Şefik Özcan on the AI bubble, regtech and what gets cheapened

We accidentally went live again. Is it live or something, bro? Exactly. Do you know what happened this time ? This time, it's not LinkedIn or YouTube, Riverside has launched something new, I guess. Exactly, man, I really don't know what we're going to do with technology . I've become like a tech boomer. I really feel that way. So the rate at which people become boomers will be accelerated significantly earlier.

For example, let's say someone in their 50s right now is in "boomer mode," but here you can be a boomer at 35. I have a feeling a feature like that is going to be activated. Bro, I'm definitely a boomer right now. So, the number of AI boomers in particular has increased significantly. I think I'm pretty good at using chat, GPT, and Cemin, in my own way. I met a kid in Kadıköy the other day .

Programmer. He showed me some models like this. Video creation stuff. I was shocked, they're doing such crazy things. Yes, brother. So, I think they'll say something like, you know how back then people used to make fun of people who would type google.com into the search bar first, go to Google, and then search? I think people will soon be making fun of the prompts and things like that they write.

I have a feeling there's going to be a whole bunch of "Are you a boomer? Is that how you write prompts?" kind of stuff. We had a course called Object-Oriented Programming. In software engineering, for example, the professor might display a code written for an assignment on the screen and say, "Hey guys, don't write like the person who wrote this ." He was shaming us all by saying that if we did that, we would use up all the system power and freeze the computer .

I have a feeling those shaming attacks are going to intensify even more. I'm not even including those things . So, these NATs and all that stuff are getting out of hand . So, on your side, Berkay, who's involved in analytics, shares things occasionally, I see it. are there any MCPs or anything like that over there ? Do you see the sale? So, right now we're dividing the AI side of the product into two parts.

First, to be honest, our prediction is that most tools that offer human-centered solutions are facing a very big risk . This includes those in Turkey as well as those abroad. So, if you're presenting something human, the biggest risk you're facing is what Open A might say. So, when you say "Hey, now you can connect your Google Ads and Google Analytics accounts to GPT chat," all the human tools you offer suddenly feel like they're falling into a void.

That's why we try not to get too involved in that person's side of things. What we're trying to do is take your data from Hardal and send that data from Hardal wherever you want. We are now working on an additional agent internally as well . That agent will do something too. So, especially for those with large data pools, there will essentially be an agent in between to be able to retrieve the report or dataset they want .

The interface will be a little easier to use. Someone might come along and say... So, basically, when I say I need data broken down by source medium like which products were purchased in the last 7 days and which were added to the cart, we're working on a system where it will directly give me the dataset . So we try to stay on the data side as much as possible.

Well, if we divide the process into three parts, it's collecting the data, extracting the data, and then extracting meaning from it. The riskiest aspect when dealing with AI is the process of making sense of things. The others definitely involve a manual workload. Oh, and we're very cowardly too, by the way. well, that's a bit interesting... the dependency thing is really messing us up there. Like what ?

So, listen, the problem you're describing is actually exactly the same. So you're making an investment, you're getting the thing working. You are running the process. If tomorrow GPT chat says "I'm shutting down this API" or "these queries will no longer work" or starts doing something on its own, then the business model is in serious trouble. So, that dependency issue is one of the things investors ask about, and in the server-side, if Google were to bring it in and make it a more basic interface, a simpler format, maybe your business model, the pivot you did, was a direct risk.

The more dependencies you create, the more problems. I've seen so many companies fail because of this, and we're very wary of that. By the way, they did something about it, 100%. Let me also touch on the agent issue . There, we are directly creating a data vulnerability that can feed the MCP servers of these companies, rather than just through the agent. So, within that context, for example, why do people use MCP?

Since they don't want to transfer their own first-party data (meaning data within the company) to the cloud, and because the data at Hardal is also within the company, we're creating a connector with that. So basically, there's no agent or Insinight or anything working there at all. So, even if the dependency changes, the system they're running locally remains the same. Meanwhile, Google has already simplified these server-side measurement and other systems.

So, he went and made a deal with CloudFlayer. Google launched this tech, launched that, and so on, but in our case, it's the complete opposite. So we're comfortable because we've removed these dependencies from the core of the product . I wish it would get easier so people wouldn't have to ask us for SGTM anymore and could just use Hardal's normal, basic features . Right now, it's reached a point where we don't want to sell server-side solutions or be server-side wrappers anymore.

We have currently developed a mechanism that can create a first-party stack using only hardwood. That's what makes us feel so comfortable. We don't even want to get involved in those aspects of sales, but the AI side is just hype, a bubble . By the way, if the sound is good, since we're already hyped up, we can jump straight into this topic for the broadcast. What do you say?

I think it's quite good. Okay then. Super. Okay, let 's start right away. I'll start then. Greetings everyone. Welcome to the new episode of our podcast . Our guest this week is Şefik Özcan from Efili. Greetings, Şefik. How are you? Hi Berkay. Hello Barış. I'm perfectly fine. Everyone is in good spirits. How are you all? Oh, that's great! We're doing just fine too. Barış, how are you?

I'm doing great too, honestly. I'm in a good mood . We started talking like this before even starting the podcast. Because of his enthusiasm, these kinds of issues are already piling up. I'm eagerly waiting. Honestly, I forgot; I just saved it so we could do a test . It was actually 78 minutes later We can publish the things we talked about in secret. So if they do something like that, if they write to us, maybe we can send those people this recording, like that behind-the-scenes thing, you know, on YouTube and stuff.

When people click the join button, they show them other content or something. Maybe we can do that on LinkedIn too. It could be . Well, anyway, I guess my joke wasn't very well received. He does that. It would also be helpful for MRI scans. It also contributes to income. YouTube, games, etc. And now we have to deal with that too . This is how we issue invoices, right?

With paperwork and all that. And Barış is in Istanbul, by the way. We'll meet with him too. Finally, we'll also have a team meeting this week. Well, Şefik abi, before we even started talking, we were already discussing, I mean, before the broadcast began, both artificial intelligence and business logic. Well, I could actually just pass the ball to you right there . So, actually, let me ask you the question you asked us.

So, what are your thoughts at this stage? So, regarding artificial intelligence, do you have a strategy in place right now on your side at Efilli? Are you taking any action regarding it? Or how is artificial intelligence used within the team ? How do you see this side? We'd like to get that very short thing there. We would like to receive their information. Of course. So right now there's a huge amount of hype in the world, as we just talked about.

And with all this hype, a lot of work is being done related to artificial intelligence. You even had a meme the other day. So, how did we secure that $20 million investment ? Look, the presentation right now says AI... Really, that's how the process is. But we've seen this before. We saw it in Metavors as well . We've seen this in different kinds of crypto processes, including NFTs.

Well, with all this hype, the vast majority of the bubble is bursting. Perhaps 5%, perhaps 10%, perhaps 1% survive. Right now, everyone is trying to throw a stone into the well, hoping to hit one of them. Well, it will transform things. EA will definitely be a new world, a new internet. In fact, one of the topics discussed, especially in Silicon Valley, by futurists is this: In the future, jobs will be divided into two categories.

One, tasks ahead of EI, two, tasks behind EI. The jobs that AI will have in the future will be people who take commands from AI and execute those tasks. Secondly, there will be people who code give prompts, or do EA engineering. Well, it could be any sector here. Regardless of the sector, there's a prediction that job descriptions will be divided into two categories. I believe that too.

Well, because it's started to affect our lives . Many professions are now becoming distractions. Well, you know, a few years ago, when the pandemic started, there were fewer programmers in the software industry. Demand for digitalization suddenly skyrocketed. Software developers have become incredibly valuable. Subsequently, this demand decreased. The number of software developers has increased. Now AI has arrived. Also, the effort and productivity of a programmer can be increased by working with a team of 3, 4, or 5 people instead of 10, plus tools like Curser, RLI, or COP .

At some point, you might start replacing the programmer. At some point, you might start replacing the architect. Maybe, maybe at some point you'll start replacing the doctor. Aayla. Well, now this is a reality. Well, there's no way this won't change the media, digital marketing, and digital analytics industries. Well , good, different, and useful things will come out of this. But from what I see right now, the AI transformation needs to be planned very carefully and with a long-term perspective.

When we talk about this in the workplace, we always focus on the concept of being future-proof, meaning being able to do something that will survive in the future. Right now, we're thinking about transitioning core systems to an AI infrastructure immediately . Rather than focusing on becoming an AI native company, we're focusing on what we can do in the future so that AI won't disrupt it, or what we shouldn't do so that AI will disrupt and destroy it.

Since we are currently in a more remote area outside of our core services, namely the rework technology field, it is naturally a somewhat delicate process. We can't trust blindly. We are trying to develop some of our complementary services using AI . Again, our software developer friends are accelerating AI development even more. Yes, we also use AI in our marketing and other processes, but this time we are an AI native company.

We're flying like this, we're running away like that. We created our own language model. I can't just say, "So we took this from somewhere and created GPT like this ." Well, I think we have a little more time for that. But things will change, they will evolve. Well, we need to be ready for this. So, regarding what we talked about before about being a boomer, we'll try to capture that as much as possible.

We'll try not to be too much of a boomer when it comes to that . Either I'll wait there for a short while. I think that's a very logical approach. The system we follow is very similar to this. Well , because there are two points. In today's world, it's very difficult to make long-term plans. For example, I have some friends who have children in elementary school. Now, when we send those children to school, one of the added values of schools is this.

We are teaching software development to the child. They start learning software development as early as the second grade, and so on. Now, for example, those families are in the "is it really necessary for our child to learn software development ?" mode. Because the child is only 7 years old. And perhaps in 10 years, there won't be such a thing as software, or coding, anymore. Well, it will be a very different world.

So why did it get to the point where we're teaching coding to kids in the 7th grade? You could actually take this and apply the same logic to companies in certain areas as well . I don't remember where I heard it recently, or who said it, but there was some talk about it. So, elephants are dancing. All we can do is try to guess where their next steps will be and not stop there.

Because it's very easy to be crushed right now. And as you mentioned, we are an AI native company in every way. This is how we get the letter We're acquiring an AI. We're putting it in another EAI. The kind of thing where we package it with EA and sell it to another EA company... well, that could disappear at any moment. They could be crushed under those huge elephants at any moment .

So, even though everything is changing so fast, I think moving forward with sure steps is just as valuable. Because, for example, there's something like this... Now, I don't know what your AI stack is? Which AI tools do you use in your daily life ? But nobody, for example, if we only consider elm platforms, uses Perplexity, chat GPT, Cloud, and Gemina all at the same time. Everyone has a preference.

Because everyone already has the ability to use an AI tool. Therefore, when you go and create your own GPT, it really needs to offer significant added value for that person to choose it . Or perhaps people are starting to shy away from it too, saying "oh, another AI tool" or "oh, another AI solution." So, how do you maintain that added value, and as you said, how do you strengthen your complementary services with small AI tools?

I think that will have a bigger impact on long-term survival . I don't know, Berkay, do you have a different perspective ? I think the same way. So, rather than us knowing that this is hype, we're not denying it in any way. But what we haven't reflected, as the CEO also mentioned, is how the company uses EA, how it implements it in its own product or service, how it speeds up its processes, and how it actually generates profit.

So, if you're doing a job at a rate of 2 units and AI can reduce it to 1 unit, then yes, that's added value. But when it comes to the hype, the way everything is labeled as AI, the way every project is actually about AI, or something like that, and how it's constantly shoved in our faces, I feel like it's actually killing the added value of the project.

Otherwise, while it might be easy to do things or join a system, I argue that it shouldn't be just about packaging something and using an API, ignoring the underlying mathematical operations and models. While you were talking, I was looking at something else. according to I Combinator, in 2025, what percentage of companies will be AI? So, around 90% AI. So right now there's nothing in Combinator besides AI .

That scares me a little, to be honest. I feel like it's somehow diminishing the importance of the matter . Yes. That situation has frankly turned into something else entirely. Well, it's back in the startup business too . Whibinator was a very different model, in my opinion. So, I think we need to evaluate that place from a slightly different perspective now. I'm also curious, by the way, Şefik, after hearing your comment, how has this broadcast positioned itself within your industry?

Because we talk a lot about personal data and privacy. AI always seems to me like it's on the other side of the coin, especially in the current process. I'd like to hear your opinion on both of these matters as well. Of course . So, we believe that in our field, which is the relatively new digitalization of legal and regulatory technologies—what we call Recttech and Legaletech—it would be healthier to use AI more for auditing or monitoring purposes rather than directly applying it to the application itself .

Because when you overlook something in a project, which is very likely in large structures, it's a challenge to do so . So there are a lot of units. Too many units are implementing too many things, and sometimes these things go beyond the scope of regulation . It could be a policy, an information statement, a cookie management system, or something else. There are millions of regulations now. So, even in our country, almost 1000 regulations have been created.

There are different regulations on the health side. There are different regulations on the food side. There are other regulations on the energy side. There are different regulations in the telecommunications sector . Well, using this to find vulnerabilities in companies seems a bit more logical to us here . Fixing this vulnerability, while I think it makes more sense for it to still be a human task, is something we believe would be easier to do manually rather than having people manually scan thousands of things they might overlook.

So, that's why we also use AI in the auditing processes we've developed internally. Again, before classifications appear to the human eye, they are initially classified by an AI. Then the human eye looks at it and corrects it. This is, of course, a topic that contributes to e-learning as well. Because in its next classification, it can make sense of it. Of course, it's a process that varies greatly from case to case, so doing this is quite difficult, but what we're trying to benefit from here is the audit processes, and along with these audit processes, there are things that companies need to pay attention to when manually reviewing thousands, perhaps thousands, of pages, documents, and contracts.

I can say that this location is a little easier to implement for us, less risky, and... comfortable in terms of operation. But that will definitely change in the next phase. This has started to change significantly, especially at LegalTech . yes, and its negative effects are also being seen. So, in a document presented to a court in America, there are two examples of cases that didn't exist, involving EI hallucinations.

There's an example, and it's a completely messed-up situation. Well, examples like this have started to emerge. So, a defense petition is being written from the AI's perspective. They cite precedents there, but those precedents don't actually exist. They investigate and find that nothing of the sort exists. The AI was hallucinating. So when you say, "We trusted blindly and tried to get an outcome from them," well, if there's a criminal aspect to it, serious problems can arise.

Similarly, in marketing, if someone sees a visual from a poorly performing campaign as if it were performing very well and bids on it, they can ruin the campaign. This can lead to deviations in the marketing budget. So, I think this topic will lag behind a bit in some sectors . Or, again in the health and medical sector, when an analysis is done based on a blood test result, a mistake can even lead to death .

Well, we need to pay attention to these things . It happens to me sometimes too. For example, I'm going in for a health checkup. I'm putting the outputs in Is there anything you 've seen that relates to me? Is there a pattern? I should actually include personal data to see if it's a news reporter's report, but... well, I'm including it purely out of my own personal curiosity .

The results it gives are sometimes very accurate, and sometimes none at all. These are things like, for example, someone says they have insulin resistance . Oh yes, I have insulin resistance. This is something I know. There is a genetic predisposition to insulin in the family . A diabetic is being transported. I'm saying I need to pay attention to this . But it can also pull out some completely unrelated, absurd thing from somewhere else, and it makes me paranoid .

These are matters that require very careful consideration . That's why I never trust someone blindly. Hey how about you do something, like, send your health data, maybe something like, an image, but... can you hear me right now? It's coming, it's coming, I think. Uplon. Heh. Oh, I was going to say... So, what if the AI generates an image after you send it your health information, and then a doctor appears and does this...

What should we do, man? What more can be done about this, and so on. Oh, I have something to say, by the way, regarding the health side of things, I don't want to digress too much, but it seems to me that broadcasting could be one of the most valuable areas. Because there's already something in the healthcare system about a person's body; if you have a heart problem or an issue, you go and usually only your heart is examined.

A full body scan, to see what your other organs are doing, is something a doctor alone would find very difficult to do. That's why it seems like AI could be very valuable, but on the other hand, as you said, it's a very risky area. And that point you just made always makes me think . So, every piece of personal data you upload there, you 're sharing your personal data with a very large system.

Seltman just came out again recently and said something... He said something like, "People don't know that what they say in GPT chat can be used as evidence in court," and so on . I think there's a huge gap on that side of things right now, and nobody seems to realize it. Company data, financial metrics, company secrets, personal secrets— they leak everything like this. What will happen to all that eGPT and stuff in the end?

What will it turn into? I am very curious . So, in the meantime, we cannot and do not upload any data related to the banned company to Chat GPT . that technique is strictly forbidden. Well , because a financial statement or calculation you submit there today might come back to haunt you tomorrow or the day after, from somewhere else. Yes, of course, there's also that other aspect to it.

Now the scope of data collection has expanded greatly. Back then, there was just one computer. We were generating data from there. Now they're everywhere. There are cell phones, there are IoT devices. My smart ring is currently measuring all my metrics and has increased my stress level. The sleep device dropped and it's somehow going to the cloud. Behind the scenes, the company is probably doing a lot of AI analysis related to this, and it's giving me a summary by analyzing my health data.

In a world where so much data is being collected , processing it and deriving an outcome from it is, of course, valuable. This applies to marketing as well. Marketing professionals also collect store data . So, it's collecting data from people who go to the store based on the advertisement. It directly extracts various metrics from digital sources, such as how many times a person shops, the frequency of those purchases, and so on.

They develop statistical models, and all marketing is essentially built upon that . So it's based on what optimization I can do here. It's something completely focused on optimization. AI is actually very valuable in marketing because it reduces the human factor. It's the same for your work; when you put that same effort into it, it's like the cherry on top of the cake. But it needs to be done with a very accurate model and in a way that will produce the right output .

Yes. So, I don't know, on the other hand, there's this thing where we're being tracked from all sides, and then we act like the use of personal data is based on a consensus, but in reality, so much nonceding data is circulating that in the end, how logical it becomes is definitely debatable. And you say, "Well, nobody's tracking us, sir, we have a plan for fortune telling, and we use it."

So we don't include all the financial data like that, but we use it when building things. So, at its simplest, we have a wipe coding strategy. For example, when doing wipe coding, we draw certain boundaries and don't allow it to go outside those boundaries, and then we create a control phase. So, our entire design system and everything is like this, it's all a written rule set. Well, for example, we use it quite well in the coding side .

So, Barış, in which areas do you use your skills? We use it constantly, to be honest. We'll do that soon, I mean, we'll do our own thing. We use cloud computing, by the way. Yes. we use a cursor. Well, frankly, we also play a critical role there, because, fundamentally, we're on the open startup side, and there's a bit of that too. So, most of our information is already publicly available.

Except for certain critical, high-confidentiality cases. That's why we're being a little more flexible. We are in a completely supportive position in this regard. So we're positioning it in areas like an AI that supports the systems within the Hardal, or an AI that makes the Hardal easier to use . Or rather, the product, or rather, the " thing" approach, is definitely not something we have either. We need to make the product AI.

It's not about what to do, but really, Hardal is already a gateway, meaning you buy it, add it, and it makes it easier to collect and manage data . The easier you integrate it , the easier it is to use, the more advantageous it is. And how can AI make this easier? So we're in a place where we generally focus on optimization and simplification. For both internal and external use.

And the aim there, as Berkay said, is to be able to do it within certain rules. implementing and using it within certain rules and so on . Well, you can't establish them today because you don't know what you'll face in the long run, etc. And then there's this aspect of it that we were actually very much against at first. Well, some of us have even posted about it.

This is the point regarding whether to do things in-house versus outsourcing. While AI isn't normally a big fan of in-house production, it has frankly made certain things a little easier to do in-house . So, it made it more possible to have a world where we take certain solutions, customize them for ourselves, and run them on our own servers . That's why it's in so many places. For example, our entire system— pricing, quotation generation, etc.—its basic structure has actually become a new product.

Hardal switch, you know. So, switch to Hardal. This switch Hardal, for example, we actually created together with AI. Then we implemented it , launched it, and now it runs as a separate structure. There are also some complementary products, somewhat similar to yours, that we use for both internal and external use. Because we also have that part there. So, that's good, we need to use it carefully. Because there's a point on our side that heavily touches on privacy as well.

So, even though we tell our clients that they are in control, that they manage which information goes through, which doesn't, which needs to be hashed, and which needs to be redacted. Even though we say the control is in your hands, as the system providers, there are still some precautions we need to take. Here too, how the broadcast is used and implemented is very critical. Because we shouldn't leave any loopholes open just because we're going to add something to something .

So, I can say that we're trying to be sensitive in that situation. It's a bit more about the security side of things. Super. well, the topic there gets a bit... very technical. Let me change it a little now. And most recently, I did something like this using AI. there's something I have. There is a security device. Like an alarm system or something. the battery ran out over there in the garden.

I thought I'd open it and take a look. They connected the batteries in parallel like this. You just connect a bunch of these AA batteries in parallel and that's basically how you power the LEDs on that thing. I also tried to light it up again, using two AA batteries . Now I do these kinds of things all the time with that chat chip pit. So, at home, at work, for do-it-yourself projects, like how to change a plant's soil, how to connect batteries in parallel, or other things like that, I started using it constantly for these kinds of topics and I started writing in Turkish .

Back when it first came out, I was constantly sharing code like this. I was asking you to explain this code . So, with this code, I was saying things like "refract it," and so on. Now, for example, my personal communication with chat, whether through GPT or artificial intelligence, has evolved over time, becoming simpler and more simplified in Turkish. So, is there such a difference for you ?

So, since it first came out, I've honestly started to use it more in ways that would benefit me. For example, Şevik, how do you start using it? Like, that thing, for example, is a nice thing. You know, it's a good thing to be able to add comments, personal health data, and then get an output from it. Are there any other similar examples, etc.? For personal users, I think there have actually been quite a few positive developments with the release of bit 5 .

Well, the outputs given started to change a lot. I started uploading images and asking questions related to those images, and I began receiving very serious, well, positive responses. So, I can say that my usage has become a bit more intensive recently. I'm dealing with the more strategic aspects here . In terms of strategy, for example, when I do market research, I gather certain data from there. Because otherwise I would have to go to 10 different data sources .

Okay, so the market I'm going to, for example, let's say, a Latin American market. I'm looking at the market, I'm looking at the total. Except for one country, Spanish is spoken in all of them, but that country, which is Portuguese-speaking, is Brazil, which accounts for 60% of the total Latin American market. Now, normally when you try to make a decision, there are many strategic factors to consider.

Or how many businesses are there? What regulatory areas should these businesses comply with ? There is a lot of data like this that I need to research . When I write the right prompt, I can get some really good market research results. Of course, I also need to create an electronic proof of this . it's easier when I do my profile externally. I studied these things. This is the result I obtained.

How can we proceed with this? When I went to an expert at this point, they also said, " Yes, these are accurate, these are correct data." That's a valid point, but when you think about it, you also need to consider this. Because you can't do that in an established company there, he says. Things are much, much easier this time . So we can eliminate a lot of that time wasted in between.

That's why, on the strategic side, I try to use it as much as possible. Well, besides that, as Berkay mentioned, it's also useful in many aspects of my daily life. But one of the things I've enjoyed doing lately is this... I have an idea. I want to get an outcome related to this idea . We could even call him a pre-MVP. So I want to make a prototype model before the MVP (Minimum Viable Product) .

So, within a few hours, he can actually produce something decent, a functional UIUX, a piece of software. Well, this is one of the areas I'm enjoying the most right now . So I want to try something. I'd like to get myself a personal note-taker. I want to do something else. click click click click, right away... I can get the build within 2 hours. And this, for example, has become one of the things I enjoy most right now.

I think this is one of the things that will significantly change the startup ecosystem. By the way, well, the product itself is n't that important anymore in the startup world . Wow, you've made a really great product! Is he an MVP? So, is the product completely sustainable? Is the product, in other words, technically sustainable? And also, it looks at marketing and product market fit. on that side, so that's why the speeding up of software and so on, that's the effect.

And on the strategic side, market research is truly a lifesaver. Well, in my personal life too, I honestly sometimes experience things like this... So I'm also wondering if there are times when it kills my creativity. For example, I often use it when I'm cooking . Well, sometimes I don't know what to do and I might send the product list from Puzzle Lab and ask for some ideas, for example .

Well, or for example, when it comes to something I'm going to do for the first time, I'm not entirely sure. I don't often use it like an assistant, sending a picture and asking, "Does this look good, or should I cook it a little more?" No way. If only you'd let me taste it, I mean, if only I'd let you taste it, it would be gone already .

There's nothing else left, is there? I am not sure. Do you understand? It's like something happened, but then I think, "How can I test if it actually happened?" And I've heard this a lot too. Its use has become particularly prevalent recently in automotive- related contexts . So, someone wrote somewhere recently that they couldn't start their car. The reason the car won't start is because... well, one of the brake warning lights or something has blown.

It's some kind of strange relationship, like it wasn't triggered acutely because it exploded or something, and actually, the system recovered when he fixed that light bulb . He says, "If I took this to the service center, who knows what I would have gotten into..." and so on. It's used for very basic car repairs, like oil changes, minor scratches, etc., or for understanding basic problems. I think it's become quite widespread in everyday life.

You know, even for small handyman tasks like how to fix my toilet or change the shower head, it's become quite useful. You 've probably seen some minor things recently. You've probably seen VO. Autonomous driving and driverless madness. Now, I was in CS in Las Vegas last year too. we saw unmanned mining vehicles, and they operate in the dark. unmanned factories, dark factories. Now, at some point, we'll enter processes that are operated solely by EA, perhaps even including repairs and maintenance, and in 5 years, we'll just get in the car, enter our destination, and it will take us to our destination via the most optimized route, interacting with other vehicles in some way, without involving us in any accidents or traffic .

I think we also need to consider the dystopian world that EA created here . Because that's basically the question everyone's talking about right now . Will it take our jobs away from us ? So, for example, we are currently on the administrative side. We have a lot of decisions to make . Some of these decisions are right, and some are wrong. Inevitably, at the end of the day, they'll say, " Yes, indeed."

From a human perspective, we work entirely with the predictions we have in mind. Some experiences we've had in the past... some things come to mind. We do a calculation and make a decision. Yes, let's do this or let's not do that . But there will be an AI working in the background processing a lot of data, and maybe the AI will be doing the decisions for us without us having to make them, and that could take our jobs away from us .

Now, life expectancy is increasing at the point where humankind has arrived. Perhaps EAL will be extended even further. I'm just wearing this hundred-dollar bill, and since my health metrics are being monitored, I 'm speaking purely hypothetically. If I'm going to die of a heart attack, this way I won't die, but... if my lifespan increases while I'm unemployed and starving, another problem arises. Well, begging sounds very interesting to me.

So when we look at what EA has brought and taken away, I have questions like, are we heading towards a less enjoyable and robotic world? And meanwhile, when we were with Barış in California, seeing those Vymos and things like that, when we went to the city center, you see that dystopian world very clearly and you feel it. For example, there are homeless people everywhere, but on billboards there are AIs and all that stuff, and there's so much difference that it's frightening.

Even that is frightening. Meanwhile, I'm perfectly fine with that. I'm perfectly okay with all vehicles becoming driverless. Because that's one of the most fundamental things that will solve the traffic problem . But I agree with the rest very much. So I think San Francisco was a really crazy example from that perspective. So, like Belke said, we're looking at billboards. Billboards... well, someone who isn't in the business would n't understand them.

So it says something like, "designers can ship now," and so on. Someone might look over there and think, "Hey, what does this guy mean?" As he said, on the one hand, the systems operate in a very automated way . You make the purchase, the product arrives, and so on. But on the other hand, you might be about to go into a restaurant when some drunk homeless person comes along and blocks your way, or something like that.

But you just got out of a driverless vehicle and all that. And a lot of things are going to a... a strange place. It's very difficult to predict what will really happen in the long term . There, among those people we always follow, on the basecamp side, DHÇ said something that really stuck in my mind . He says that as humans, we always think that most things will increase linearly.

The best example of this right now is VR glasses . It's been quite a while since VR headsets first came out . That's what people expected. Oh, we'll hold our meetings from here too . Ah, now reality is about to completely disappear. Then, with the pandemic, it reached its peak even more. If things were following a normal, linear path, we would currently be doing everything with VR.

But at this point, very few of us actually have VR at home and use it regularly . That's why it's very difficult to predict that thing at certain points. So AI is very busy right now. And once upon a time, the metaverse was like that. We used to talk about web3, we used to talk about crypto . How much are these a part of our lives ?

Yes. Much more intimate than before . Something is still happening at Metavors. Crypto is still emerging from very unexpected places, and with much higher quality uses. And I honestly think the same thing applies a little bit to as well. So it will affect our lives, it will change things a lot, but only time will tell how much of an impact it will really have and what the consequences will be .

Because nothing increases linearly. At a certain point, things slow down, and then the hype actually starts and... well, it changes. For example, I think unmanned vehicles are a very good example of this. How much will it cost? Where will it be? So, for example, how many years will it take before we see driverless vehicles in every province of Türkiye? It seems to me there's still some time left, but...

Well, I could be wrong. Because, as I said, things are very, very difficult for me to predict. It was one of my new projects. I was doing some kind of waste tracking project related to smart cities. A sensor inside the trash containers. We had designed a circuit with a few sensors using ARD . It measures the fill level of the trash can . It measures the gas content inside and so on.

He's planning routes for garbage trucks. For the overflowing trash cans. When the garbage truck goes around, it doesn't stop at that street for empty bins, and this allows the garbage truck to collect the garbage quickly, saving fuel and eliminating hygiene problems, especially during the hot summer months . So, we're going to start prototyping. We are in contact with the municipalities, but there is a problem. Some places do n't have trash containers.

Because it's a hill and people leave their trash on the ground. Here are some places where underground waste is found. in some places there's above-ground trash, and there are people who sort through the trash and collect recyclable things . Now, when we look at it, Türkiye isn't suitable at all, but Finland is doing it right now. In Finland, there's a lot of research being done on issues like the amount of garbage.

This issue is particularly prevalent in Nordic countries. Because it has a systematic structure. It's well-made and easier to adapt. I completely agree with what you said. So, well, the infrastructure and the country itself need to be very ready for it . I was talking to a German friend in an electric car the other day . He said he didn't think it could be implemented in Europe. Because he said, "Think about Italy."

How are you going to install an electric charging station on a street where it's difficult for even one car to pass? the guy can't find a place to park his car there. How is the man going to charge his car and travel from one place to another? Or let's skip that. The station you built, the electrical infrastructure there was built 200 years ago, and that's how the infrastructure was supported by the buildings there 200 years ago .

You can't renovate it either, because we don't have anything like urban transformation, or because the roads aren't constantly being dug up and the infrastructure isn't being changed, so their infrastructure is old. Now, when you put a few charging stations there and people start charging their cars, that's when infrastructure problems start to arise, he said. For example, I had an epiphany there about something I had never thought about before.

So the structure needs to be suitable for this as well. As Türkiye, we are relatively fortunate to be a newly developed country. But in older, more developed countries, there's probably not a very good chance of doing that. Or it will be necessary to develop an alternative technology . Let me put it this way. In Spain, for example in Barcelona, phone reception is poor in certain areas of many houses .

So, there's still no internet or phone signal. From where? Because... well, they don't allow too many base stations to be installed. Nobody wants a base station near their home Secondly, the systems are very old. The average age of the buildings is around 100 years. And they have thick walls, all the buildings are very large and deep. How can you build a smart home when you can't even get internet access to certain parts of the house ?

How can you create a smart system ? There are countries or cities that deal with these things, and the city I'm talking about has a significant population and attracts a large number of tourists . When you go to younger children , when you go to the blind, as you said, it seems like there will definitely be very, very serious problems . Where was that thing? Dubai. Was he in Dubai?

It was somewhere . They are essentially building a box-shaped city in the desert, near the desert edge, far from the city . It's like a huge shopping mall, but it also has social facilities inside. There are apartment buildings here too. Like a long block, a block that stretches for kilometers. Doesn't it look like a line? He was in Arabia. In a place like that, for example, instead of restoring old cities, demolishing buildings and rebuilding them, destroying the old cultural heritage and making it smart, they seem to be moving towards a new kind of urbanization.

But of course, I think, for example, imagine trying to do something like this in Istanbul, in a situation where we don't know what to do in case of natural disasters, fires, etc., everything is crammed together, 20 million people, that's an enormous thing. How are you going to make him intelligent? Which part are you going to modernize? Or where exactly will you find the humanitarian aspect? It's a bit...

Difficult. So, I think the technological aspect of the project can be a bit disrupted, even if it's with a delay. But I think people tend to buy everything that's accessible. I mean, I've said this somewhere else before. So, when iPhones and Androids first came out, people were buying them saying, "Oh, I'm just buying it to check my emails," but people who didn't really need smartphones started buying them, and now almost everyone has a smartphone in their pocket, and it has become a necessity .

So you go into WhatsApp, and if you have Instagram, you go into that too. Or perhaps my mind has become unable to live without my phone . I think that as accessibility increases, whether it's VR, VR or something else, or the internet, I think everyone will become familiar with it . There is also an antitheory to this. Well, in the long run, we're already seeing that social media will become an entertainment for the poor, that the rich will no longer use social media and will live a more privacy-oriented life.

Well, things like that still exist today. There are also those "Rich Kids" type of people trying to show off their wealthy lives, but the main context envisions something more like a technology detox, people who don't use phones but use flip phones, people who aren't involved with social media, and people whose digital banking transactions are handled by assistants. So, I read in theory that a certain segment of the population will try to escape this digitalization and live a more reclusive life in order to preserve the qualities that make them human .

Well, it feels both close and far away to me at the same time. Well, it's about lifestyle habits and being connected, like you go on Instagram, you discover things that are related to what you like. If you like football, you see football; if you like music, you see music; or if you're interested in a different hobby, you see that. So, in terms of whether I want to see this or not, well, seeing it definitely nourishes me, but on the other hand, it's an irrelevant world, ridiculous, TikTok-like.

Barış loves them very much. She's constantly watching hair extension videos and stuff like that on TikTok. Frankly, this is my perspective. Well, Berkay's social media usage is quite low anyway. In terms of sharing... I looked at that thing last time too. The last post I made, from my own Instagram account, was about 2.5 years ago . atali. Now, I'm thinking it would feel weird to post anything now, so I'll probably not share anything for a very long time.

Well, there's something like this on social media too. So, there really is a trend in the world. You know how there's a general healthy living trend, like less alcohol consumption , less partying at night, fewer bars and more places that serve healthy food? I think social media will also be positioned within this trend. So that's why I think the number of people going on social media detoxes and things like that will increase .

Actually, I have a brother who's... well, in his early twenties. For example, he has friends around him who do n't have social media or who use social media very minimally. So that's why these kinds of behaviors exist in the next generation as well. I think that the generation that grew up with social media will definitely have more unconventional people, and I think that will be a large group.

I never turn on social media. As you said, using non-smartphones or more basic-level smartphones— I think things like that will definitely attract a following in the long run . Well, among those who are richer or have a higher income level, there's already a certain behavioral pattern now. Returning to nature, retreating into solitude, living outside the city—it's essentially the digital version of that privacy I mentioned, stepping away from social media, or perhaps developing different platforms .

Therefore, I think we will definitely see such a move. I agree very, very much with your theory. So, people over there are looking for jobs. Also, those who are constantly digital, as if they are so immersed in it, I don't think they need a detox, but I argue that reducing usage, rather than a detox, is a necessity . So, those blue lights, you know, that's why I wear those red glasses .

It's really changed, I mean, my body's balance is shifting now. There's a screen over there, a screen here, and... a screen in the bathroom . You're looking at your phone. So, you check your phone when you go to bed at night. Or maybe I'm just trying to get out of this psychological state by doing all sorts of things now. Well, for example, I use regular traditional sticky notes, like those little Post-it notes with a pen and paper, and I stick them on walls and stuff.

Oh, just using these things concretely, helps me to discern that balance. Well, otherwise, this would be something that would be very difficult to overcome. And if we're experiencing such a severe shortage of this, if we consider it a necessity, just imagine what would happen if the electricity suddenly went out. So those skills of ours are related to whether we are human or not. What we're going to do is a bit...

Well, I really want to prepare for all of this. For example, I'm interested in topics like what to do in an emergency, those kinds of things . Just imagine, what if the power went out right now, and there was no electricity in the world? So what are we going to do? Everything is electric. That's why that thing there scares me so much. Preparing for it is somewhat important.

There's also another point to consider from a philosophical perspective . Well, back then there were printed publications . So you were buying a physical book. So you used to buy musical works on cassette, CD, vinyl record, etc. As things went digital and access became easier, they began to lose their context. Because previously, it was more of a desirable object, something that spread by word of mouth: "This album has come out, we need to listen to it."

You either miss a minute of it or you miss it, or the rerun isn't on, or you have to wait until the next week, or you need to catch the rerun on the weekend . Now that everything is at our fingertips and easily accessible, it loses its value and meaning, and as we consume it quickly, we just want to get the next thing. You move on to the next one and use it up quickly.

That fast camel actually operates on a very similar logic in AI as well . So, okay, we found an AI that makes sound . Very good. this time EI is doing something else, EIay is producing the video , okay, something else, EI that can produce three-dimensional things, holograms, that's something else. We are constantly focused on consumption, in other words. Well, honestly, I don't think that much consumption is very natural in our human instincts either.

I'm going to move now. 1500, 2000 records. There are still the same number of books. Well, half the house is being moved in as books and records. Now, under normal circumstances, with one Kindle, I could buy all the books I wanted. Either on a single hard drive, or you do n't even need a hard drive at all. I could actually be listening to all those records on Spotify.

Exactly. Now, under normal circumstances, it's a very optimal scenario, but when I physically hold it in my hands and listen to that music, it gives me a completely different feeling of pleasure. Or, for example, I'm going to an e-record festival with a friend . We're discussing an album there. It reminds us of our human side for 2-3 hours I feel like digital technology has made me much more lonely.

So I put on my headphones. I 'm going to the Explore page on Spotify. It's a completely different world. Maybe there's something that would greatly improve my musical culture, but it doesn't feel human to me. By the way, as someone who loves plag music very much, I should also mention this. Well, in my opinion, the biggest damage Spotify has done to the music industry, or let 's say digital music in general— I shouldn't single out Spotify alone.

Listening to albums. Well, if you look around now, most people aren't listening to albums. For example, let's say artist X released a new album. Several songs from that album become popular anyway. Well, there are a lot of people who know those popular songs but never listen to the rest . So, for example, who do you like? Tarkan. Oh, have you listened to that album of yours? Yes.

I know two of his songs . There's a point at that level. But listening to the entire album, from beginning to end, which is something that can also be done on Spotify, is not preferred. Well, I think that means showing a little respect for that artist and their work. There is actually a composition that he is trying to create there . You're taking this from beginning to end, and...

The topic shifts a bit towards music, but there's a point to be made. Well, when you were listening to music on a plug-in or CD, back in the day, your main function was to listen to music. So you were sitting there listening to music, say, with headphones on. Okay, you used to walk a lot, but now the music has just become background music. With the digital age, it's become a world where music plays in the background while you're doing something, and in that case, you're not really listening to music anymore .

So it becomes background noise, and as you said, it actually minimizes the value of something that is one of the most fundamental things a person can do – like enjoying art, listening to music, watching a movie, and so on . Because all of this becomes a background complication when you're struggling to multitask. Well, that definitely reduces value. There was that thing a while ago, I forgot the name of the pasta company .

Baliet? Baren? Barill, for example, if you want your pasta to be cooked al dente, you cook it for about 67 minutes. For example, he would give you a 67-minute playlist . When you finish that playlist, you add a song that's, say, 2 minutes long. That was the playlist, for example. When you listen to the Aldante 3 song playlist, your pasta is ready. They had created something like that psychologically there.

From a marketing perspective, it was a good thing. But I definitely agree with the multitasking part as well. So we're not advocating for that particular type of music, or for it to be this way or that way. So, you're just going to sit down and listen to the music . Rather than the idea of doing nothing while listening to music, as the conductor mentioned, physically taking the record and going to that place actually enhances the enjoyment of listening to it .

Or the setup you've created while sitting and listening, the headphones you're using—whether they're lossless, HiFi, flag format, etc.—all of these things, even though you could have been aware of them, actually made things worse. I mean, even if people were to listen, they don't even have the patience to listen anymore. Forget about them not listening because of multitasking . People ca n't even tolerate four minutes of music.

It might seem complicated, you know, like the constant use of minor and major scales. Just letting the music play is enough for me. It's always that rap music and stuff. Anyway, let's not get into that now. With songs that are so simple, we're heading straight for the boomer side right now . Brother, last year, I swear, Şefik is also involved with music, and when he came to see me, we had a lot of guitars and stuff like that.

Brother, I took a taxi the other day. Full sound, I guess. The taxi driver had AirPods, those in-ear headphones, but he couldn't hear the music from the car stereo. He turned the volume up full blast. I couldn't do anything either, I felt really nauseous . I couldn't say it either. Brother, Bim started playing rap music from beginning to end. No, it's not like Ebru Gündeş with a rapper, or Demet Akalın with a rapper.

They've just mixed in old music and stuff, and the lyrics are so... like, I went to the barber, got a shave , did I do something... I swear, I asked about it because of that, and I've lost interest in music. So, Şefik, do you listen to Trap music? Let me bring the topic from here to there . Right now, that's the best connection point. I can't listen.

This is what I'm trying out these days. Well, the style I like is more psychedelic, Middle Eastern psychedelic, and ethnic styles . For example, I've started making songs using AI . I've discovered a few channels like that on YouTube as well. They do it too, and I really like that. Oh, make some music. Because I have something in my head . I mean, I'm thinking, what if Deep Purple and Cem Karaca did a JEM session, with Cem Karaca doing vocals, and them playing music and all that?

Now, when you give 'e' the right prompt, and so on, it can produce really great music. So, I'm having her write something else in a song lyric. I'm printing from AI. Such interesting things started to emerge. There are also some very interesting works on YouTube if you search for "Turkish Psychodelic AI" . For example, they covered Haydar Haydar in a very interesting way . There were always these things swirling around in my head.

So, what if we were to turn this folk song into something like this ? I'm thinking, what if we set up a simple, normal drum kit next to Elğaman, and there was a heavy drumming setup, a flow of music, etc., how would that sound? It gives really strange results. Well, maybe in another 5 years, more than half of the music produced will be about family. So, for example, in Turkish rap, there used to be an instrumentalist , a vocalist, and various other things.

Now there's something called autotune . So you don't need to read from that sheet music . Ototune brings you to that note. Or, in the background music, samples are generally used in hip-hop. You take a specific part of a song. You could write a drama about it. Just copy and paste those samples. Let it fall into a certain flow. It's not easy to create samples, by the way, but then you write a hip-hop rhyme over it and perform it.

Now, at this new stage, everything is just copy-paste. Copy and paste the beats. You're looking at the Rys. Ryel's have become completely like pop music rhymes. So, you know, with all that " cistak cistak barking" stuff, a strange kind of R culture has formed. Now you're looking at this... You're looking at American hip-hop of the 90s . This is what Black people experience: oppression, social injustice, gentrification, and so on.

When you look at Turkish rap, you see more underground lyrics, like Ceza's songs, Sagopa Kim's songs have a slightly more literary language. Now you look at where we are. There is a degenerate form of hip-hop music both in America and in Türkiye. If this trend continues and degenerates, and if this degenerate music finds a market, then in 5 years AI distrap will probably become a music economy .

So, who will own the copyright this time, for example? Will AI receive royalties for this? Will Chat GPT receive royalties for this? He'll probably claim that he produced this song and perhaps even obtain the copyright to it. So we're in a period that's ripe for some very interesting begging. He was barking, wasn't he? Chit keeps barking like he's going to throw a towel or something . He'll say, "I'm not producing it for you anymore."

Honestly, I think incredible things are going to happen over there. Even Spotify is heading in a very interesting direction. So, those streams are fake. the music that was made is fake. It goes something like, "There's an economy going on behind the scenes, but there's no human factor involved," and so on . If we start talking about this topic, I think we'll have at least half an hour.

Since we're almost an hour in, I think now is actually a good time to थोड़ा pull ourselves together . Well said, Şefik. It was a really good session. You called me here to discuss business. That's very philosophical, honestly. I also don't understand why we might not have been able to discuss any of the topics we wanted to talk about . I'd say we really need to make a second one .

Honestly, I think it turned out much better. Thanks a lot, Şefik. Wow, great job, it was a really enjoyable conversation. I think we started off a bit more technically, but I wanted us to talk about more everyday things . So, I don't want it to feel like a typical webinar, but I think if you'd like, we could do a second one and talk about other topics again.

So, within the context of the topic we just discussed, since we've taken on the entire Turkish rap community, or if we talk about artificial intelligence or other topics that could lead to online harassment, well, at least there might be an audience for it. If there 's anything left, then there's nothing left. I would like to emphasize this point right away. So, it's not like we're excluding anyone who doesn't use any particular musical genre or anything like that, or who is n't connected to these topics .

We would like to emphasize this. We are only able to see the difference very easily because we are from the generation of three people who witnessed this transition from the past I would like to add a footnote for the new generation. For example, that "haf haf haf" sound I mentioned earlier, in the example the Chef gave—it was n't meant to be sarcastic. We want to emphasize that we tend to lose sight of the meaning behind this.

So, even people in this generation might not know that the barking there actually means, let them talk nonsense . We're in such a disorganized state now. So even our perception can be that insignificant. So, showing what has replaced that rebellion, that anarchism, that kind of hip-hop and rap from the 90s is pretty important. It existed in the 90s too, by the way. Woe to the people of Who at the dogs out.

Exactly. Heh, Dogy Dog, exactly. Doggy Dog. It's actually Snoop Dogg's thing, but I'm absolutely sure that when the new generation listens to them, they think it's like some guy made a song about barking. Perhaps some people don't understand it that way, but I think that's exactly what we want to emphasize here . Just leave the things I threw at you so I can't lynch you . He's extra scared because he lives in Kadıköy.

Kadıköy rappers, Kadıköy. Exactly. Oh, this place is so scary. I'm really scared . They're going to corner us one evening. Okay, let me say one more thing very briefly . The anecdote is... well, I wo n't mention names. well , a fairly well-known gentleman lives in my apartment building. So, a sports commentator. Well, honestly, every time he goes out, people keep making fun of him behind his back.

I mean, it's not just teasing, they're constantly verbally abusing the guy . Ah, so this isn't a step that's being taken? no, but he's a sports commentator, and quite popular on forums and online dictionaries, you know . So, while I don't really know where these kinds of topics are going, I just wanted to say something like that. Anyway, there's not much to it. Becoming famous in Türkiye isn't easy.

So, Şefik, you mentioned moving in the meantime. So, where are you moving to? I'm in the Kadıköy area again. Getting too far away is great. I'm around here. Yes. Great, excellent. Okay then, honestly, let's make sure we do this physically all the time . Thank you so much for joining again! It was a great episode. It was quite enjoyable. Thank you. Well, I think it was nice to talk about so many philosophical topics outside of work .

Well, it was enjoyable. We'll cover different topics, different things, different headings. Exactly. Super. Honestly, the music topic went well. I think you already have extra work and projects, too. I know it exists. we talk about those things too. Good things will happen. So, Barış, would you like to say one last thing ? Closing. Brother, I also thank you. Okay, take care of yourselves. I say, until we meet again, and goodbye .

Goodbye then. I wish you good luck. See you.