years of industry experience. Bavarian's laws also allow professors to go back to the industry for half a year every five years so that we stay in touch and we understand the industry needs and so like we we we understand what we preach. Um so all this um parts actually align all these values and ideas align very well with what I believe in firmly. So I think that uh to some extent there are a lot of healthy ideas uh in in the current system but it's just there is a lot of overhead and

00:37:03

inefficiencies that actually could be u trimmed and that's what I'm working on myself [Applause] right I remember when I graduated um one thing that I did recently was actually teaching at the university where I graduated from and I wasn't a professor but I was teaching my own course uh that actually evolved into course on product design and I tried to completely like do it the opposite of what I didn't like about the education. So yeah, exactly. We were like I I didn't have exams that

00:37:34

where people had to I don't know whatever recite from memory different things. It was like all project based and uh they had to do this project in order to get their grades and it was interesting. Yeah. I tried I tried to share some of my early experience of work. Um but I'm curious then like if we just expanded into what's so just just coming back to just a small example, right? So how uh for example I I have a course that I specifically designed for our master students not only for AI

00:38:00

masters but also for other masters that for example work in this area of digital society and and this course is called um entrepreneurship for engineers and it's one semester course that's based on the idea that people actually prepare pitch of their startup uh and develop of their startup and uh the criteria of getting a best grade which would be one in Germany Germany the best grade is one and the worst is five you fail fail if you get five. So to to get one, they actually need to have either the first customer

00:38:29

um paying for what they built or the first seed investment, right? So so far nobody got one but that's the that's their um direction in which they need to go and then the whole semester is structured through some project based work where they have several sessions of something that resembles Shark Tank where I ask my uh peers from my professional network who are founders themselves or who are actually early stage investors, angel investors to join for a zoom call to see the pitch decks of the people to the current ideas and

00:39:00

basically say okay this is something that I potentially invest in and this is something I would not and so the com the the the teams that didn't manage to get through the Shark Tank uh phase they have to pivot or they have to join other teams so there's merger and acquisition round after every um every uh Shark Tank type of Zoom call because you actually need to be validated by external people on what you build and if the external people don't believe you well you failed then you need to do something else right

00:39:27

a and I think as an example um this course creates a lot of value for the students at least this is their overwhelming feedback uh there are some out startups that actually spin-offs of the course they don't happen within the course so far uh but but they happen after so at least two companies started after the course um and I did I do it for the second I did it twice so far here in Verdsburg I'm doing it third time this semester but but there is like clear tangible value both subjectively

00:40:00

on the side of the students and objectively in terms of like just entrepreneurial activity that happens as followup of the course and I believe that this is the direction in which um you can kind of combine and I also use a lot of outside materials I use books I use y combinator YouTube channel as a course material literally so I don't I try to reduce the amount of time where I talk to the whole class in front of the whiteboard and I actually try to offer them educational materials that they

00:40:27

have to study on their own while we use the time in the class to interact directly and to do this type of mentorship um craftsmanship type of interaction where they can learn directly from my personal experience or from the personal experience of other people who I invite. And I think in this configuration we create first of all we create subjectively at least I would say far more value and then and on the second side I think that pushing more courses in this direction by example I think is a good thing because ultimately

00:40:57

in my my belief that this is much closer to how higher education should function. Moreover, this was the model for example that ancient Greek used, right? So in ancient Greece famously all the schools like the term academia was actually a private school ran by Aristotle and and like people paid to go to the garden and walk with Aristotle in the garden and just to talk with him, right? Um that that was the whole model of education in ancient Greece and Greeks figured a lot of things, right? They figured out

00:41:22

democracy. They figured out um a lot of fundamental ideas that still uh power our scientific uh thinking and all these things happen partly through this mentorship type of interaction without any systematic uh coursework without you know sitting at the desk and writing stuff and memorizing them. So so this idea is not new we just forgot that this is how education works best. Yeah. And can you assume like having AI professors uh in the university like we now have so many language tutors or u other AI

00:42:00

agents uh providing education. So how do you imagine it? I would be I would be very happy to be replaced. I mean I don't I'm not a designer. What will you do? How will you manage your spare time? Well, I got a bunch of things that I want to do, right? So uh design sneakers. Yeah. like you know um probably we'll go into design apparently most of the designers don't design the sneakers I want to see no but like uh jokes aside um there is first of all research something that I generally like

00:42:31

a lot I like teaching but it's actually quite taxing right so like interaction with with the class it's actually very like it's very demanding uh on a physical uh level uh so if uh to some extent teaching could be autom automated so that I have more time for research. I'll be happy to do that. Research, open AI, deep research modal. Yeah. Well, it's amazing. Yeah, I will. Of course, I will do research with EI. Of course, I don't I don't plan to do research without it. Uh, no question about it. It

00:43:02

also allows me to research far more and get far more results then guys, it's okay. Um so um I don't think that research and I think here is one clarification. So um so far if we do have free will I don't see any technological uh gadget that could create free will within AI system. Yeah. So if you believe that humans have free will, so far I don't see any technology that enables free will in the machine which means that all the creative activity uh which is centered around free will

00:43:42

fundamentally uh is available for humans for a while uh unless we don't have free will at all unless we are biological predetermined complex machines well then the whole question becomes kind of obvious which is what Seapolski is arguing for I guess but I haven't read his like full book on I think he's not that reductionist. Um, but but but but I I think that's that's a very big assumption, right? So, uh, once again, and this is the question of belief, so I believe that we do have free will. Uh,

00:44:12

if anything, using the the Pascal argument of like God exists. So, famously, Pascal formulated that if God didn't exist, inventing God would be a good idea because it actually helps with a lot of issues that humans have. Believing in God is just utilitarian, has certain utilitarian value. So similarly believing in free will gives you certain utilitarian value in my opinion. So uh and since I do believe that I have free will and I don't see any mechanism that would enable AI systems with the free will so far I

00:44:41

think that uh fundamentally creative uh activities are available to humans. Moreover, I would argue that humans don't do enough of creative activities. We don't do enough early education on how to be creative. I think fascinating statistics that I I came across. So my daughter is one and so we were thinking about different you know early education systems and surprisingly enough Montasauri advocates have a very interesting statistic that goes for them that most of the unicorn founders in the valley

00:45:14

were Montasauri educated when they were small kids. Wasn't it Waldorf system? No no not Montasur system. Uh I mean let me let me find the stats. One of the plausible arguments um take your time, take your time. The toys are available for kids and the kids needs to choose um they have to make their own decisions. Yeah. And so Jeff Bezes uh um for example is a part of the so-called Montasauri mafia. This was a term coined by a Wall Street Journal in 2011. So, uh, Jeff Bezos, Lori Page, Sergey Breen,

00:45:51

Will Wright, uh, the designer of the Sims video game, uh, they are all, uh, Montisauri kids. Uh, and there are bunch of other founders who got this Montasauri early education, which actually highlights freedom of choice by the kids and certain self- agency. Yeah. So, um, the idea is that the kid needs to choose the toys they're playing with. the kid learns the discipline not through the fact that there is a structured education when the the class is over and the next class starts but rather with the fact that if you play

00:46:19

with a toy and you finished playing you need to put it back where you took it. So um the system has certain proven track record to be developing certain agency of the kid very early on and agency I think is very important in context of creativity. So famously creativity needs agency. A lot of very creative people are very difficult to be around because they are so agentic that that they really don't tolerate um this kind of infringement of their freedoms and and and so um I would say that our

00:46:53

current system actually prioritize compliance rather than agency uh for the most part and may I'm also not saying that Montasauri is the only right system. I'm just saying that we do have statistics that seems to hint that if you prioritize agency in kids education, then the kids tend to be more self-sufficient and more creative. Um, at least this is anecdotal evidence. I'm pretty sure that there are also some stats on it that are probably uh more comprehensive and maybe the story is not

00:47:24

so simple as I as as I believe it to be, but at least we have some preliminary data on on the fact that it does matter how you teach the kid. Yeah, it's really difficult to research because you have to have like a really longitudinal study in order to understand the effects. Indeed. Indeed. But I was going to say like that actually resonates strongly and we ourselves um our older daughter goes to what's called Sadbury School which also like blows mind of my friends when I talk about it. But that's a

00:47:50

school with no teachers, no classes, no curriculum, no grades. It's just like an environment where kids can work together or do different projects together like and there are lots of facilities like the arts room, computer room, library like fantastic sports facilities, musical room, playroom and they can and I know the kitchen and they can just choose what they want to do and decide it on it like themselves um set their own goals understand together how they can achieve it then I know make their

00:48:19

own judgment as to whether that was successful or not but I wanted to expand this into probably a question to you like what do you think then are the skills that are important in this century like what what skills should people learn or develop or build within themselves to become successful in the future? Yeah, sure. So um I think um first and foremost certain learning flexibility. So the ability to learn new things and um you can foster this through several very applied clear aspects. on one on one um side you need

00:48:55

to actually work hard on resilience to failure right so learning through failure is the fastest way if you if you fail you learn way way more and way better a little bit painful though so painful just trust me I'm u sorry uh yeah this is I'm so anxious about this topic because this is it is painful uh at the age of 27 I just started learning playing piano and playing tennis at the same time and playing to uh manage sales in a company and it just takes so many resources and so many energy and my

00:49:36

therapist says it's a best um exercise to feel yourself loser and to learn through this through this um state of mind and um yeah it's it's hard if Do you have any any advice on that? Uh it would be just valuable. Failure is amazing. Fail more. Fail more. Like like think about like I would I would say that if during your day you didn't have a single possibility to be exposed to failure, you wasted your day. Right? Because exposure to failure means that you are very close to fulfilling your potential. Right?

00:50:19

Because if you do stuff where the chances of failure is zero, then you literally underinvesting. Yeah. Only if your failure chances are close, like non-trivial. You actually strain yourself. Yeah. And then if you don't fail, well, good for you. If you fail well, you learn a bunch. But like if your day is a day where you had zero exposure to failure, then you literally, you know, didn't fulfill your mission as a human being arguably, right? Like you could do better than that and you specifically decided to play it safe.

00:50:55

It's so funny that as people become experts, uh, they don't tolerate failure as much in their area of expert. Oh. Oh. I think it's a very it's a very wellstudied effect and I I have personal take on it. Um once again this is not a scientific statement it's my personal belief which is that on one hand side as you get older your neurop plasticity reduces so it's much harder for you to learn new things and much harder for you to tolerate failure and I think what is interesting is that for example

00:51:24

scientists uh it's this one is proven people who do like intellectual heavy intellectual work tend to have significantly lower levels of certain senity issues as they age. So they they tend to be having sharper minds longer and and um this is something the psychologists call the theory of mind. So some people believe that the mind is something that's given to them and some people believe that mind is a muscle and like if you exercise your mind you stay sharp and fit longer and and to some extent um it

00:51:55

seems to be reproduced at least partly that like if you learn through your life and if you are not afraid of failure and if you keep on failing and learn from it then you you retain this possibility to learn for longer and you stay sharper for longer. So I would say that this this expert shell is actually certain in vulnerability that's on one on one hand side physiological like the person feels insecure uh that they are actually not so sure about their mind and not so sure about their expertise and this insecurity is

00:52:26

projected into the world through uh you know this kind of shell that's backed up by credentials. Um that's one thing and another thing I think that there's also cultural component to it. You know um Ken Robinson has two fantastic TEDex TED talks about education. And in one of them he mentions how he took a panel discussion with the Lai Lama and Dalai Lama was asked some question and and he paused for a while and the silence became uncomfortable and then after a very uncomfortable period of silence the

00:52:54

llama said well I don't know and he says that's such a liberating experience because western culture to a large extent assumes that if you're an expert you know things and that's that's a very like humans are social animals so there's certain punishment or at least anticipated punishment that you that that that there is there if you feel oh I'm not compliant oh I'm I'm stepping over the line oh people expect the answer from me and you say I don't know um and and so getting being able to to

00:53:23

have certain level of intellectual honesty and bravery to overcome this cultural push I I think it's difficult not every person manages and coming back to just like I I just want to return to this question about like failure so two practical things one thing in the companies where I work I try in introduced this idea and pardon my French of night and night is is is a corporate event we do it once a month and the idea is that every it's it's uh in the evening somewhere let's say Friday evening or maybe Wednesday

00:53:54

evening uh so uh and we meet with with the team and there are several presentations about fuckups uh so people people actually present um how they hire yeah yeah so so like just you know just a story. It could be a personal story. It could be your professional story from your previous position. Once you have certain level of trust in your organization, it could be even the actual story about a within your organization that happened before. Uh so the people learn from each other and they also learn to tolerate

00:54:24

this thing and and and foster this culture of trust and you know uh build upon assumption that everybody can uh make an error and can make a mistake and the important thing is that you learn from it and you analyze it and you propagate the information about it to the others. So on um on average your organization becomes more resilient and more more uh stable and um of course you need to start with sea level people right. So if sea level people don't admit that they up themselves then that never works like then it becomes a

00:54:54

very weird um flagagillation type of activity uh that that I I don't believe is right. But if you actually start with senior people and you mix them up uh from time to time into the pool so that new candidates or new hires as they come to the company they they adopt this culture of being honest about things that didn't go well uh being able to analyze them and being able to share them with the peers then I think it does a lot of good. Yeah. So this is one practical very very I think rather

00:55:23

sensible so much against this high performance culture where it's almost like uh impossible to fail and you have to deliver amazing results all the time from I'm not a fan of I'm not a fan of high performance culture. I'm a low performer. Thank you. Thank you. I'm a low performer once again. Thank you for saying dudes. I I'm not I'm not afraid of fuckups. So yeah, if you want to hire me I going to be underperforming. I'll be I'll be the good thing I will not overpromise too, right? So I will not

00:55:49

overpromise and will definitely underdel. Be be aware about it. But I think I think this high performance culture is actually very interesting because it's a very interesting cultural phenomenon, isn't it? Because like any person like once you get high performer drunk, they hate it, right? Like if if you were in the bar and you had a few shots with a high performer and if you ask them like, "Okay, so dude, like what's your feeling about all this thing of like the stuff you do?" Most of

00:56:21

them would uh would actually say, "Oh, that's actually very tough. I'm I'm so under pressure. It's so stressful. Uh my therapist tells me I I gonna die young and stuff, you know." So, so it's a very interesting self perpetuating evil that people recognize as evil on an individual stage, but it's the thing that's well described by sociologists as the so-called spiral of silence. um which is that if you have 10% of very loud people in the room that that are stating certain position loudly clearly

00:56:56

in a in a rather aggressive fashion so that everybody else believes that contradicting those people might have some negative social implications for them. Then you can actually have an organization where like 10% of loud people stating an opinion that nobody shares could kind of modify the whole culture into their stated opinion. Even if they they themselves state this opinion partly out of fear that if they don't say the truth like like that if they say the truth and everybody like you know even they themselves might not

00:57:27

share this opinion but just choose this tactic as the way to kind of socially advance within the organization because they believe this is the desired behavior in the organization right and and and so I think this well first of all it's documented phenomenon it's well known second of all it's sometimes it's actually useful indeed if your organization is in danger, under heavy stress, if if it's a life and death type of situation, you need this high performance from your uh team. But I

00:57:54

would argue that the organization is actually honest about its goals and about its culture and has this culture of working with errors, being able to propagate this information, first of all, would have lower chances ending up in this dire situation when like you actually need to work 20 hours a day, 7 days a week to to survive. Yeah. Because a lot of these things actually ultimately happen because somebody up and quite often somebody high up up and is not ready to admit that they up and instead makes

00:58:19

everybody else to cover up for for their And and and second thing even if just this is the market reality like you know there's certain paradigm shift in the market and you as organization are under enormous stress. I would argue that having honest relationship with your team and saying you know this is how it is we need to pull it off. this is sort of an emergency type of deal and we need to actually work on it the way we uh never worked before because it is a life and death situation. I would say

00:58:47

that once again the chances to overcome this crisis in the organization that has this healthier culture around error are actually [Applause] higher. Let me just go go back to the original question. what skills are really important for this century and said like learning quickly failing learning quickly learning from your mistakes having certain tolerance to errors uh because that's the way that's the fastest way to learn I would say um being able to uh interact with other humans and uh tools yeah so um having

00:59:25

the understanding of relevant technology landscape and relevant expertise aspect and being able to leverage those two Right. So not only you need to systematically be aware of how your technological landscape change, you also need to systematically work on your network of professional contacts that you can use when you need uh extra expertise or guidance once again to learn faster or to adopt certain tools faster or even to monitor this tech landscape. Because to some extent once your uh professional

00:59:58

network is stable and and big enough you inevitably start to be more aware of the landscape at large because now you're a part of this collective that sifts the information about relevant professional topics and shares with each other. Right? Um so this um awareness uh to a large extent both of the technology and the professional skill set um is something that's actually quite hard to develop because you need a bunch of soft skills uh on top of like technical understanding of what you care for. Now,

01:00:30

um the third thing that I believe is is important is um being um being able to allocate your timing in uh like consistently so that you improve and explore. Yeah. So, professionally, people either explore because, oh, I lost my job. I need to find a new one. So, what's up? What's going on? Let's learn some stuff. or improve because I'm on the I'm in this high achiever type of culture. If I underperform, I lose my job. So everything on on on red. Um and I would say that in a once again in a changing

01:01:13

landscape, you actually need to healthfully combine the two. So you constantly allocate I don't know 20% of your time for exploration of new things and trying things you didn't try before. And you need to actually make it in a way like you know if you have to-do list or whatever it should be on your list. It it's like it's not optional if you want to be healthy within it. It's not optional. Exploration is not optional. And on the other hand side the execution part uh once you become good at it once you

01:01:44

incorporate this systematic exploration then very soon uh depending on what you work on but like probably within a month if you systematically I don't know every week dedicate let's say eight hours of your week to actually explore things relevant for your job you find some things that you want to try and then you can actually merge the exploration and execution in a very natural way and then you'll actually find out if you if um I mean at least most of the time I never heard a story when people tried to merge

01:02:13

this exploration and execution and then they became less effective like that's that's actually very funny because you you intuitively think oh I'll waste eight hours a week for exploring things instead of like delivering and yeah you will reduce your productivity by 20% for a few weeks at the best but in reality there are a lot of people a lot of smart people that think about your industry and come up with a bunch of tools that you never tried right? Because you never had time to look around for them and

01:02:40

test them for your needs. And once you actually dedicate those, I don't know, 80 8 to to 20 hours just to look around and list those things and say, "Oh, maybe I can try that and I can try that." And by the way, most of them you can try for free. And then you like implement them in your work routine for certain tasks. And on average, probably in a month or two, you'll see that your productivity actually increased because yeah, well, now you pay a few bucks for a few Gen AI tools or maybe you manage

01:03:07

to dedicate some time to set up your Kendly link so people can book meetings with you faster without this back and forth on the chat that you do 75 times before everything fits. And maybe you did a few automations of exporting certain things in in in some I don't know, notepad or whatever. So, you have the the the automated update of your to-do list based on your work email and you know, a bunch of these small things that you just tried because they were around and you had those eight hours to

01:03:35

try them and oo surprise, your productivity actually grew and probably not by 20% but by more and you still have those eight hours to continue exploring and testing new stuff. Um, delivering more than you delivered several months ago. Um so yeah I I think you know what's funny like some people stay in this cycle of always um like executing and never exploring but some people also may get stuck and always exploring and not get to execute anything. Ah yeah that that's also I think that's a great um thing. Yeah. So

01:04:06

um and that's by the way this culture of underperformance. So I actually believe that the culture of underperformance is very good because it doesn't allow perfectionists to survive. Right? So if you're perfectionist, you cannot fail and you cannot underdel. And that's good because that means you can ship stuff that's not perfect. Yeah. Because because that's by definition how we ship stuff here. We ship stuff that works. It's not perfect, but we ship it anyway. And yeah, people will write to you that

01:04:35

it it sucks here and you know there's this bug over there and this is not clickable and stuff, but that's exactly fine. We we do it this way here. Um and and that actually makes things uh better on this delivery side, right? Because in fact, delivering stuff is incredibly incredibly rewarding. Um but once you deliver before you deliver, it's actually super scary cuz like, oh, what if they find something that doesn't work? You know, like even if you publish a podcast, right? You start the podcast

01:05:06

and and the hardest thing is to publish the episode. Yeah. any anybody listening to this can start the podcast tomorrow and you don't. Why? Well, because it's actually super scary because you'll publish an episode and there will be like four people listening. And if you think about it, it's a super irrational decision because well, first of all, four people is higher than zero. So, somebody's listening and it's your first episode. After all, if more than the number of guests on the podcast, then I

01:05:29

think it's a success. Yeah, success. True. But like, you know, first of all, like it's your first podcast anyway. and and and u you don't expect to have viral popularity from the day one. Uh probably and that's not what you fear actually. You just have this irrational fear of starting. Yeah. Of just shipping something that's not perfect that I don't know I will look not look good. I'll say something stupid and there will be one random Joe 66 that will comment on it and say that I'm stupid. Right.

01:05:55

And it's fully rational and it prevents you from shipping stuff. And so if you have the culture that's actually honest about errors on one hand side and on the other hand side allows people and pushes people to to ship things so that others test them before they're perfect then you can like this is actually hard but I believe that you can have this feedback cycle of very very close to this lean startup type of uh cycle where you work with the customer and the customer tells you how to improve your MVP but you can have it

01:06:27

for any type of job within your organization because in At the end of the day, you have the end customer within organization or outside of it. And if you propagate this culture of trial and error, being honest about what's what went wrong and correcting for it as you iterate, then the whole thing becomes very rewarding for people because people learn that, you know, shipping stuff is rewarding. I think the worst part, another bad thing about perfectionism is that you spend way more time on things that probably are not

01:06:54

that important and then you ship them and you still got four listens. Yeah. So coming back to this podcast example like you know one person just took up took out their iPhone and just like sat with a bunch of friends in a bar and just recorded their chitter chat and then published it there and got four listens. You bought like the whole setup for the studio. You made the list of the guests you want to have. you reached out to three of them that are just perfect guests and like one of them agreed and

01:07:20

then you did I don't know whatever hired a professional plus production blah blah blah and then you got the same four listens cuz it's your first goddamn episode what did you expect? Uh you know and that's actually despite the fact that you ship stuff and you ship stuff that's far better than the first option. This might actually be depressing. Whereas with the first option, woo, we just recorded something on iPhone in a pub and four people listen to this Wow, that's pretty cool. How about next

01:07:45

time we record it in a quiet room without a background noise? Yeah. So by shipping thing that that's undercooked, you also give yourself space to develop and develop in a direction based on the feedback from others. Not within your like head of like you know not sitting in ivory tower and anticipating what people want but actually listening to them. I will send the transcription of your speech to chat gibbery and prompt it to cheer me up anytime I um I'm afraid to start something or I can

01:08:16

tolerate my failures. the yeah quad word in I think the biggest insight that is pretty well known is that most people don't care for you right and you know that deep inside you know most people don't most people don't care for you like most people like like most people don't care for you like you have care about then I thought they should have care about me like other things to care about why is that like if you're lucky people care like people from your family care for you at

01:08:46

fast, but most people not a very good thing because well, you still care what your parents think of your job and your whatever whatever. But like it's it's at least something, right? So, they actually would actually bother to provide you with feedback and like if you answered to them, they would probably answer back and blah blah blah blah blah. Like most people, even even if they have the the the the motivation, which is actually very hard, and they write a comment below your YouTube video

01:09:14

about, I don't know, how your makeup sucks, that would be the end of it, right? Even if you return to them and explain how this makeup is actually perfect for the situation, probably they will never respond because, you know, it's not what they do in life. They just trash talk the people under some YouTube random videos. Yeah. Uh and and I think that's pretty liberating in a sense that like most of the things you can do them and you'll be fine. Yeah. You know this is what I tell to my students like

01:09:41

students come to me for career advice and I just tell them look first and foremost you'll probably be fine. Yeah. So like don't do hard drugs and work and you'll be fine. Yeah. You live in Germany for God's sake. You know unless there's this world not being as dangerous as it used to be 200. Exactly. Right. Like you know there there's no random knight going rumaging through your town and burning people for no reason. You'll be okay pro usually. Yeah. So, I mean, we need we

01:10:10

need to care for certain things like we don't want to have global third world war or stuff like that, but as long as we stay on track, you'll be fine. And that that's that's actually an assumption that's actually very hard to believe in for younger people. That makes them very riskaverse. And in the meantime, if you just read a little bit of history, you'll find out that we live most of the people on on the planet live in the most uh risk controlled environment ever. Yeah. So there is this

01:10:36

fantastic study from Norway when um a psychologist noticed that that there were safe playgrounds for kids with like a lot of this rubbery stuff that it's very hard to hurt yourself. And kids were way way way crazier on this playgrounds than on the normal metal playgrounds. And they actually studied it and their hypothesis is that the kid learns by exposing themselves to failure which we talked earlier today about. Yeah. And if you make environment safer for the kid, intuitively a small kid actually tries to do even riskier stuff

01:11:12

in this environment because they don't feel that they perform on the on the verge of their cap capabilities, you know. Yeah. So we do have this internal drive for exploration and by nature we nature kind of forces us to be risktakers at least in the earlier stage and then through education through certain structure of life we get this assumption that risk-taking is not necessarily good. It doesn't pay off. By the way it's not true in the modern world we see that a lot of the very successful people are actually very good

01:11:41

at risk-taking. Yeah. And the school most of the time tries to actually enforce the idea of that risk takingaking is not good. So to a large extent it's a constructed belief. It's not rooted in reality at least I believe. And then if you look at the biography you know like my favorite well I have several several heroes like you know I think that um I read the biography of a person I'm like that's amazing that that's that's the person I want to be like when I grow up. And then

01:12:10

one of the examples I I can list multiple ones but like one of them would be um Schlowski. So Victor Schllovski the the founder of structuralist approach to literature. And when I say structuralist approach to literature, most of our listeners just yawn immediately because you know um and so they think I know it's fine and and you think about a philology professor but then you read the biography of this guy and you know the guy was actually involved in the terror group of uh social revolutionary party of imperial

01:12:40

Russia helped to organize several terrorist attacks on some Russian officials. Then during the first world war he became the commander of a of of a of a tangle battalion. So he because he he because he could you know apparently um then there was the revolution. So during the revolution the bolshik came to power and social revolutionary party that he was a part of was not in favor so they were jailed. So he spent some time in a desa jail. There is a episode of his biography where he was actually

01:13:08

playing chess with the boalik um overseer and the bat was basically his life. So he wins, he's allowed to run away. If he loses, he'll be shot on on site. So he wins, he runs away, lives a little bit or all over the place, come back to Soviet Union, whatever, you know. And in the meantime, he also writes poetry and some works on structuralist literature. That's the that's the biography of a scientist. That's a biography of philos philology professor 100 years ago, you know. Uh

01:13:35

compare it with with your biography. Um so are you sure you don't have enough risk in your life? Are you sure that your life is super stressful? Are you I mean like another another perfect example would be uh Everest Galwa the famous mathematician who came up with with a concept of certain concepts in mathematics that we only understood why we need them in the 20th century because they turned out to be very useful for aerodynamics when we started to build airplanes. So some of the mathematics he

01:14:04

published and then he died on a duel when he was 20 because he had a fight like some political conflict about France because he was a Republican and and and monarchist kind of didn't like him a lot you know um still became a great uh scientist very well known um if you look I don't know another example that I like to to um uh to Yeah, we're still um despite all these examples, we still suffer from having a toddler, for example, and we think our life is so busy and so hard at this point. So, I think it's

01:14:47

it's a matter of um the year we're living in. So, today, no duels, but other category of problems. Yeah, of course. Of course, we're our well-being is not comparable with what we had 100 years ago. But yeah, you know, I actually wanted to circle back to what you mentioned as a very important skill is being um aware of uh um technological landscape and I actually have no idea how to organize it correctly because one day I just uh uh turn out to Costa and ask him to share with me all like tech resources

01:15:30

he's reading every day so that I am uh uh in course of everything and I got so overwhelmed with information I just muted all these telegram channels and uh unsubscribed from this newsletters um how to keep this balance because I I want to know new things I want to learn um new apart from build an AI bot that will help you it's very easy take course download corser uh and build an AI bot that will help you to monitor telegrams Right. So it's very easy. If you have subscription to OpenAI, what you can do,

01:16:07

you can actually have an Telegram bot that gets all the posts from the Telegram resources that cost a recommended, filters them down according to your interests and sends you a short list of five most useful things once a week. Yeah. And you can build it your on your own and learn a bunch while you do that, right? And you can build it just with Corser. just free version of Corser and OpenAI API and you know just just coding a little bit with Corser maybe asking for help from one of uh dev people and just just just build your own

01:16:38

agent that would basically curate your telegram subscriptions for you and probably it will not be perfectly it will miss some important things and list you some unimportant ones but first of all you learn a lot while you were building that second of all you will not be overwhelmed with information third you'll be super proud that you built this yourself Yeah. So I'm already proud that I'm thinking that I will do it one day. Even this imaginary idea is like like for example like last week we

01:17:08

discussed an idea of an app with my wife and then and then I just you know just build a prototype with it via Corsor. It's actually it took me like one night. Uh maybe if I didn't know programming it took me like it would have taken me three I don't know but it didn't take me that much time and um yeah we kind of testing it now maybe we'll we'll publish it maybe not but my point is that the distance between the idea and prototype never was long never was shorter right it's it's very very short now especially

01:17:36

when talk about digital products so and and then by the way I'm pretty sure that people get overwhelmed with with Telegram channels and you're not the only person and probably there's a product that's doing this aggregation already and you can just pay five bucks subscription and if there's no product and you have this bot that you built maybe you should charge five bucks subscription and here you go you're a founder now you know um that's how the world works now people just don't

01:17:58

understand but that's that's literally how it works you you you explore problems that human have you think how technology can help with those problems you think if you can build something based on technology to solve those problems and if the answer is yes you build it and at least you solve your own problem and at best you become a founder There is a really nice book um by James Wooden I think um so he's uh uh he's one of the partners at Balderton Capital James Weiss sorry James Weiss and and and and

01:18:28

um uh I think the book is called founder century where he basically makes a statement that AI uh specifically generative AI enables solopreneurs in a manner that was never seen before and the percentage of population in the 21st first century that actually becomes uh business owners uh probably will increase drastically uh ju just by the nature of the availability of those tools. Yeah, that's that's a future of work that many people predict like solopreneurs or creative economy as like some some other

01:19:04

um branch perhaps of that. I wanted on that actually I suggest maybe another skill that's I believe is really important and that's proactivity or being self-driven enough so that you can set your own goals decide what you want to do actually do it u which is a skill that I think like is under underestimated and to your point previously modern education system is not always helping uh if not like oh the opposite on the opposite like most of my classes uh the homework homeworks are open question homeworks So I don't tell

01:19:36

humans students what to do. I ask them yeah I ask them what to do. I tell AI what to do. Yeah. So it's it's not accidental. But like with humans I actually leave open sour open questions type of homework. And even like when we talk about master students most of the time master student reach out for me for the topic and I say well I don't have any like because it's it's your master thesis. So, how about you come up with the topic? And that gets them so puzzled because it's literally for some of them

01:20:02

it's like the first interaction in their life when when like the person just tells them, you know, build stuff you want to build and we'll see how it works and and you know, tell me what you want to do and they're like, whoa, I I never heard that question education system. And fundamentally, I think it's a big big problem, right? Like if a smart well-meaning person um hears what do you want to do first time when when when they are already in their masters then you know something's wrong

01:20:32

in the Danish kingdom something's fundamentally like all the previous time they've been told like here's an assignment here's what you need to do way and follow this structure and whatever yeah and and like surprisingly enough I I would First of all like and and here we come again to this question of beliefs. So like I think believing in in the freedom of um of uh uh will is important because if you believe in it then you believe that the natural state of human is actually being proactive.

01:21:01

You fundamentally believe that that's that's what we are born with. And if you look at any kid by the way like you have a toddler you know like toddlers are super agentic like like toddler has like you know like telling toddler what to do I mean like good luck. Um uh you can show you can stop them. They will do it again. They are very like toddlers are very goal oriented. They are very very stubborn in what they want to do. They have zero trouble communicating it to people around them. Failure tolerant and

01:21:33

they're super excellent. They actually enjoy failure quite a while. Like my toddler fall down and smile and laugh about it and try again. Yeah. So I fundamentally believe that this is how we are designed. And then the problem is that if you want to plug this human being into the factory and make them work on a conveyor belt, then all those things are very very very very unhelpful. Yeah. like to work on a conveyor belt all those things actually make things worse and just create a lot of trouble

01:22:06

and you don't know don't want no trouble with people working on conveyor belt and so education system is largely designed or if you send the person to the army as a private all those things for most of the armies uh actually don't work uh famously for example Israeli army is one of the few armies in the world that actually fosters the agency US army to some extent tries to do it and claims If you redesign or think the way home reparations should work based on this agency of humans, then you can actually

01:22:35

be more efficient even at that. But this is a relatively new approach. And then this classical proian army where you just need to stand in line and follow and do this thing and marching in the right direction. All those things are very very annoying for the general. So so we built an education system to actually pervert to some extent human nature. At least if you believe that by nature humans are um you know risk-seeking uh self motivated individuals and so we perverted the incentives and we kind of beat into the

01:23:08

humans the things that are very very unnatural for us as human beings. Yeah. So like we are hunter gatherers you know hunter gatherer is a very like to be hunter together you need to be super agentic you know you need to find out what you want to gather and you need to find out what you want to hunt and if you're a gather you on your own or maybe with your family and then you bring stuff back to the tribe and you need to be good enough so that people actually take you back and stuff and like with

01:23:34

the hunters the same thing you need to construct your own tools maybe you have a few people together uh who you can you know convince that you are a good company to hunt a mammoth together, but it's a risky business to hunt a mammoth. So, you actually need to be good enough. You have something showing for you so the people follow. Um, so like biologically we don't have this problem. Moreover, by the way, if if you ever saw how kids play computer games, you can see that like computer games hijack this

01:24:04

internal uh mechanism in a in a fantastic fashion, right? So like good computer games are designed in such a way that that humans are very very self motivated to do stuff. Um no problem at all. Yeah. Uh and like people I I I always say that arguably uh Sid Mayors and his civilization did more for historical education globally than any other teacher of history, right? Any book on history. Yeah. Minecraft probably taught more about certain chemical and physical properties of the world than any book on physics or

01:24:37

chemistry. uh because like humans internally are very explorative by nature and we actually spend a lot of money to make our kids less explorative and even less explorative when they're taking a higher education. So I think that that definitely should change. The good thing though is that technology is already changing that we see the examples of the schools like Costa mentioned. We see um people in higher education discussing those things openly and trying to go against the current. We see some attempts to redesign education

01:25:07

from scratch which I think is not a bad idea again because the costs of redesign are actually drastically lower due to genai. And by the way coming back to this question about like automation of professors, AI professors and so on. The generative AI allows you to experiment with a lot of educational products uh on your own or with a very small team of founders significantly reducing the cost of uh you know production of different educational materials. um and maybe thinking a little bit out of the

01:25:38

box and you know how they should look. I can give you ana an example of a course that's taught here in TWWS. So we have this internal type of Oscar type of prize for the best course every year. So in autumn every year the professor who was chosen to to to have the best course is presented with certain trophy and one of the winners. So I was really impressed with it in the first year. Uh in the first year of the course that got uh um the the best uh um um price uh was the course taught by by the faculty of of engineering and the

01:26:13

name of the course was natural rubbers. Yeah. And it sounds like rubbers. Yeah. Natural rubbers. And it sounds like a crazy crazy name for the best course. But then when you look into how the course was structured, well, apparently the people were split into teams that were rubber producers and they got the the technical specification of a rubber they need to produce and a a possibility to participate in the auction for a bigger corporation that's looking for a rubber supplier and they have a bunch of

01:26:41

books and materials that they can study and prepare their uh application for the auction with the price estimates, with the production time estimates and so on and so forth. And the whole semester the course is structured in such a way that the teams progressively get hotter and harder tasks where more and more factors of rubber manufacturing should be taken into consideration in order to actually win the auction. And the professor is just the guy who kind of guides them through some educational materials that

01:27:10

are available. He's around for the whole class while they're working on this thing available for their questions. And he's the person who's making judgment on the final bids for the auction and choosing the the teams that actually wins the auction. That's it. That's brilliant. And you know what? Like people also learn social skills the best by playing and by interacting together. So they they not only taught them about rubber production or whatever, but also I'm assuming there was a lot of their

01:27:34

internal debate and I know group dynamics and that they learned a lot from each other in terms of those. Yeah. Another example, I think there was this uh front-end school. I think it was called El Bruce in in St. Petersburg and what they were offering they were offering an express course on like front-end development uh and they were boarding school so for grown-ups so people were paying to come and stay I think for 3 weeks uh in the in the dormatory and like every day you have eight hours of working as a front-end

01:28:04

developer from day one you are added to the slack chat you have the manager this person is not called your teacher the person is called manager so you you're you have a manager you have your slack group you has other you have other channels you have the tools like standard tools the people you have Jira and and stuff like that so in your first day of the course you get access to all the tools that you're supposed to to get you get your manager you have your team with which you work on certain projects

01:28:32

and then you just have standups and you have reviews and you have a performance review in the end of the week and then in three weeks you are a junior front-end developer not only because you actually shipped I don't know end projects of different front ends in a very intensive manner because you are in the office for eight hours and then offline you have like if you're running on the deadline and like you cannot meet the deadline you work on it on your own in the dormatory with other people who

01:29:00

also try to catch up the deadlines but also um they had this practice they had a bell in the middle of the room in the middle of the open office and whenever somebody found a job they were ringing the bell so everybody knew that like oh yet another because in the middle of the course they had this career mentorship block with HR that explained them how to structure their CV and how to apply for new jobs. Yeah. And and and and so starting from the second week, people already apply for the jobs based on some

01:29:27

portfolio they built in the first week which is crazy timeline but that's working. This is the model that actually example such a fantastic example. Yeah. of action. The Oscar for the most positive optimistic attitude towards technology goes to Ivan. Thank you. Thank you. I would like to Yeah, I think it's been we discussed lots of things already and that was a fantastic conversation. Thank you so much for your time. Thanks a lot. Thanks for having me. Sorry for talking so much. I guess no, I'll share my bot for

01:30:03

uh exploring the tech channels with you uh in a few weeks. Nice. Awesome. Yeah. All right. Thanks everyone for tuning in and listening. If you like this episode, please like, repost, share, leave a comment. You know what to do. Uh we'll be very very thankful for all things like that. Uh and until next time, [Applause] [Music]