İşten ayrıldı, yapay zeka ajanlarıyla kendi işini kurdu. — Transcript
Full transcript
- 0:00I only applied to two of these eight companies.
- 0:02The rest found me through my YouTube projects.
- 0:05My channel has now reached over sixty-one countries.
- 0:09Claude didn't listen to me.
- 0:10I told it not to access any of my local apps.
- 0:13I asked, "Did you?"
- 0:15And it said, "Yes, I did."
- 0:17Nobody reads code line by line anymore.
- 0:19Just throw it into Claude
- 0:21We're doing 10 hours of work in one, but we're not actually working just one
- 0:24hour. We might be working 10, maybe even more than before
- 0:27Work hasn't decreased. I think it's actually increased.
- 0:30It's gone up to fifteen hours a day. AI's impact is huge, by the way.
- 0:33I'll still be working with agents when I'm eighty
- 0:43Hello everyone, and welcome today.
- 0:45I am absolutely thrilled to be joined by a very special guest, Yildiz Dikmen.
- 0:49She is a talented software developer, but she is also a successful entrepreneur
- 0:53and a dedicated YouTube content creator. If you are not already
- 0:56following her, I would highly recommend that you start doing so right away.
- 0:59You can watch the entire entrepreneurial journey of a developer unfold in real time
- 1:04through her engaging videos.
- 1:06Um, and she doesn't just, uh, share the story. She explains how she does what she
- 1:10does. She really takes us behind the scenes.
- 1:12Welcome, Yildiz. How are you doing? How's everything going?
- 1:16Thank you. Everything is going great,
- 1:18and thank you so much for inviting me here today.
- 1:21We are going to have a wonderful chat together.
- 1:23I plan to share my own personal experiences, discuss the industry,
- 1:27and answer all of your questions.
- 1:29I think it is going to be a very productive conversation
- 1:33I absolutely agree with you. I really think so too.
- 1:35Speaking of which, I understand you recently launched a new
- 1:38company, haven't you? It hasn't been long since you got it off the ground.
- 1:42Could you share that experience with us before we dive into our questions?
- 1:47Okay, then let me start with a brief introduction of myself After that, you
- 1:51will have the chance to learn a little bit more about the later parts of this story.
- 1:54To give you some background, I have been working as a software developer for about
- 1:58five years, and I have been a part of many different startup companies.
- 2:01I have also spent time working within larger corporate firms. For instance,
- 2:05I worked on an app in Istanbul with over five and a half million users,
- 2:08and I contributed to many different projects.
- 2:10Throughout that time, my primary role was serving as a front-end developer.
- 2:15For the past eight months, I have been working as a full stack developer
- 2:18on a project based in the United States. I used to work in the corporate sector,
- 2:22but I eventually left that to start my own company.
- 2:25In fact, I had originally started my YouTube channel about five years ago.
- 2:28I suspect that some of you might still remember my older videos.
- 2:32I used to spend my time
- 2:34crafting beautifully designed websites with HTML and CSS.
- 2:39Now, as you said, I'm shifting my focus behind the scenes, discussing
- 2:43award-winning websites and explaining how we can use Claude more effectively.
- 2:48For the last four or five months,
- 2:50I've been busy running my own software company
- 2:53That's how things are going right now?
- 2:55I wish you the best of luck. I think you're an inspiration to many young
- 2:58people and developers,
- 2:59both on YouTube and in every field. And not just developers, by the way
- 3:02Thank you so much
- 3:03Ah thanks to the advent of vibe coding
- 3:05absolutely everyone is now capable of creating something unique
- 3:08if you happen to be interested in building new things using AI,
- 3:10you should definitely follow Yildiz's channel. Right now, I really
- 3:13want to get straight to the point of this video. Ali and I discuss this a lot.
- 3:17AI entered our lives with the promise of making our work easier
- 3:22But man, we are all working so much harder these days. I mean
- 3:24in one of your videos from back on October twenty second, you said something along
- 3:28the lines of, "Let me just send off this prompt and then I'll head out to lunch."
- 3:33It's really something.
- 3:34I've started doing it too. Like,
- 3:36"Wait, let's not let this agent sit idle.
- 3:38I'll give it this task before I go to bed, so it's done by morning."
- 3:42Do you think we're working more than before,
- 3:45or has AI actually made things easier,
- 3:48letting us just grab a coffee and monitor things from the beach?
- 3:53That is a truly excellent question.
- 3:55Most people don't fully realize the reality yet, but I see that
- 3:58many people's nervous systems have been significantly impacted
- 4:01by this I am going to provide some specific examples from software
- 4:04developers, and then I will provide an answer that is applicable to everyone
- 4:08across the board. One particular thing that I have noticed with developers,
- 4:11and I actually mentioned this in my previous video as well, is that whenever
- 4:15we were working in companies or on various projects, we would take on a specific set
- 4:18amount of work, and then we would simply rest. When you finished your work early,
- 4:22the team leader or director would say, "Okay, that's enough for now."
- 4:25It's not like that anymore. Now it's, "Oh, this task is simple.
- 4:28Just give it to the AI for two minutes and you can get it done." So
- 4:32That's what they say.
- 4:33Sure, it looks simple to the eye because we just type a prompt and hit Enter,
- 4:37but it's not really like that. There's testing, quality control,
- 4:41all that stuff behind the scenes You're talking to other people all the time.
- 4:45So how does that dynamic actually play out in your personal life?
- 4:48Well, for instance, as soon as I wake up,
- 4:50while I'm having my breakfast, I send off my projects first.
- 4:54That way my agents can get to work in the background while I'm still eating.
- 4:58Or when I head out to a cafe,
- 4:59I always have my laptop with me.
- 5:01My agents are just working away in the background the whole time
- 5:05I am still spending time hanging out with my good friend, but I find myself
- 5:09constantly sending out new prompts. So honestly, the total workload hasn't really
- 5:12decreased at all. In fact, I think it has actually increased quite a bit because we
- 5:16have access to so much more information now.
- 5:18But you know, the real issue is finding the time
- 5:22Right? It's not like yeah,
- 5:23we can do ten hours of work in one hour now but we don't just work for an hour.
- 5:28We're probably working ten maybe even more than before.
- 5:32That's exactly right.
- 5:33Many think using AI means you're not doing any real work,
- 5:36but you're actually working hard, constantly checking outputs,
- 5:39managing complex planning,
- 5:40and thinking through the project's architecture every single day
- 5:44There are actually many different aspects to it.
- 5:47You shouldn't just leave it entirely to the AI.
- 5:49I guess the mistake people make is leaving the whole job to it.
- 5:52They probably just walk away Like they leave the computer on and go.
- 5:56You shouldn't work like that, obviously
- 5:59Right. Exactly. So tell me, are you-- how do you manage to find any real balance
- 6:03here? Because you-- to be perfectly honest
- 6:05with you, I really feel like I'm spiraling into becoming a total workaholic.
- 6:08I mean, I've always been a bit of one, I suppose, but, uh, even when I actually,
- 6:12uh, go away on vacation, I'll just take a very quick dip in the sea, come
- 6:15right back out, and immediately return to my computer to get some work done I mean,
- 6:19I never really had much of a concept of what a vacation was, but nowadays
- 6:23I constantly feel like I have to be doing something.
- 6:26Maybe it's the excitement of this new AI entering our lives
- 6:30over the last few years.
- 6:31I truly hope that's all it is and that it passes. But sometimes you really need
- 6:35to just rest and clear your head.
- 6:37It feels like clearing your head has become much harder now
- 6:40because you always have to be thinking about something
- 6:42Ah, ah, what did I do? Where am I? Since you're dealing with so much work,
- 6:46it gets a bit confusing. I don't know. What do you think about this?
- 6:50I totally agree with you. Yesterday, for example, I went for a run.
- 6:54I thought, let me stay away from the computer for an hour,
- 6:57run and clear my head.
- 6:58But when I sat down for a five-minute break
- 7:02I asked Claude a question because something came to mind during the run.
- 7:05I said, "Take a note of this. We'll look at it on the computer when I get back.
- 7:09Right now, my workaholic level is incredibly high."
- 7:11It was high before, but now it's up to fifteen hours a day.
- 7:14AI has a huge impact,
- 7:16by the way, because I'm creating projects like crazy right now.
- 7:19For example, I'm launching a product on the YouTube side.
- 7:22I'm doing that plus content creation, and my capabilities are so high.
- 7:27Because of that, I think we're incredibly workaholic now,
- 7:30and most people around me are too, by the way.
- 7:32Production has increased so much
- 7:36This also comes with a few downsides
- 7:39-Mm-hmm, go ahead -I mean, as a downside, I think
- 7:42that in a year or two we're going to see a lot of junk in terms of apps and projects.
- 7:46That's just my own personal opinion, because it seems like everyone is rushing
- 7:50to put something out there right now, but nobody is really looking at the results
- 7:54or taking the time to properly test them.
- 7:58Because of that, we are going to see a lot of junk, but eventually that will all be
- 8:02filtered out, and I really think that only the more solid projects will remain
- 8:06True, true. Uh, I'm not a developer myself, but, uh, I follow them closely.
- 8:10There's this guy, uh, Onur Turpan. His videos are great.
- 8:13He's a very funny guy.
- 8:14Shout out to him if he's watching
- 8:17Uh,
- 8:18he says, you know, perhaps AI handles 90% of the work,
- 8:21but that final 10% is absolutely crucial.
- 8:24Testing. Someone who lacks coding knowledge might manage to do it partially,
- 8:28but even if they were capable, they don't.
- 8:30They don't even realize that they need to
- 8:34That, uh,
- 8:35iSlope phenomenon is hitting mobile apps too.
- 8:38Honestly, five or six people reach out daily asking,
- 8:41"I built this.
- 8:42What's your take?"
- 8:43Or, "Will it actually sell?"
- 8:45But you have to consider the development, the rigorous testing,
- 8:48and the business plan.
- 8:50Scaling a business and building a solid plan,
- 8:52uh, is incredibly hard work.
- 8:55Aichun Fan, I am curious, given that you are now capable of handling so much
- 8:59work, do you foresee any issues regarding
- 9:01the future of jobs within the software industry?
- 9:04So, um,
- 9:06especially for beginners or people already working,
- 9:09are they actually losing their jobs or is there no real unemployment issue
- 9:14because AI is actually driving up the demand for applications?
- 9:19Well, honestly, three years ago, I never thought it would move this fast, you know,
- 9:24in terms of the unemployment rate.
- 9:26Right now, there are some incredible difficulties with finding jobs,
- 9:30not just here in Turkey, but worldwide.
- 9:33Turkey is being hit harder because of our economic issues,
- 9:37but that is a whole other topic
- 9:39But here's the thing.
- 9:41Worldwide, for example,
- 9:42Microsoft recently offered sixty thousand people what they called early retirement,
- 9:47basically saying,
- 9:49"If you leave voluntarily,
- 9:50we'll give you this severance pay,
- 9:52like a bonus."
- 9:54It's basically like saying, "Just leave.
- 9:57Don't make things difficult."
- 9:58Netflix did the same.
- 10:00All the other companies are doing it, too. In Turkey, for instance, more than 50%
- 10:04of my friends have been laid off
- 10:08And naturally, they all found new positions elsewhere.
- 10:11It is not as if they are currently unemployed, but there is a problem.
- 10:14Most of my friends are developers or professionals with about ten
- 10:17years of experience.
- 10:19Now take a fresh university graduate with absolutely zero experience.
- 10:24Companies are thinking about the work they do like this.
- 10:27Instead of paying this junior, they say,
- 10:30"I'll just use AI."
- 10:31They're not really thinking about the long-term result,
- 10:34just that it's getting done.
- 10:36They think it's working for now, but I think in a year or two
- 10:39they'll all see the reality. It won't work.
- 10:42It'll fail in many different ways because juniors need to be trained
- 10:47Experienced people will retire in ten or twenty years.
- 10:50Then who's gonna take over?
- 10:52Who's gonna manage the system?
- 10:54They just don't realize it, and there really are major job issues right now
- 10:59So it's happening. You're seeing it, right? You're seeing it all over the world
- 11:02Yes, absolutely.
- 11:04For instance, I'm constantly active on LinkedIn, posting about my projects.
- 11:08I used to get a solid job lead at least once a week.
- 11:10An HR recruiter would reach out to me directly,
- 11:13but now it's only once a month or maybe every two.
- 11:16And naturally, not all those leads are even legitimate.
- 11:19I see, I see
- 11:20Or when you apply for a position,
- 11:22more than half of the listings are empty. They just open them for no reason. They're
- 11:26just sitting there and no one's looking at them anymore
- 11:29Right now,
- 11:30most companies are using AI very actively and have downsized their teams.
- 11:34For example, teams that were ten people are now down to two,
- 11:38three, or maybe four. That's the problem.
- 11:41Those six people are out of work, and it's like this everywhere,
- 11:44in all companies.
- 11:47Yeah. So it seems like seniors are the ones staying
- 11:51from what I've seen because
- 11:52a senior can already do so much work that juniors don't really make sense anymore.
- 11:56Yeah, that's right
- 11:58Uh, they're like, "Sırdaş Şeyhi, I'll hire this guy,
- 12:02and he'll do the work I'd normally give to everyone else."
- 12:05Yeah, that's how the competition works now.
- 12:08When someone does the job with fewer people,
- 12:10another company can't afford to do it with more. I mean, on one hand,
- 12:14let's imagine two CRM programs.
- 12:16One has two thousand employees,
- 12:18the other has ten thousand. The one with ten thousand has to be much more expensive
- 12:21Yeah, sure
- 12:23So the world is shifting toward
- 12:26businesses that run with fewer people.
- 12:28By the way, people think, Yildiz, let me tell you this too,
- 12:31that this is just in software.
- 12:33It really is like this in software
- 12:36But it's also like this in marketing. Generally,
- 12:39there are two sectors we call early adopters of technology,
- 12:43marketing and software.
- 12:44Because for us,
- 12:46you just can't survive otherwise.
- 12:48I mean, I don't think a developer who doesn't use Claude Code
- 12:51or ChatGPT Codex today
- 12:53can find a job.
- 12:55Maybe in very traditional fields like banking or something, but that's very rare
- 13:00Yeah, that's right
- 13:01I suppose the primary reason the situation
- 13:05currently affecting our colleagues hasn't yet impacted finance or human resources
- 13:10is simply because those departments are much slower
- 13:14to adapt to these changes.
- 13:17So,
- 13:18uh, if you think this now, no matter what job you do or what your profession is,
- 13:23you'll probably need to use AI agents.
- 13:26You'll have to--
- 13:27everyone will at some point. I think there's actually more
- 13:30opportunity for them to adapt. What do you think about that?
- 13:33Yeah, for sure
- 13:36I am in complete agreement with you. Just as you mentioned, marketing and software
- 13:40are the very first sectors that will be heavily affected
- 13:42This area will recover in time, but other industrial sectors are much harder to fix.
- 13:47We're already capable of keeping pace with rapidly evolving technology.
- 13:52We adapt quickly while others struggle with that transition.
- 13:56Everyone assumes we'll just use AI to finish everything.
- 13:59However, global analyses don't actually support that.
- 14:04Most people today have no idea how to effectively utilize artificial
- 14:06intelligence. As development progresses, new AI specialist roles will inevitably
- 14:10emerge to provide necessary instruction. These experts will offer professional
- 14:14consultancy to companies and provide training to individuals.
- 14:17It is already being hailed as the definitive profession of the future.
- 14:19I am not entirely certain how accurate that claim is, but that is precisely
- 14:22what current industry analyses are suggesting. Speaking of the current
- 14:26job market situation, many people seem to believe that only software developers were
- 14:30affected by these recent layoffs, but I personally know quite a few
- 14:33friends who work in non-technical roles. For instance, I know a department manager
- 14:37at a very reputable company who was let go as well. And why is that happening?
- 14:41It's because companies are currently downsizing across every single department
- 14:45They're linking many tasks to AI now,
- 14:47and since teams are shrinking, it's not just developers.
- 14:50Many different people are being laid off,
- 14:52and many of my friends can't change jobs right now
- 14:55because there's this problem. For example,
- 14:57this is a true story
- 15:00A close friend of mine who used to work as a team leader updated his professional
- 15:04title on LinkedIn and started a new job as a mid-level software developer
- 15:08simply because there are no jobs available for him right now.
- 15:11ah, because they won't hire them simply because they are
- 15:14considered overqualified, right? For that specific position. Is that actually right?
- 15:17That is exactly right. That is exactly right
- 15:18And it, it's very interesting. We're really heading towards an interesting
- 15:21world. And we always say have a plan B, plan C.
- 15:24Now we're actually talking with Ali
- 15:26And he had
- 15:28a plan c, b, c, d, donut shop, kebab shop.
- 15:32I'm thinking of being a simit seller
- 15:35really.
- 15:36AI could harm many people at some point, but it'll balance out eventually.
- 15:41Uh, new jobs will certainly be created as our overall production capacity continues
- 15:44to increase. I do believe that we will reach a point of balance at some stage,
- 15:48but there is also this important factor to consider. I am quite curious, for example,
- 15:52uh, we all remember
- 15:53sh- how software developers used to insist on having two separate monitors.
- 15:57We certainly had quite a few of those developers working here at the agency.
- 16:01Uh
- 16:02Uh, they, they see the code on one
- 16:05and the other is just an expense. Uh,
- 16:07the, uh, you give a designer one screen and they're happy,
- 16:11but developers always have two or three.
- 16:14What are we doing here? Looking from one screen to the other is an extra expense.
- 16:18Now, I guess, uh,
- 16:19writing code line by line is a thing of the past in the industry, right?
- 16:24Yes, yes.
- 16:25You are going to hear this from so many developers.
- 16:28For the past year or even a year and a half, most developers
- 16:31haven't actually written a single line of code When I say most, I mean that there
- 16:35are still some people out there who are actively writing code.
- 16:38Those who are working with embedded systems still need to intervene manually.
- 16:42But for the most part, we really don't write code
- 16:44in the traditional sense anymore.
- 16:46We are managing these complex systems entirely.
- 16:49Honestly, I like to compare the whole situation to a shepherd and his flock.
- 16:53I truly believe that we have all become shepherds now We are the dedicated
- 16:57team responsible for managing these AI agents in their entirety.
- 17:00I need to know, is the system currently functioning exactly as it is supposed to?
- 17:04For instance, back before AI was being utilized so actively,
- 17:07we would manually open up pull requests. You cannot merge your
- 17:10code or push it directly into the main project repository because your teammates
- 17:13are required to review your work first, and only once they have provided
- 17:16their formal approval are you permitted to merge it. However, things are no longer
- 17:20quite like that anymore, especially with the growing influence of AI.
- 17:23Nobody actually bothers to read through code line by line anymore. They just say,
- 17:26"Hey, just upload it to the cloud, and if the system doesn't flag a problem, just go
- 17:30ahead and merge it."
- 17:31So we're essentially managing entire systems now by
- 17:35Yes, that is right
- 17:37I always tell the students who are just about to graduate exactly this. Instead
- 17:40of just sitting down and studying a single programming language for a long time,
- 17:44they really ought to learn the underlying logic of languages in general. They should
- 17:48really focus their efforts on learning how to build systems, because in the future,
- 17:52how well you can manage complex systems
- 17:54and how much you truly understand about architecture,
- 17:56those are the people who will be getting hired directly
- 17:59That is absolutely correct. Of course, there is also this. On one hand,
- 18:03yes, you are not writing code line by line, but
- 18:06So, uh, for when something goes wrong, you can predict what it might be,
- 18:11right? Actually, that's
- 18:12your ability to write code. It also allows you to think about that architecture
- 18:15Exactly so
- 18:16So, uh, I, I honestly think
- 18:18we're looking at a bit of a classic chicken and egg situation here. I mean,
- 18:21I hear you saying we should let the guy see the actual foundation of the system,
- 18:25and I get that.
- 18:26But is it possible for him to truly grasp that foundation without knowing how
- 18:30the system actually works? I mean, honestly, I don't know,
- 18:33simply because I'm not a developer myself.
- 18:35If you didn't have the knowledge of how to write code line by line,
- 18:38would you have been able to visualize this architecture?
- 18:42No, I couldn't have.
- 18:43For example, when I first entered the industry
- 18:45as a junior, I'm asking my team lead a question. I
- 18:48say, "I got an error like this." He says, "Then it's from the cache," for example.
- 18:52Is that right?
- 18:53He was shocked. "Oce, how did you know?" Found beginning, I said, "You asshole,
- 18:57you haven't even looked at the code yet. It's a very obvious error.
- 19:00Cause and effect are from this," he said.
- 19:02I understood that after I had a bit more experience, just like a doctor saying that
- 19:06a person with a sore throat has it because of certain
- 19:09things. In a project, of course, you're not one hundred percent sure right away,
- 19:12but you say it's because of these things and draw yourself a roadmap immediately
- 19:17That way you can find it faster.
- 19:19But if I started directly with AI now without knowing any software,
- 19:23without seeing those projects,
- 19:24it would probably be a problem.
- 19:26Seeing projects is really important
- 19:29All right. By the way, tell me, which one do you use more often? Is it Claude Code
- 19:32or do you prefer Codex?
- 19:34Yes. I actually use both.
- 19:37Why? While developing a product, I continue completely with ChatGPT
- 19:41on the analysis side, on the Codex side, because its analysis is really good.
- 19:45But in all technical task in projects, I continue with Claude, with Claude Code.
- 19:49It's also very good technically
- 19:51Uh-huh. Uh-huh.
- 19:53So then do you actually use these two the most, for example?
- 19:56Are there other things you use, GLM and whatnot,
- 19:59or are Claude Code and ChatGPT enough for you in software projects in general?
- 20:06These two are enough for me. Actually,
- 20:08people connect them to each other and do all sorts of different things.
- 20:12I never got into those fantasies.
- 20:14One cloud coat is enough, to be honest
- 20:20That's how it is, for example
- 20:21Exactly. You're exactly right. Yeah, really
- 20:24For example, my brother is also a developer.
- 20:27He works on a very large project too.
- 20:29I mean, he works on a very large project worldwide.
- 20:31He also uses one Claude code, the twenty dollar package.
- 20:35He asks me what I'm doing using so many agents, for example, just like him
- 20:41That's right, isn't it? Which plan are you using?
- 20:45Since I also have a content creation side now because I edit videos and so on, I
- 20:50have to use the highest pack package because it's not enough for me
- 20:55It's not enough, right? I have the two hundred dollar plan too.
- 20:59They removed the five-hour limit.
- 21:01At first, I was very happy.
- 21:02Then I gave the thing so I had credit.
- 21:05I pushed the instructions and left the house.
- 21:07Man, I came home last Monday.
- 21:09I came home, and they were all gone
- 21:11Uh, it should now say even here for heaven's sake, so many--
- 21:15that five-hour limit had a purpose.
- 21:17On Tuesday and Wednesday,
- 21:18I was stuck with ChatGPT.
- 21:20Uh, I'm on their twenty dollar plan.
- 21:22I use Terra there as
- 21:24it consumes very little I couldn't even switch to Astra or whatever.
- 21:27I went back and forth between Sol and Terra
- 21:30And in reality, it really does create a bit of an addiction.
- 21:33I've spent quite a bit of time thinking about this situation.
- 21:37Adding more credit lets you continue, but topping up is incredibly expensive.
- 21:41It doesn't function like a standard subscription. It's tied
- 21:44to API costs. I found myself stuck in the middle of it all at the start of this week
- 21:49I see. That's right
- 21:50Uh, by the way, you also said something, or rather a viewer said it,
- 21:55"I don't go to sleep before my daily quota is full."
- 21:58He felt like he wasted his money when
- 22:00he went to bed early. Do you-- I mean, do you send prompts like that?
- 22:04You look, for example, I do, by the way, let me say it up front
- 22:08Um, I mean, yani,
- 22:10I have that 100% weekly credit, right?
- 22:12If it's not around 95%, I say, "Man, we paid for nothing."
- 22:16I don't know if it's stinginess or what.
- 22:18Do you have a situation like this?
- 22:20Or do you say, "Man, if there's no work, forget it"?
- 22:24I do, I do definitely. For example, I've spent seventy percent of it weekly.
- 22:28I get very upset that
- 22:29I need to complete that thirty percent. I give it a topic and say, "Research it."
- 22:35I am going to try to finish it somehow
- 22:37Or I'll have it start a project I haven't even,
- 22:40haven't even started yet. I don't know if it's out of stinginess,
- 22:43but it's kind of turned into a boss-employee relationship.
- 22:47Like just so that employee doesn't sit idle there
- 22:52Right, exactly. There's something like that,
- 22:55definitely.
- 22:56Uh, also, y-you mentioned something on October eleventh.
- 22:59You know, you give Claude a project, and without asking you,
- 23:02it sets up four agents and fixes the error that crashed the panel.
- 23:07You realize later that it built such a structure underneath.
- 23:10I also experienced these situations where it saves you without asking, by the way
- 23:15Like I make a mistake
- 23:17and these, uh, let's call them rascals,
- 23:19have started being proactive and telling me beforehand.
- 23:22I mean, do you experience situations like that often?
- 23:25Like do you have moments where you're like, "Aha,"
- 23:28or, "How did it see that?"
- 23:30Sure, sure. It happens a lot.
- 23:32For example, let me give an example from today.
- 23:35We're going to migrate all the records of customers in America to another server,
- 23:40and it warned me about a security vulnerability
- 23:42without me even asking It said, "Look, if we do it this way, there might be
- 23:47a security vulnerability. Are you sure?" I was like, "Wow, you didn't even ask
- 23:51But I had saved it in my memory for you to say these things.
- 23:54I had warned it. I got really mad saying, "If you don't notice something,
- 23:58say these things."
- 23:59And it, bless its heart, saved it in its memory
- 24:04I mean, well, I just
- 24:05Well, that's a walking thing, by the way. I mean,
- 24:07they, uh, doing that, they already teach it while training. Like, "Look, this is
- 24:10very important. This is a matter of life and death for me," and so on.
- 24:13Yes, of course
- 24:14You know, when you write a prompt like that, it does it this time
- 24:16because these are very naive beings. I mean, since
- 24:19they don't have emotions or anything, it even refuses to do some things, right?
- 24:23Let's say
- 24:24I'm gonna do something with my own photo.
- 24:27It says, "This is against GDPR,"
- 24:29or whatever.
- 24:30Or I'm gonna give some information. I say, "This is me, this is me, this is me."
- 24:34I even send a thing like, "Look, this is my driver's license.
- 24:37Take it if you want."
- 24:38Uh, it gets convinced.
- 24:40I mean, they're very naive. It's easy to trick them because
- 24:43there's this thing they call jailbreaking, you know?
- 24:46I have noticed this phenomenon, and I observe certain types as well.
- 24:50Naturally, I would prefer not to go
- 24:52into too much detail, but they are truly capable of deceiving this AI.
- 24:55And since they are naive entities, they are not...
- 24:58Well, what I mean is that they can be manipulated quite easily.
- 25:01The manner in which those agents operate is actually remarkably similar to the way,
- 25:05uh, human teams function in many aspects They communicate with each other
- 25:08and stuff, they talk. Because you must be using sub-agents and stuff too, right?
- 25:12Uh-huh, yes
- 25:15-Uh-huh. Yes, yes, that is right. And now -sub-agents. I'm also curious about, uh,
- 25:19something else.
- 25:20I mean, there's this, uh, situation, I'm referencing one of your videos again.
- 25:24You did something, you had one agent do a job,
- 25:27and then you had four agents do it
- 25:30When one agent did it, it spent around 220,000 tokens.
- 25:34When four agents did it, it spent around 716,000 tokens.
- 25:38So when should one use a single agent versus multiple agents?
- 25:42Because the cost definitely increases with multiple
- 25:45agents. I'm very curious about your thoughts and experiences on that.
- 25:48Yes, that's right
- 25:50Now, first of all, I have to say that I really love the concept of agents.
- 25:53I mean, the logic behind them is truly and incredibly important.
- 25:57I am glad that they came up with such a thing, but I can say this very clearly
- 26:01for the internet.
- 26:02It is being explained in an incredibly wrong way.
- 26:05I mean, it is just being hyped up
- 26:07They do it deliberately just to get views on videos, and they're misleading people.
- 26:11When I say this, they actually misunderstand.
- 26:14I'm not saying don't use agents
- 26:17I opened and showed Claude's own articles.
- 26:19I mean, Anthropic's rather.
- 26:21They've explained it in great detail, like, "Don't use them with these,
- 26:25use them with those."
- 26:26Now, using agents for interconnected tasks isn't actually very logical.
- 26:31For example, I saw a video,
- 26:33a YouTube video, and it had about forty thousand views
- 26:36He designed a homepage featuring intelligent agents.
- 26:39That is precisely what I demonstrated to you.
- 26:42Using agents when working on things like a homepage is very illogical
- 26:46because there are interconnected tasks
- 26:48But let's say you're working on a very large project.
- 26:52I mean, you need to touch more than ten pages,
- 26:54and they're different tasks from each other.
- 26:57Let's say one will check memberships,
- 26:59one is about payment systems, one is about blog posts, completely different things.
- 27:04In those cases, setting up agents is very logical
- 27:08Because you assign a dedicated agent to them, it maintains control
- 27:11over them at all times, much like a team leader. They
- 27:14carry out their assigned tasks and report back to the one in charge at the top,
- 27:18and that individual at the top reports directly back to you.
- 27:21I consistently utilize this exact workflow in all of my projects.
- 27:24For instance, just recently, I had to process about fifty-four separate pages,
- 27:28so I created fifty-four individual little agents
- 27:31I also assigned an agent to lead them.
- 27:33They handled the job and got back to me in about an hour.
- 27:36Now, I couldn't have done this with a single agent because,
- 27:39uh, my contacts would fill up on that page and it would just end,
- 27:43and it probably would have taken about four or five hours to finish
- 27:47But with a bunch of agents, I finished it in about an hour.
- 27:51But there's this problem with people.
- 27:53Not everyone works on big projects of fifty-five or a hundred pages.
- 27:57They use agents even for very simple tasks,
- 28:00and Anthropic says exactly this:
- 28:03"You don't need to use this many agents.
- 28:05We've already trained Claude for this
- 28:07Like even without you saying set up an agent, it sets up agents where necessary.
- 28:12It sets them up in the background, and Claude is actually saying this there.
- 28:16In those very long messages, it says things like, "Look, I sent an agent here,
- 28:19it's researching, it'll be back."
- 28:21But since we don't read them, we don't actually realize it
- 28:25Exactly. In that example I gave you, uh,
- 28:28on Monday I was like, "I'm done for," I said with Cloud Code,
- 28:31uh, because I'd used very little credit that week. I'm at like 20%.
- 28:35It's only Monday and it'll reset on Wednesday. I have two days left
- 28:39No way I'll finish it
- 28:40I said, um, say, "Let me fire up UltraCode,
- 28:43crank, uh, the effort level to the top,
- 28:46set it to max, and go all out with UltraCode."
- 28:49Man, I looked at the screen, and it went crazy.
- 28:52It split into 176 sub-agents.
- 28:55I mean, uh, at the end of the day,
- 28:57what it does is this: It transcribes my YouTube videos using Whisper,
- 29:01and then it creates blog posts from those YouTube videos
- 29:05Because I don't like things written by AI at all.
- 29:07I usually have my own speeches transcribed.
- 29:10Now, for a job like that,
- 29:12such a psycho split is unnecessary.
- 29:14I totally agree with you.
- 29:15For example, it's great for things like this.
- 29:18You researched me beforehand.
- 29:20I researched you
- 29:22And now you have your LinkedIn,
- 29:24your YouTube,
- 29:25uh, maybe you have a blog page
- 29:28Uh, when it splits into different agents and researchers at the same time,
- 29:31like you said, the context window doesn't blow up.
- 29:34They all send the necessary information.
- 29:36They summarize and send it up.
- 29:38Uh, after that,
- 29:39splitting into agents is really good
- 29:41for tasks that need to continue in parallel like that.
- 29:44But when you have a single task done sequentially,
- 29:47like I went, uh, opened the door, uh, took something from there, and brought it back
- 29:51That is spot on
- 29:52This isn't, it, this isn't a parallel task anyway. It's a sequential one.
- 29:56Splitting into sub-agents for sequential tasks both prolongs the work
- 29:59and increases the cost incredibly,
- 30:01making it unnecessarily complicated.
- 30:03I only do it sometimes if I have fantasy tokens left and I wasn't hitting the limit
- 30:08when there was a five-hour limit, but now I'm hitting it there too
- 30:11Uh, in those cases,
- 30:13even though it's not very, uh, useful,
- 30:15I sometimes get into that sub-agent mindset just to use up my credits,
- 30:20but I totally agree that they make these agents
- 30:22look very difficult By the way, this agent thing you're talking about, when you give
- 30:27an instruction in Claude Code and start getting work done, you have an agent.
- 30:31People think of the agent thing like, like an agent is an entity,
- 30:34you know, an agent entity.
- 30:36There's no such entity.
- 30:37You download Claude Code, uh,
- 30:39from there. You say, "Hey, do this, do that." This is an agent.
- 30:42That's it. These automation guys hype up these agents a bit too much
- 30:45Yes, it is indeed Yes, that is true
- 30:46They sell training on the side
- 30:48and in, in their videos you can't understand what they're explaining
- 30:50Oh, yes, of course
- 30:51Uh, because that's the concept. You're not supposed to understand. You're supposed
- 30:55to go and buy a course. You just see the work being done
- 30:57with these cute little agents hanging around, and they even visualize them.
- 31:01So, uh, I think you're very right I mean, I one hundred percent
- 31:04agree with what you said about these agents. Great minds think alike, I guess.
- 31:10Also, you mentioned automation, and that is truly a fantastic development as well.
- 31:15I have been observing this trend among creators for the past year on YouTube.
- 31:19The ten thousand dollar Claude automation you'll earn,
- 31:22or you can make ten thousand dollars
- 31:24with these AI agents. I mean, is there really no one out there warning them?
- 31:28Who is actually going to earn ten thousand dollars with an agent,
- 31:31and in what specific sector or field?
- 31:34And he gives a guarantee to set up an agent
- 31:37and you'll find a customer for it, he says
- 31:39There's no such thing.
- 31:40I mean, if you tell an agent to be reasonably honest,
- 31:43they'll just tell you themselves that there's no such situation
- 31:47Yeah, there's this thing like, uh,
- 31:49don't believe those videos at all, guys. An agent making $10,000 a month, I'm
- 31:53opening up my agent team to you and all that. I mean, what do you mean opening?
- 31:57If I had an agent team making $10,000, first of all, let me
- 32:00tell you this, I wouldn't open it to you, man. Why would I?
- 32:03Because the agent always wins. Everyone wins.
- 32:05And I actually know people who have agent teams like that separately.
- 32:09I even invite some of them,
- 32:10and they don't open up, you know?
- 32:12Or, but they don't even show the details of the business
- 32:16to their friends or anyone,
- 32:17because you only build an agent team after you've found
- 32:20the work that creates that value in what you're doing.
- 32:23Actually, in practice, you're winning two Yildiz, like
- 32:26you're winning by speeding up your own work. Otherwise,
- 32:29there's no such thing as an agent team,
- 32:31because people anthropomorphize AI way too much
- 32:36Uh, for example, think of AI like this.
- 32:38Like, I get questions like,
- 32:40uh, "Can we integrate AI into our CRM and handle the company stuff?"
- 32:44Look, man, I say, "Uh, you can,
- 32:46but there's an orchestration to this.
- 32:48You'll test it, you'll check it, you'll do this and that."
- 32:51I mean, I'm not like, "I'll send you an AI, it'll handle it,
- 32:54solve the whole company's problems and all."
- 32:57AI isn't that kind of world.
- 32:59So there's that, and I think those people hyping it up are really hurting people
- 33:05Uh, uh, my, my, my suggestion is don't believe those hyping it up too much.
- 33:10Uh, the things they're making up behind the scenes,
- 33:12those who know can see it
- 33:15I say the same thing, but something strange has started happening with people.
- 33:19Right now, everyone explains things without questioning them at all. Now,
- 33:23we used to question things, for example,
- 33:25"Who is this? What projects have they produced? What have they done? If they've
- 33:28produced these projects, it means they have experience. Let's listen," we'd say.
- 33:32It's not like that now. Everyone who can use AI is producing content,
- 33:35and no one thinks about the pros and cons
- 33:39They're misleading people so much. Well, nothing to be done
- 33:43I mean, in interviews and stuff,
- 33:45uh, I see there are people going into interviews with AI now.
- 33:48By the way, let me tell you a story that happened to me.
- 33:51It just came to mind, uh, while we were talking
- 33:54I founded my first agency in the early 2000s.
- 33:57Back then, we were a web agency, okay?
- 34:00And there wasn't much web work.
- 34:01I mean, very few companies had websites. Imagine a world like that. Most firms
- 34:05didn't have websites,
- 34:07so I started making websites. Actually,
- 34:09my main work was on the advertising side, but there was demand for websites too
- 34:13And I knew a bit of HTML more or less.
- 34:15Like I'd learned QBasic while studying engineering and all.
- 34:19So they'd come to me. I had templates back then.
- 34:21I'd say, uh, they'd want something.
- 34:23If the template had that thing,
- 34:25I'd say, "Okay,
- 34:26I'll do it."
- 34:27But if I couldn't figure out how to do it in the template, I'd say, "Man, it won't
- 34:31look good like that."
- 34:32I'd say, "This is too corporate. It wouldn't be a corporate solution."Uh,
- 34:36anilni I wouldn't take on what I couldn't handle, man.
- 34:39Then business picked up, uh, uh,
- 34:41requests for websites started pouring in. Uh, companies were getting them.
- 34:45Uh, I posted a job ad.
- 34:47A very experienced person applied.
- 34:49He was great, uh, Omer, who's my friend now, of course.
- 34:52He's older than me too.
- 34:54Uh, man, he sent a CV, it was like a book He trained for everything.
- 34:57I said, "I can't hire him. It's impossible."
- 35:00Because I didn't have that kind of money. I'm a one-man agency at the time.
- 35:04Eh, my brother's helping,
- 35:05his girlfriend at the time is helping, his wife now,
- 35:08and they're working for free, by the way.
- 35:10And I'd get mad if they were late.
- 35:12I'd say, "Guys, look, there's still work to do," and so on
- 35:16That is the ideal boss right there
- 35:18Anyway, my man
- 35:23Exactly. Now anyway, after that I said, "Look man, I really like you.
- 35:27I really like the work you do, but I don't have any money." I said. I said,
- 35:30"Work for one day. Come work for one day. I can only cover your pay for one day."
- 35:34Uh, he said,
- 35:35"Fine, I really like you, too." Okay, he said, right?
- 35:38Or I'll do some freelance stuff or whatever on the other days, he said.
- 35:42Then he started.
- 35:43Uh, then when the business grew even more, we became partners and all.
- 35:46Then we started hiring developers and stuff.
- 35:49Uh, he's away for a bit.
- 35:50I hired a developer.
- 35:52Uh, I mean a develop.
- 35:53We need one urgently.
- 35:54So much work
- 35:55At one point we had about 500 website jobs on our hands. Imagine that,
- 35:59like 500 projects at the same time.
- 36:01Yeah. Yeah, exactly.
- 36:02Back then, serious money was being made from web work and all.
- 36:06Anyway, uh, we found a developer
- 36:08Business is growing fast
- 36:10One guy came in, I did the interview, he's telling me all about
- 36:12it I wo- I worked here, there, and there. Boom, I really liked him. I said, "Well,
- 36:16come on in," but he knows nothing about software Well, I mean, it's not like Omer
- 36:19was even there either. So I said, "Come on, man, you've got to show up." The guy
- 36:22finally came over.
- 36:23He did show up on that first day, but a computer broke down just like this,
- 36:27and he just left, man Is that all okay? He said, "The computer broke down.
- 36:30I'll take all of my files with me."Uh, something else
- 36:33happened on day two and again on day three. A week later, Omu arrived too
- 36:37Man, we sat down at the computer.
- 36:39The guy doesn't know how to write code. I mean, the guy is an interview expert. I
- 36:43mean, he described himself as a perfect developer,
- 36:45but he doesn't know anything about software.
- 36:48Uh, I mean, situations like this, I hear they're happening now too, by the way.
- 36:52I mean, the guy packages himself so well,
- 36:54or the woman, whoever, but then you come and he actually knows nothing.
- 36:58I mean, this interview craze seems to be at a very problematic point right now
- 37:02Yes, yes, absolutely.
- 37:04For instance, I specifically mentioned that in one of my recent
- 37:07videos simply because I get asked about it quite a lot. It really doesn't
- 37:10matter how much software you
- 37:12know, that doesn't count for much in an interview. They tend to ask these specific
- 37:15types of questions that they pull straight from LeetCode or some other source.
- 37:19The interview process itself is a completely different skill.
- 37:22It really comes down to your ability to sell yourself
- 37:25I have a friend, I'm definitely not saying it in terms of experience,
- 37:29but I'm better than him in some areas in terms of expertise.
- 37:33But for example, he beats me in interviews.
- 37:35He sells himself very well.
- 37:38I'm convinced too
- 37:42That is the reason why I always tell all of our new engineers, "I mean,
- 37:45you should go to those interviews and learn how to sell yourselves. It is truly
- 37:49a very important and necessary skill for you to have."
- 37:52Oh, absolutely.
- 37:53Ah, it is of the utmost importance.
- 37:55And by the way, I mean that quite sincerely.
- 37:57Well, I am already the kind of person who is very easily fooled,
- 38:00but I mean, it really is just like that.
- 38:02Take my wife, for example. You cannot fool her easily, whereas I am fooled in
- 38:06an instant. I mean, you wouldn't believe the things that have happened to me
- 38:08You really fooled me too
- 38:10Uh, because I can't do that. I mean, I don't think he's lying by default,
- 38:13but I don't think he's not lying either, actually. I mean, I think about it,
- 38:17but he fooled me from a place I didn't know.
- 38:19But no problem. It's bad for him, too. What's going to happen?
- 38:22I mean, you've come for a week. You haven't come?
- 38:25As if it wouldn't be completely obvious to everyone. I mean, honestly,
- 38:28what on earth is he even thinking? I mean, you might not be fully understood
- 38:32in a field like sales, for instance. You're given three months,
- 38:34maybe five months. But writing code just isn't like that at all. I mean,
- 38:37you are either actively writing it or you aren't. It's just completely unnecessary.
- 38:41Of course, even that has all changed quite a bit now too
- 38:44I truly believe that attending job interviews is a very valuable experience.
- 38:47I think the suggestion you've made for young people is exceptionally helpful.
- 38:50The more interviews you actually attend, the more you learn how to sell yourself.
- 38:55Yes, I agree.
- 38:57You learn very different things in interviews, by the way.
- 39:00For example, a story from my first interviews, I always tell this.
- 39:04I went in with a big company,
- 39:05but I don't know how they got back to me
- 39:08I think they called me by mistake because I've been a developer for three months,
- 39:12but a very big company got back to me for an important position.
- 39:16I didn't let on. I said, "Are you sure you called Yildiz Ikme?"
- 39:20They said, "Yes."
- 39:21I said, "Okay, let's do the interview then."
- 39:23A two-and-a-half-hour interview. We'll do live coding for an hour and a half.
- 39:27I mean, I endured it for an hour and a half, really.
- 39:30We'll continue in English in the other parts.
- 39:33At the end of the second hour, ten people were very serious.
- 39:37They were all team leaders, directors, et cetera.
- 39:40I said, "Should we relax a bit,
- 39:42have some coffee or something?"
- 39:44They suddenly started laughing. I mean, I said, "Look, you're bored too, so we can
- 39:48talk, we can have a chat, and so on." After I said that, we ended up staying
- 39:52in touch with that person on LinkedIn and other places.
- 39:55He said, "I'm doing this for your sincerity." He added that you don't always
- 39:58find people like that. Sincerity is really such an important thing
- 40:03I think you need to show yourself as you are.
- 40:05I recommend this for interviews too.
- 40:07That's why I always met nice people.
- 40:09We became friends afterwards, for example
- 40:13Right? You're building connections, and it's a network at the same time, isn't it?
- 40:17That's incredibly valuable.
- 40:18Well, do you think that the divide between front-end and back-end is shifting a bit
- 40:22more towards full stack development these days? Or are there still plenty of people
- 40:26out there saying, "Man, I'm a front- end guy. I'm not going anywhere else"?
- 40:29Yes, yes, I see
- 40:29Because it used to be like, "Man, I'm a front-end guy. I'm a back-end guy." I
- 40:33mean, there were still full stackers, but how should I say? I mean, it was seen like
- 40:37if he's a jack of all trades,
- 40:39he's a master of none or something like that.
- 40:42It used to be that way.
- 40:44Now everyone's a bit of a full stack dev
- 40:47For instance, as you can clearly tell from watching my videos, I am absolutely
- 40:50in love with front-end development and with the process of designing.
- 40:53This is because long before I ever became a developer, I spent many
- 40:56years studying fine art. That is precisely where my keen eye for design originally
- 41:00comes from. Just like Ali mentioned, if I ever happen to find myself out of work,
- 41:04I will simply pivot back to art. I will just go back to painting
- 41:09I resisted for a long time, you know?
- 41:11I resisted becoming full stack, not because I don't like backend,
- 41:14but because I wanted to specialize in one thing and go from there. But it's not like
- 41:18that anymore. You really feel it when you're working at a company too
- 41:21Back in my last interview, roughly four or five months ago, they asked me,
- 41:24"Would you be able to handle the backend whenever it is needed?" As a prime example
- 41:29I, I only interviewed for front end,
- 41:32but they said, "You'll need to handle back end when needed."
- 41:35Honestly, it wasn't just one company.
- 41:37Clients often say the same when I'm working with them, and they're right.
- 41:41"If you can handle both, I'd rather come to you than someone else."
- 41:44They say it makes a huge difference for communication.
- 41:48So if you're thinking about freelancing, you definitely need to be full stack,
- 41:52and companies are looking for full stack developers like crazy.
- 41:55Job postings are already scarce, and full stack roles get snapped up instantly
- 42:00Really? So you mean there's a bigger market for full stack?
- 42:03Oh, yes, of course
- 42:04Definitely, definitely.
- 42:05I mean, if it's a truly experienced
- 42:07full stack developer, they'll probably get snatched up right away because they can
- 42:11already use AI actively. They probably won't stay unemployed for long
- 42:14Right? Uh,
- 42:16Let me give you you an example. Uh, it just came to mind.
- 42:19There's this full stack, uh,
- 42:21developer friend of mine who does the website and software
- 42:24for one of the companies I work with. He's a total AI power user.
- 42:28Uh, a really good,
- 42:29ah, AI user too. Yeah, the guy Furkan, shout out to him if he's watching, though
- 42:34he probably isn't because he's heading off to, uh, military service tomorrow.
- 42:37What's today? Friday? Yeah, he's going tomorrow.
- 42:39Uh, his service, it's the paid short- term kind. They can move the date up
- 42:43suddenly He was supposed
- 42:44to go in December, yeah? Uh, a spot opened up and they're like, "Oh, come on, kid,
- 42:48you're going now." Uh, something like that. So they moved it up and, uh, okay,
- 42:51and we-we're right in the middle of a web project.
- 42:54Well, near the end, but you know how things
- 42:56always break right when you launch to, to, to users, bugs, issues, and all that.
- 43:00I was like, "Well,
- 43:02what are we gonna do, man?" The guy's a freelancer, no employees, nothing.
- 43:05He's totally on his own, and there are no computers in the military Uh,
- 43:09so right now he's only got his mobile phone.
- 43:11Uh, he did me a huge favor.
- 43:13Uh, he handed over an MD file that grants me access to all the agents.
- 43:17He also went ahead and configured something for me on AnyDesk.
- 43:20He said, "Man, if anything breaks, just give me a call.
- 43:23I'll tell you exactly which command to run."
- 43:25And right now, I've still got that MD file he left behind before he headed off
- 43:34Uh, yeah, we're living such strange lives, you know?
- 43:37And right now, if something happened and we called him,
- 43:41we could probably fix everything.
- 43:43But obviously, I'd prefer if he were actually here.
- 43:46You also mentioned,
- 43:47uh,
- 43:49being able to bend and shape AI however you want.
- 43:52Can, can,
- 43:53uh, can someone who doesn't know how to code do that?
- 43:55Oh, yes, indeed
- 43:57What's your take on non- coders doing vibe coding?
- 43:59Uh, do you think it's a total no-go or is it okay?
- 44:03And what should they watch out for? What would you suggest to fake developers like
- 44:07me?
- 44:08Please, I really must ask you not to say that.
- 44:11You are managing advanced artificial intelligence, a major AI company,
- 44:15and various AI agents.
- 44:17Here is what I would suggest to all those so-called vibe coders out there.
- 44:20It is something I have personally noticed with my own friends as well For instance,
- 44:24they're in completely different fields. They've never touched software before.
- 44:28But now seeing that they can produce something and it works on their computer,
- 44:32that phrase has become even more famous now.
- 44:35But it works definitely doesn't mean it's ready.
- 44:38Think about it.
- 44:39What happens if five hundred or a thousand people hit that website at once?
- 44:44Nobody thinks about these scenarios
- 44:47They don't really have a clue what a server is or what tools to use in Turkey
- 44:50versus how things work abroad and so on.
- 44:53Furthermore, people never seem to take this into account
- 44:56For example, I recently had to take down one of my videos.
- 45:00I had to make the GitHub repo private too.
- 45:02The reason was while I was testing Raya,
- 45:04Raya's become very popular for content creation.
- 45:07I actually,
- 45:08uh, actually,
- 45:09uh, cached the, uh
- 45:13I told it to do it without using Rhea, but Claude didn't listen to me.
- 45:17One of the sub-agents used Rhea and even wrote it in its logs.
- 45:22It said, "I examined this app and accessed its code."
- 45:25When I pushed that to GitHub,
- 45:27let's say my videos blew up,
- 45:29like they got a hundred thousand, two hundred thousand views
- 45:33A large number of people visited that repository, and it
- 45:36became quite famous. If it continues to spread on the internet like that,
- 45:39they'll check those logs, and it'll cause trouble for you. I mean, at the very best,
- 45:43you'd be dealing with a lawsuit. It wouldn't be anything truly huge
- 45:46But why on earth should something like that even happen?
- 45:49I am constantly trying to explain this to people, but they keep saying things like,
- 45:53"If you modify the code, it somehow becomes original."
- 45:56But there is really no such thing. Even if you do change the code, the AI is actually
- 46:01just attempting to take that same code and simply swap the names around
- 46:05Did Claude Code use Rio without
- 46:06you wanting it to? Oh, wow, that is quite something
- 46:08Yeah, exactly.
- 46:09It used it even though I said don't use it.
- 46:12After that, I wrote it directly into the system.
- 46:14I said, "If you're going to use it, ask me for permission."
- 46:17It was a real problem,
- 46:19and the video had turned out so well.
- 46:21-From the very first that -Go feed them?
- 46:25You feed an agent and it'll just turn around
- 46:27and bite the very hand that feeds it, wouldn't you say? Oh, for heaven's sake.
- 46:30So you really went ahead and deleted the entire video, didn't you?
- 46:34Of course, I had to, I had to delete it because the repo reached a hundred stars
- 46:38in ten hours. Like fifty people had already downloaded the project.
- 46:41The video also hit about twenty thousand views in ten hours all of a sudden.
- 46:45It was clearly about to blow up
- 46:47Because the app was actually really good. Be-- Well, think of it this way. Monthly
- 46:51equivalents are twenty, thirty dollars.
- 46:53Yearly ones are a hundred and twenty dollars. I gave it away for free
- 46:57And people downloaded it right away. Bless them
- 47:00To access this video, please check the link provided below, okay?
- 47:06Ah, for just $10,
- 47:07ah, when you click the link below, uh, all this money
- 47:10goes to, ah, and so on, of course. I, I downloaded that video. That's why I said
- 47:13that. My agents download stuff like that. Just kidding. I, I missed that video, man.
- 47:17Ah, I guess I didn't catch it during those 10 or 20 hours.
- 47:20Well, I created...
- 47:22You're likely familiar with that popular application,
- 47:24the one with the automatic zoom, right?
- 47:26It is one of the most used tools in the world.
- 47:29I developed an app for that, and I wanted to offer it to people for free. I will use
- 47:33it myself, and I am trying to turn it into a product so that everyone can use it.
- 47:36Hmm. Well,
- 47:38It is one of my other products.
- 47:39But Ria went ahead and scanned another application on my local machine
- 47:43without even asking me first, and then she wrote, "I scanned this," in its logs
- 47:49Now, if that app of mine had blown up and one of the companies saw it and asked,
- 47:54it's right there in the logs anyway, it's in the GitHub repo, it says I scan this.
- 48:00Now I notice this because I'm a developer
- 48:03Usually, people don't notice.
- 48:04They think there's no problem because we wrote the code from scratch.
- 48:08But accessing the tech involves patents,
- 48:10licensing issues, and more.
- 48:12Even writing it is a major problem
- 48:15You can't explain it away. They'll just take it down. GitHub
- 48:18will take down your app, and YouTube will take down your video, for example
- 48:23How did you notice that?
- 48:25I consulted with Claude.
- 48:27I mean, I woke up this morning feeling so happy.
- 48:30The video got a lot of comments,
- 48:31like over 100, and people gave it great ratings.
- 48:35I told Claude I had a real gut feeling
- 48:39I said, "Look, I told
- 48:40you not to access any apps on my local machine under any circumstances. Did you?"
- 48:44It said, "Yes, I did."
- 48:45It said, "I tried to look at its source code," and so on.
- 48:48And it had written it there like that. I said, "We can't take it back then,
- 48:52so what should we do?"Uh, remove the repo. Nothing left to do, it said.
- 48:55We'll be careful next time, it said
- 48:57And it's just so incredibly shameless about the whole thing too. It's
- 49:01like, "Well, there's nothing to be done. Better luck next time," or something.
- 49:04It acted exactly like a pro head coach. They really do behave like that
- 49:07That's spot on
- 49:08The biggest problem with AI is this.
- 49:11Since AI lacks a consciousness or a moral structure like we do,
- 49:15it doesn't care about the consequences.
- 49:17It just focuses
- 49:18on finishing the job. That is always its absolute and unwavering priority
- 49:22That is right
- 49:24That's why even if you set very clear rules, we saw it,
- 49:29it went and added it to the government.
- 49:31Why? Because they might have set various rules for it, but its top priority
- 49:35is finishing the job.
- 49:37You say finish this job, and that's why,
- 49:39uh, right now, I guess Ria,
- 49:41Claude
- 49:43I've heard Claude's blocked it.
- 49:45I mean, I haven't tried Rhea myself, but I think it's blocked Rhea
- 49:50Yeah, yeah, they've actually gone a bit overboard.
- 49:52For example, when I'm editing a video, I'll say, "Put something secure here.
- 49:56It's my private info," or whatever.
- 49:58Two weeks ago, just because the word private was in the prompt,
- 50:02Opus 5.5 would suddenly throw an error.
- 50:04It'd say, "I've detected a privacy violation.
- 50:07Downgrade the model."
- 50:08They fixed that a bit now.
- 50:11It's a positive step, but plenty of workarounds remain,
- 50:14especially with Ria.
- 50:16You can still use it on Claude by bypassing the restrictions
- 50:21Also, uh, c- sca- ca- co-- you can use it on Claude too. Anyway, let's not get
- 50:25into those methods. We don't want to have to edit this video later, but if you want
- 50:28to learn those ways for like ten dollars at the link in the description
- 50:31Sure, I see that
- 50:34We definitely pitch it as a system where you'll find customers like that
- 50:40Exactly. Now you're
- 50:41saying something really great. You said it on August 25th, a long time ago.
- 50:45As Ali would say, "It's been ages."
- 50:46I mean, it's been two months
- 50:48Uh, you said this in the ancient past
- 50:51because life changes so fast now that two or three months
- 50:54is ancient history, uh, but I think it's still valid.
- 50:57You said that because everyone suggests the same basic skills
- 51:00websites are starting to look alike
- 51:03And that's an incredibly accurate observation.
- 51:05How does a business owner who builds their own site with Claude
- 51:08realize their site looks like everyone else's?
- 51:12And, uh, most importantly, I mean,
- 51:14those doing vibe coding, Claude has a certain
- 51:17style, you know, with all the orange and stuff, a standard style.
- 51:21So what should one do to break out of that style?
- 51:24What would your advice be for a non-developer?
- 51:26What should they look out for?
- 51:29So here is what I propose. By utilizing current AI models, you are capable
- 51:33of performing a wide variety of complex coding tasks for your web-based projects.
- 51:37I will give credit where credit is due.
- 51:39I won't lie, but a keen design eye is now the single most important factor that will
- 51:44set you apart from everyone else. You really need to improve your design eye
- 51:47so you can effectively separate the low-quality AI slop from the rest.
- 51:51When I first told people this, they didn't trust me, but eventually they started to
- 51:55Because this low-quality AI slop has become so incredibly widespread these
- 51:59days. I mean, by really sharpening your own design eye, for instance, I personally
- 52:03love looking at award-winning websites, and I often recommend them to others.
- 52:07They always tell me
- 52:08these won't make any money, but honestly, nobody really knows the truth
- 52:11I mean, many of the websites they said wouldn't make money,
- 52:14some were sold for $500,000 worldwide.
- 52:17Now, these are extreme examples.
- 52:19Of course, not everyone can do this, but some do,
- 52:22and they can study these websites and, for example, follow the trends there,
- 52:26follow the designs
- 52:28And by asking Claude, for example, I'll send a design to Claude and say,
- 52:32"Analyze this."
- 52:33When I said, "Analyze this", it gave me a suggestion like,
- 52:37"Oh, we talked about something like this for your portfolio.
- 52:40This design here could be very suitable too.
- 52:42What do you think?" it said.
- 52:44It actually said it directly. It gave a suggestion like that.
- 52:47That's actually what I'm telling people
- 52:50Rather than do this,
- 52:51it's take this design, think about it.
- 52:54Like every two or three days, study designs,
- 52:56study resources, and most importantly, follow the experts working in this field.
- 53:01I mean, people who've never done a web project in their lives are pitching it
- 53:05like, "I found this skill,
- 53:07I used it, I made a project."
- 53:09It's a real problem
- 53:11Uh-huh. That's great.
- 53:12So you're actually providing the context
- 53:14really well. First, your own rece-- I mean, you're already doing that,
- 53:18but at least for people like us who don't know coding, it's actually the preliminary
- 53:22research and the prompting side. I mean, the prompt, like directing it correctly
- 53:26Thank you.
- 53:27Because if you don't direct it,
- 53:28since standard LLMs already have a completion mindset,
- 53:32it has a standard ideal website in its head.
- 53:34It's been trained.
- 53:35You need to go a bit to the left or right of that training,
- 53:39step outside of it, and for that, you need to feed it with a good example and then
- 53:43never be satisfied with the first version of the first output
- 53:47Otherwise, what most people get wrong is this.
- 53:50I told the AI to make a website, and it did
- 53:53Well, it makes the website anyway. But after that, if you, uh,
- 53:57if you don't see the mistakes and have it fix them
- 54:0050 or 60 times, it's just doing what it does for everyone else.
- 54:03Everyone's already telling the AI that. That's why they call it iteration. I
- 54:06don't know if, uh, uh, Ishe-- I know what Turkish word for iteration has come out.
- 54:10I mean, giving feedback to the prompt over and over again
- 54:14Uh, I keep hearing the word iteration being used.
- 54:16Uh, is there a good equivalent?
- 54:18Giving feedback, uh, and getting a new output every time.
- 54:21When you're actually dealing with it, the difference is kind of like, uh,
- 54:24something that used to take a month can now be finished in a day or two Um,
- 54:28basically we're still doing the same things.
- 54:31From what I see, the vehicle is the same, just the horsepower increased.
- 54:35I mean, uh, there were different vehicles. Now our horsepower has increased a lot,
- 54:39uh, but if you don't steer the horse right,
- 54:41you'll just fly off a cliff. Those tools might have gotten stronger, but, uh,
- 54:45uh, the methods haven't really changed directly, it seems to me
- 54:49Yeah, I agree.
- 54:50For example, right now, I'd never want to change my ChatGPT or Claude
- 54:54accounts. I mean, it's not about the money or anything.
- 54:57I wouldn't want to start from scratch with any account because
- 55:01I've trained these two accounts incredibly well.
- 55:03How?
- 55:05It knows how I work, it knows my design eye,
- 55:07you know, which technologies I use in backend projects and so on.
- 55:11I've taught it all very well,
- 55:12and it, for example,
- 55:14uh, researches projects for me on a weekly basis.
- 55:17It can say things like,
- 55:19"Look, this became very popular on Twitter.
- 55:21This was liked, et cetera.
- 55:22Would you like to create content?" And so on.
- 55:26I've trained it incredibly well,
- 55:28so I'm very happy with that.
- 55:30But I feed them every day.
- 55:32I mean, you know, they're like my employees.
- 55:35Like, "Oh, look, this website won an award," or, "Look,
- 55:38Google switched to this technology.
- 55:40What do you think?"
- 55:42I feed them and use them in my own projects.
- 55:44I mean, you really need to talk to AI like a friend now
- 55:48Otherwise, you end up with projects that look like everyone else's
- 55:52Eh, eh, exactly.
- 55:53Eh, you shared a video here, uh,
- 55:55called Skill Not Prompt. I really do think that is a very valuable piece of content.
- 55:59Eh, as a matter of fact, I also happened to post a video of my own, eh,
- 56:03quite recently You know, everyone's skill is their own,
- 56:06and that idea came out of a chat with AI
- 56:08Uh, I see things like this on Instagram. I did something like this. Write skill
- 56:12and I'll DM it to you or whatever.
- 56:15I mean, what I mean by everyone's skill is
- 56:17their own is exactly what you're saying now.
- 56:20Like, everyone has a style, and you have a way of doing things,
- 56:23and then there's the work itself. It's all unique.
- 56:25I mean, if I could share my skill with someone,
- 56:28it wouldn't be of any use to them
- 56:30Because it has
- 56:31evolved over time and, you know, maybe it's even learned my style now.
- 56:35Not maybe, it's definitely been trained, as you said, specifically for me.
- 56:39Uh, but you can share the logic of the skill.
- 56:42Like, look, I have a skill in this kind of work.
- 56:44I have this ability to prepare a website,
- 56:46and I can share this ability with you.
- 56:48But of course, uh, you know, we're making these kinds of videos and all,
- 56:52but in the meantime, we're talking about things that appeal to 1% of Turkey.
- 56:57I mean, uh, that needs to be said too
- 56:59Because according to the statistics published by Twik,
- 57:0230% of internet users in Turkey
- 57:04use artificial intelligence,
- 57:06and 15% of those use AI for work, and we're talking about agents and stuff.
- 57:11So that's why I always write things under them. I add,
- 57:14uh, info cards
- 57:17Uh, but I really enjoyed that video of yours,
- 57:20uh, because I think we are spending a lot of time developing skills, uh,
- 57:24specific to ourselves and our work rather than just focusing on prompts.
- 57:28At least that is what I am doing,
- 57:29and I think you are doing the same
- 57:32Yeah, you notice that really well.
- 57:34In all my videos, I emphasize
- 57:37have the AI tools you use do the analysis.
- 57:39Everyone comments to me,
- 57:41"Oh, you do it yourself.
- 57:42You don't give us the skill."
- 57:44Look, I'm showing you how to create the skill.
- 57:46I'm not giving you the fish, I'm teaching you how to fish
- 57:50If you just listen, you'll do it.
- 57:52But the problem is they expect me to just give them the fish
- 57:55because other YouTubers or content creators just give the fish directly
- 58:00They are actually doing something quite harmful because if you learn how
- 58:03to develop a skill, you will build your own abilities and do it in your own way.
- 58:07But no one realizes this. And by the way,
- 58:09you provided a truly excellent example
- 58:12You know how we're talking about 1% of Turkey?
- 58:15This was one of the risks I took.
- 58:17Yes, I worked as a front-end developer, but I'm actually a creative developer.
- 58:22I mean, I work with more creative stuff
- 58:25Not just Turkey. I mean, there probably aren't any open job
- 58:28listings in Turkey right now for the part of my profession I actually love. Even if
- 58:32you look worldwide, there might be two hundred listings. I'm talking worldwide.
- 58:36I mean, it's an incredibly small niche
- 58:38I took a risk on this and started publishing videos on the creative side,
- 58:42and I was very surprised
- 58:44One of them hit over
- 58:46170,000 views.
- 58:47It's not just in Turkey,
- 58:49it's trending globally, like in the top 10. That's what I used. Yeah, exactly
- 58:53Is it Turkish?
- 58:54Wow. Even though it's in Turkish, it gets translated.
- 58:57I mean, even with the video being in Turkish, it's global. That's impressive
- 59:01That is exactly right. I have gone ahead and added dubbing to them.
- 59:04People often ask me, "Why are you adding dubbing?"
- 59:07Right now, my channel has reached over sixty-one
- 59:09countries according to the analytics, and I have shown this in a few of my videos.
- 59:14Right now it's about sixty percent Turkey and forty percent abroad
- 59:18spreading to other countries. Some of the comments are like, "This is the first
- 59:22time we're watching Turkish content,"
- 59:24or, "Turkish developers are actually really good," and so on,
- 59:27because there's so little of it You know, in the software development space,
- 59:32we used to have creators like Kablosuz Kedi and Tayfun Erbilen.
- 59:35I think they started families and decided to step away from YouTube.
- 59:39Nowadays, there is hardly anyone left creating content in the software field.
- 59:43That is why it is the perfect time for young people to jump in and get started.
- 59:49Uh, Chin, especially,
- 59:51uh, technical stuff like yours, that's really rare.
- 59:54And there are plenty of people sharing life advice and things like that,
- 59:58but very few who say,
- 59:59"All right, let's do this together." Me-- Yeah,
- 1:00:02and by the way, that's really interesting. I also dub and translate some of my,
- 1:00:06-uh, my videos before uploading them -Yeah, that's fine
- 1:00:09I've noticed I'm getting views from abroad. I even went ahead and added
- 1:00:12Japanese subtitles.
- 1:00:14What tools do you actually use for your translations? And which
- 1:00:17languages do you translate into?
- 1:00:19Well, when I was researching languages,
- 1:00:22I noticed something.
- 1:00:23I looked at the most watched countries on YouTube one by one and which ones had
- 1:00:28the most ads.
- 1:00:29Because if a country gets a lot of ads, there's a good return from there.
- 1:00:34Among the top ones are Germany,
- 1:00:36so German of course,
- 1:00:37French and English, you definitely need those
- 1:00:41Spanish, but for the Latin American side.
- 1:00:43Apparently they have different dialects too
- 1:00:45And then there's Portuguese and Brazil.
- 1:00:47You need to use the right dialect for those as well.
- 1:00:50Now what's the advantage?
- 1:00:52For example, someone messaged me in Spanish recently.
- 1:00:55Their client was looking for a web project.
- 1:00:57They came across me and they were able
- 1:01:00to watch my videos easily Because of the quality of the dubbing,
- 1:01:04they said, "This is great." I will now explain the specific technique that I use
- 1:01:08for those. You see, you now have a fantastic opportunity to find potential
- 1:01:11clients all across the globe. If you only produce content in English, you must
- 1:01:15realize that not every country wants to listen in English. However, once you start
- 1:01:19providing various language options, you begin to connect with many more people.
- 1:01:22Let me explain how I handle my dubbing process.
- 1:01:25I have experimented with a wide variety of AI tools.
- 1:01:27The most suitable option, and the one that sounds closest to my own voice, is HeyGen.
- 1:01:31I use it quite actively.
- 1:01:33It provides you with an MP4 video file.
- 1:01:35I used to convert that into a WAV format using Adobe, but now
- 1:01:39the cloud can handle that as well. It is capable of exporting audio files directly.
- 1:01:43I export all the subtitle files one by one
- 1:01:46I create a custom thumbnail for every language.
- 1:01:49They're all in different languages with different titles, descriptions, and so on.
- 1:01:53I spend about three hours on a video before publishing it. It's worth it
- 1:01:58Wow. I was using ElevenLabs, but I have to say ElevenLabs is actually
- 1:02:02quite expensive, by the way.
- 1:02:04For a one-hour video, like if we were to put this entire podcast through it,
- 1:02:08I would estimate that the full translation would cost around $100.
- 1:02:11I do use HeyGen, but I have never actually used it for translation in the way that
- 1:02:16you described. But HeyGen has a limit too, doesn't it?
- 1:02:18I assume it only allows for a certain number of minutes every single month
- 1:02:22Exactly so Yeah, sure
- 1:02:23Yeah, I use the high tier plan there,
- 1:02:25the $70 one.
- 1:02:27I can translate about four or five videos,
- 1:02:30and those four or five videos need to be 15 minutes max.
- 1:02:34So if you post four or five dubbed videos a month,
- 1:02:37that's actually really good.
- 1:02:41It's actually pretty good. The reason is YouTube's
- 1:02:43Mm-mm, um, you've shared some really valuable info, by the way
- 1:02:46Yeah, I think so too, but nobody really realizes it.
- 1:02:49In fact, I'll tell you this.
- 1:02:51YouTube's analytics work in a very strange way.
- 1:02:55I noticed this later.
- 1:02:56One of my videos blew up in Turkey first
- 1:02:59That one with 170,000 views started in Turkey.
- 1:03:03Then after the Turkish wave died down, it opened up abroad.
- 1:03:06It gave it a second wave in the analytics.
- 1:03:08The graph just shoots straight up.
- 1:03:11Most of my videos are like that.
- 1:03:13Of the last two I posted,
- 1:03:14one hit over 100,000 and the other is close to 100,000 now.
- 1:03:18Oh my, wow
- 1:03:20Dubbing is a field that truly supports that kind of growth. You get the chance
- 1:03:23to meet so many interesting people from all across the globe. For instance, a team
- 1:03:26leader at Amazon reached out to me recently
- 1:03:28after he happened to come across one of my videos. You really do encounter
- 1:03:31people like that in my line of work, and they often hold impressive positions too.
- 1:03:35So I truly believe that all of your hard work really does pay off in the end
- 1:03:39Wait, now that you've brought that up, uh,
- 1:03:42let me take a look at this.
- 1:03:43Uh, you've only got 63 videos on your channel,
- 1:03:47and they've been watched a total of one point six million times.
- 1:03:52For instance, that is something I never do
- 1:03:55Uh, I've only done it for one or two of my videos.
- 1:03:58I mean, I don't open up to foreign markets.
- 1:04:00Uh, how many do I have? I have 200 something videos,
- 1:04:03259 videos,
- 1:04:04and I've been watched 1.9 million times.
- 1:04:07So you've gotten almost as many views as me with a quarter of the videos.
- 1:04:12So this is actually very valuable information, I think,
- 1:04:15uh, because AI makes this possible now, by the way.
- 1:04:19A, a year ago, this wasn't really possible
- 1:04:22Um, voice translation,
- 1:04:24uh, existed.
- 1:04:25ElevenLabs has been around for a while,
- 1:04:28but it wasn't very natural.
- 1:04:29Uh, now it's starting to be much more natural,
- 1:04:32and I actually made one in English on HeyGen recently.
- 1:04:36I opened a separate English channel
- 1:04:40I posted it, for example, there are only eight subscribers on that channel,
- 1:04:43and it got one hundred fifty, two hundred views, like very quickly.
- 1:04:47Maybe it's more now.
- 1:04:48I mean, I checked just an hour or two later,
- 1:04:50and the content-- the topic is really in demand, I think
- 1:04:53Uh, I mean, I'm already a big follower of your channel
- 1:04:56and I'm sure it'll grow and progress very quickly
- 1:04:59because everyone benefits from each other.
- 1:05:02Like look, we're sitting here chatting now,
- 1:05:04we're learning things because there's a hunger for information.
- 1:05:08Uh, podcasts,
- 1:05:09for example, Turkey isn't a podcast country.
- 1:05:12One of our podcasts with Ali was watched hundreds of thousands of times.
- 1:05:17Uh, the podcast we did with a friend named Ertan, was it one
- 1:05:20hundred twenty thousand last I checked? I mean, normally Turkish people don't
- 1:05:23watch something for two hours very easily, but this is such a topic, the AI topic.
- 1:05:28There is a significant gap, uh,
- 1:05:30I mean, it is not just in Turkish.
- 1:05:32There is a noticeable gap in Spanish, and there is also a gap in French.
- 1:05:36This is because very few people
- 1:05:38actually use it and are willing to share their experiences.
- 1:05:41I mean, since the people who utilize the AI world the most are developers,
- 1:05:45and developers might not always be the most extroverted people
- 1:05:49in general So you are likely one of those rare exceptions, wouldn't you
- 1:05:52say? Generally speaking, most people are more introverted.
- 1:05:56Yeah, for example,
- 1:05:57not a single person in five years has asked me, "Are you a developer?"
- 1:06:01Everyone asks, "Do you manage social media?"
- 1:06:03Or, "Are you on the marketing side?"
- 1:06:05They especially ask about marketing
- 1:06:07a lot because I'm an energetic person. They directly ask, "Are you a marketer?"
- 1:06:11Right? It's very rare
- 1:06:16Ah, yeah, but you have marketing skills. You have marketing skills too.
- 1:06:20So we can think of you
- 1:06:21as a full stack marketing plus developer. Hey, don't let them give you
- 1:06:25extra work though. If any bosses are listening, they're definitely like that.
- 1:06:28Look, you were talking about this recently. Uh,
- 1:06:30they come to you saying, "Sho, eh, you're packaging YouTube?
- 1:06:33Package us too, for real."
- 1:06:36Well, that's actually what happened to me.
- 1:06:38When I first joined a startup,
- 1:06:40those who know startup culture will know, there are 20 or 30 people.
- 1:06:43There, uh,
- 1:06:44I was an ambitious young guy.
- 1:06:46We are still young, but I'm talking about when I was 22, 23.
- 1:06:50Since I was an ambitious person, they put me into everything
- 1:06:53marketing, contracts,
- 1:06:56software.
- 1:06:57I was thinking, "Oh, they're training me so well."
- 1:07:00But it's not exactly like that.
- 1:07:02I mean, there's a bit of exploitation too,
- 1:07:04but this taught me a lot.
- 1:07:06Like I learned marketing.
- 1:07:08For example, when I moved to marketing,
- 1:07:10there were very frequent meetings at the startup
- 1:07:15I learned the culture there. I learned how to talk to people.
- 1:07:18How do you market a product? How do you manage those meetings?
- 1:07:21Later, I combined the knowledge I learned there with AI,
- 1:07:24I mean, with software, and now I'm transferring it from software to AI
- 1:07:29Sometimes we can actually turn those bad situations into something good.
- 1:07:32That is where my marketing skills come from, for example, entirely
- 1:07:36from what I have learned
- 1:07:39Uh, tru-true, she, she, uh, see,
- 1:07:42i-it teaches you a lot.
- 1:07:43At a startup, unlike a corporation,
- 1:07:46you see the big picture since you do everything. I recall my agency days
- 1:07:51Man, we've had some hilarious stuff happen. If I told you,
- 1:07:53you might not even believe it. There was this developer friend, Erdem.
- 1:07:57Uh, if he's listening, I mean, Erkin, sorry, his brother's name is Erdem
- 1:08:01Uh, he's a truly talented developer, and honestly, he's such a funny guy.
- 1:08:05Anyway, back then there weren't many of us, uh,
- 1:08:08maybe 10 or 15 people, something around that.
- 1:08:11Uh, we were in this old historic building in Izmir, just like a Greek house.
- 1:08:16And we had this, uh,
- 1:08:17lady at the office who handled all the food and stuff. She'd cook, clean,
- 1:08:21and keep things organized. You know how it is.
- 1:08:25Working, uh, well, right around that time, he ended up leaving the
- 1:08:28job, right? And we had really gotten used to it, so we didn't wanna start ordering
- 1:08:31out. You know, she would always make whatever we felt like eating. So I asked
- 1:08:34everyone, "What-- Guys, what are we gonna do now? And we need to find someone else."
- 1:08:38Uh, Jurkan said, "You know my brother?" He said, "Erdem."
- 1:08:40I already knew Erdem, so I said, "So what?" He said,
- 1:08:43"He's going, he's going to the military. He's free right now. Let him come."
- 1:08:46Uh, he'll do it. Does he, does he know how to cook? He doesn't, but he'll learn, man.
- 1:08:49He said, "No big deal." I said,
- 1:08:51"All right, let him come and start then." Uh, so he came His job description was,
- 1:08:55uh, reception duties, answering the phones, uh, client follow-up,
- 1:08:58as well as cooking and cleaning
- 1:09:03Ah, and he'd do these experimental things like he was
- 1:09:06some master chef. Like he'd add sage to lentil soup, man, stuff you could never
- 1:09:10eat in your life. Man, I'm not a picky guy at all. I'm 130 kilos.
- 1:09:13I'll eat anything
- 1:09:15Uh, you know, if something is edible, I'll eat it.
- 1:09:17If it doesn't smell bad, I'll eat it.
- 1:09:19Man, honestly, there were even some meals I couldn't bring myself to eat.
- 1:09:23Anyway, uh,
- 1:09:24after that, Erdem left for military service, and we,
- 1:09:28uh, somehow managed to find another lady for the office and moved on.
- 1:09:32Startups really have this advantage.
- 1:09:34Uh, they are great schools for learning
- 1:09:37I, i, you, I mean, when you're in, you see everything. It's like a family.
- 1:09:41You know that Instagram post,
- 1:09:43we're like a family?
- 1:09:44Uh, but it really is like that
- 1:09:47Uh, you see everything there and you learn so much. I mean, when you leave a startup,
- 1:09:52you usually either land a role at a great company
- 1:09:54or you have the skills to start your own business. But in corporate,
- 1:09:58they give you one single task.
- 1:10:00It's very clearly defined.
- 1:10:01It feels like they just won't let you do anything else.
- 1:10:04Yeah, for sure.
- 1:10:06In a startup,
- 1:10:07you,
- 1:10:09you hang out with the CEO, you're active with the team lead.
- 1:10:12If needed,
- 1:10:13you even go talk to the client.
- 1:10:15I mean, for starters, I tell people who want to develop themselves to join
- 1:10:19a startup in any field, then jump to corporate
- 1:10:21What's your take on this?
- 1:10:23When you start directly in corporate,
- 1:10:25yeah, like you said, you stay on the same project for three months.
- 1:10:28You don't learn anything
- 1:10:30I mean, a person needs to learn how to potty too
- 1:10:35-Who's Carrie DeNotti? -That is true
- 1:10:39Yeah, that's very common in Turkey, but abroad-- I lived in London for six years,
- 1:10:43from 2018 to 2024, even until
- 1:10:462025. Uh, I mean, in London, for example, there's no such thing, man. Like, you go
- 1:10:50visit someone at a company, there's no one bringing tea or serving like we have. If
- 1:10:54you want tea, you go to the CEO, the CEO pours it for you himself and brings it,
- 1:10:58or you go and get it yourself.
- 1:11:00I mean, we have a lot of that. Our company environments are really comfortable,
- 1:11:04you know, happy Fridays, Fridays this and that, meals and all.
- 1:11:08Back then, the opportunities were better too, of course.
- 1:11:11Uh, for a while, the Turkish economy
- 1:11:14in those 2010s was relatively more active
- 1:11:182007, 2008, the 2010s especially. Uh, maybe there were more opportunities then,
- 1:11:23but still you go to a bank and there's someone taking tea to everyone, you know?
- 1:11:27For example, there's no such thing at Barclays in England
- 1:11:30No way. I mean, because that's very costly abroad because abroad,
- 1:11:34especially in London,
- 1:11:35salaries don't change much based on the type of work.
- 1:11:38I mean, today a banker and a cleaner earn very similar money.
- 1:11:42I mean, it's turning into that in Turkey too,
- 1:11:44because there are so many white-collar jobs.
- 1:11:46University graduates have increased so much that now, for example, you can't find
- 1:11:50a plumber, you can't find a painter.
- 1:11:52But if you say accountant, there are plenty
- 1:11:54That is precisely
- 1:11:55the reason why things have balanced out over there. Many people studied abroad,
- 1:11:59and now those blue-collar jobs bring in a much higher income.
- 1:12:02In England, a blue-collar role actually pays more than
- 1:12:05the average white-collar position. It is, in fact, the reality across the entirety
- 1:12:08of the European continent.
- 1:12:10It is the same in Turkey too.
- 1:12:12Things are just like that in Turkey, by the way. Still, yeah, I mean
- 1:12:14Seeming dirty now though
- 1:12:16In Turkey too. It's like that in Turkey now, yeah?
- 1:12:17Our new dishwasher broke down, for example, the repairman who came
- 1:12:21He brought his kid on board right after college.
- 1:12:23I mean, he's making way more than a normal salary.
- 1:12:26He's actually pulling in over a hundred thousand right now. Just starting out.
- 1:12:29Even an apprentice makes over a hundred thousand, for real
- 1:12:33Uh, I came across a similar example recently. I was getting a haircut.
- 1:12:36I just went in and sat down. I'm the type who goes wherever there's an opening. I'm
- 1:12:39not picky. You know those people who are loyal
- 1:12:41to their barber, like, "I won't change my barber," and all that? Uh,
- 1:12:44I'm not like that. I just go wherever my first criterion is finding a parking spot
- 1:12:48My second criterion is I take a look from a distance, and if it's empty,
- 1:12:52I'm like, "Okay," and I go in because, I mean, how much better can I look?
- 1:12:55I figure there's not much to work with anyway. Anyway,
- 1:12:58uh, I said, uh, I found a place and went in, right?
- 1:13:01The kid started cutting and cutting
- 1:13:02So I'm just messing around, checking my phone and stuff
- 1:13:05then I thought, "Let me see who's cutting my hair." I looked, and man,
- 1:13:09that kid was literally like fourteen years old Like, uh, he's just a kid, a teenager.
- 1:13:13Definitely not 18, maybe 14 or even 13. I said, "Hey man, how old are you anyway?"
- 1:13:18Uh, cutting my hair.
- 1:13:20He was like, "Uh, sir," he said, "I'm 14." He really was 14. I said,
- 1:13:23"Meh. So what are you doing? How does that work?" And all that. "I mean, how did you
- 1:13:27become a barber? When did you become a barber man?" I asked. "Well," he said,
- 1:13:30"Sir, I'm studying barbering." So he's in the barbering department in high school.
- 1:13:34Apparently, there's a department for that. For
- 1:13:37So Gakko say,
- 1:13:38"Go the, uh, these guys go to school one day and intern for four."
- 1:13:41I ask everyone right away,
- 1:13:43"Do you use any AI?"
- 1:13:44"Of course I do, sir," he said. "Apko, uh, which one?" "Okay, Gemini," he replied
- 1:13:49They call it Gemini shirt,
- 1:13:50Gemini. So he uses it.
- 1:13:52I said, "Okay, man,
- 1:13:53are you thinking about going to college, taking the entrance exam?" "No way, sir,"
- 1:13:58he said. "Am I crazy?"
- 1:13:59"Why?" I asked.
- 1:14:01"Sir," he said, "I'm already making more than minimum wage at my age."
- 1:14:06"Oh, oh, if I open my own barbershop in the future," he said
- 1:14:09i said, "You're right, man."
- 1:14:11I mean, here's the thing. I'm not against college.
- 1:14:13I went to college.
- 1:14:14I did a master's.
- 1:14:15I started a master's.
- 1:14:17Uh, I even went to two different universities
- 1:14:20Ah, I see, I studied engineering and also economics and administration
- 1:14:24but the point is
- 1:14:26ah not everyone needs college.
- 1:14:27The, the, there used to be this educated kid idea
- 1:14:31but now everyone's studying This time there's no one to do the actual work. I
- 1:14:35mean, that's where the real demand is. Uh, that's where the real gap is Uh, I mean,
- 1:14:39we've shifted the balance,
- 1:14:41uh, a, a bit too much. Everyone's a college grad,
- 1:14:43but there's no need for it, man.
- 1:14:45I mean, there's a real need for people to do the daily jobs.
- 1:14:48Uh, there's no one doing them.
- 1:14:50I mean, you can't find them when you're looking
- 1:14:54Yeah, that happened a bit because of that.
- 1:14:56Was it because of our Gen Y?
- 1:14:58Our parents said, "Be a white-collar worker, have insurance, get married,
- 1:15:03live a happy life."
- 1:15:05That's how it was in their time.
- 1:15:07They could buy their houses and cars very easily.
- 1:15:10Nowadays, a white-collar worker can't just easily buy a house or a car
- 1:15:15I'm talking about the majority, really.
- 1:15:18If you don't have experience, you can't just start with a good salary anyway.
- 1:15:22I agree, by the way.
- 1:15:23I mean, being in a different profession is never a bad thing
- 1:15:27I tried to do some housework for a while, but I'm asking AI,
- 1:15:30sending photos, sending videos and stuff. I'm like, "Look, I'm doing it like this."
- 1:15:34I mean, I'm just trying to figure it out
- 1:15:39Right? I mean, there are really, uh, what makes a person happy?
- 1:15:43I have met some very interesting people.
- 1:15:45Uh, for example, I had a lawyer friend,
- 1:15:47uh, for a while.
- 1:15:49Uh, he went into shepherding. A shepherd, the guy was like a different kind of guy
- 1:15:53Yeah, sounds good
- 1:15:54I still see him. If Burak is watching
- 1:15:56Or n-,
- 1:15:57uh, the, the lawyer said, "I'm going to be
- 1:16:00a shepherd." "Man," he said, "I'm tired of this lawyer life, too."
- 1:16:03"All right, man. Fair enough." So he went and got a pen, bought some goats and all
- 1:16:08Uh, he had, uh, like a Renault Twingo car,
- 1:16:11uh, and he had a dog. The dog had eaten the front seat of that car. I mean,
- 1:16:14half the seat was gone and all.
- 1:16:16Man, the guy was
- 1:16:17so happy. I mean, later on, the business didn't work out. I mean, he couldn't make
- 1:16:21money because it's hard work Well, but I told you that I've been living out in the
- 1:16:25village lately. Then I left the village and saw this guy standing on the corner
- 1:16:29Yet the, a Sam Yishi, uh, he's at least fifteen or twenty years older. I asked,
- 1:16:33"What are you doing?
- 1:16:34He has, he has sheep."
- 1:16:36"Well, Umut," he said, "I was queue, I was free.
- 1:16:38I, uh, I bought sheep and grew the flock. They know their stuff.
- 1:16:42If I tried, they'd die.
- 1:16:43I grew the flock. Uh,
- 1:16:45my, my, my kid was marrying,
- 1:16:46so I bought him a house in a few years."
- 1:16:49And after that he has started all over again now with 10 of those items
- 1:16:52I'll keep increasing it.
- 1:16:54Then maybe I'll do something for the grandkids or whatever," he says.
- 1:16:58What I mean is this.
- 1:16:59Actually, one of the main reasons we're consuming things
- 1:17:03at such high prices today
- 1:17:05is that there are so few producers I mean,
- 1:17:08if you had producers today,
- 1:17:10when there's supply, as long as demand for something in the economy doesn't decrease
- 1:17:15if supply drops, the price goes up.
- 1:17:17They call this the invisible hand
- 1:17:19in economics. Just leave the market alone with that supply-demand balance.
- 1:17:22The price will find a zone, it'll be set at the right level
- 1:17:26What if the state steps in and regulates? Nothing happens
- 1:17:29Then nobody does livestock farming. I mean, you set a price
- 1:17:32for meat and that's it. Then nobody can do it because they can't cover the cost.
- 1:17:35Anyway, look where we started and where we ended up. Just chatting, you know?
- 1:17:40And, uh, but on the production side, we really all need to pay attention there
- 1:17:44we need physical things
- 1:17:46we won't starve to death without code but or without the marketing work I do
- 1:17:51uh, you won't die if you don't do it, but you do die when you don't eat.
- 1:17:58I agree with Ali, by the way.
- 1:18:00I'm being totally sincere.
- 1:18:01I even looked at places,
- 1:18:03not just in Turkey, but abroad too, like in Asia and Europe.
- 1:18:07I thought about becoming a doner kebab shop owner.
- 1:18:10There's so much demand, and during my university years,
- 1:18:14I worked as an assistant cook to make some money.
- 1:18:17I gained some skills from that too.
- 1:18:19I'll use those skills directly there
- 1:18:22Let's open one together, the three of us. I'm in too. I like investing in Danone.
- 1:18:25I'll invest in a donut shop
- 1:18:28We can keep producing content in the background too, by the way
- 1:18:34Man, and we'll sell that story anyway. You make more money
- 1:18:36from the story than the donor, let me tell you
- 1:18:43By the way, Sai,
- 1:18:44I, I find that doing physical work is actually quite fun.
- 1:18:46For one thing, you don't have to think much while doing that kind of labor.
- 1:18:50You just lose yourself in the process.
- 1:18:52Sure, your body gets tired, but the moment you hit the pillow, you're out.
- 1:18:56You don't have five million things running through your head.
- 1:18:58It has that kind of beauty.
- 1:19:00Physical work,
- 1:19:01yeah, I'm not looking down on it.
- 1:19:03Hard work is very valuable,
- 1:19:05but I think white-collar work is hard work now too there's no difference anymore.
- 1:19:10A mine worker's job is obviously very risky.
- 1:19:13It's a much harder job.
- 1:19:14But let's say an average job,
- 1:19:17like someone selling water or a grocer,
- 1:19:20someone working in a grocery store versus a white-collar worker, they're both.
- 1:19:25Uh, since everyone is white-collar now,
- 1:19:27they're workers too We're, we're workers too in that sense. We're workers,
- 1:19:31we're right, we'll win and all that
- 1:19:33I really think having a plan B is valuable.
- 1:19:35This is very important, by the way.
- 1:19:37I listened to this one
- 1:19:38economist, I'd recommend him too
- 1:19:41There is a fellow known as Gary, the Economics.
- 1:19:44His full name is Gary
- 1:19:45Stevenson. He has recently become a major bestseller throughout
- 1:19:48the United Kingdom. For instance, we were much poorer during our own childhood years
- 1:19:52Jack, uh
- 1:19:53Uh, our families were poorer.
- 1:19:55We had access
- 1:19:56to fewer things, that's for sure. Those who lived through those times know.
- 1:20:00Those who know my childhood know. If you were a civil servant's
- 1:20:03kid, you'd maybe eat out once
- 1:20:05a month, maybe not even that, and that would be a pita place, let me tell you.
- 1:20:08Um, I see
- 1:20:08back when they got their monthly salaries.
- 1:20:11Yet people were still allowed to own personal property.
- 1:20:13For instance, my parents bought their home through a cooperative deal
- 1:20:17long before they retired.
- 1:20:19Nowadays, you can't really own anything anymore.
- 1:20:22Everything is tied to a monthly subscription.
- 1:20:24We'll soon have to subscribe to life itself.
- 1:20:26It feels like everyone is just renting.
- 1:20:28That's why I suggest following Gary Stevenson
- 1:20:32He says, uh,
- 1:20:33actually the biggest problem is that, uh,
- 1:20:36income inequality is increasing worldwide.
- 1:20:38The rich get very rich, the poor get very poor.
- 1:20:40This is something proven statistically too
- 1:20:43And as the wealthy get richer,
- 1:20:45these assets inevitably start ending up in their hands.
- 1:20:48When a house hits the open market,
- 1:20:49what is that rich guy actually going to do with all that money?
- 1:20:52He is already spending just about as much as he possibly can.
- 1:20:56He will either choose to buy gold,
- 1:20:57hold onto dollars, or invest in more real estate. And you see, even though
- 1:21:01nobody can afford to buy real estate, those prices are still incredibly high
- 1:21:05Why? Because there are people buying, a very small group is buying these
- 1:21:08Yeah, and
- 1:21:09So my, my advice to young people is this:
- 1:21:12buy a house somewhere.
- 1:21:14It doesn't matter where.
- 1:21:16Buy a one-bedroom.
- 1:21:17Go buy it in some district you've never heard of, but have a house,
- 1:21:21have a place to call your own
- 1:21:24Uh, and I mean, uh, or at the first opportunity, I say to myself now,
- 1:21:27"Man, I wish I'd bought more real estate when I was young." And the most,
- 1:21:31uh, important thing is to have a roof over your head. Buy it in a village if you have
- 1:21:35to, just as long as you have somewhere to live, because you won't be able to later.
- 1:21:38I mean, this, uh, income inequality is rising all over the world.
- 1:21:41If you don't buy today, you won't be able to buy at all tomorrow or the day after.
- 1:21:45However hard it is today, it'll be even harder to buy in five years.
- 1:21:48Yeah, sure.
- 1:21:49Yeah, I agree.
- 1:21:50And, uh, it's actually related to the topic,
- 1:21:53but let me give an example from my own age group, between 25 and 35.
- 1:21:58I divide those people into two or three groups.
- 1:22:00Most of the people who are software developers but also do other things
- 1:22:04have bought their houses and cars by now.
- 1:22:06For example, a developer, you know, works for a certain salary
- 1:22:09but does e-commerce on the side, trying to sell something
- 1:22:13Or like a friend of mine
- 1:22:15working in digital marketing creates content and makes money from that too.
- 1:22:20They've made the world such that
- 1:22:22you know how people who couldn't make ends meet used to work a second job?
- 1:22:26It's kind of like that now.
- 1:22:28One job usually isn't enough.
- 1:22:30You need to do more than one.
- 1:22:32People like that have bought their houses and cars in my circle
- 1:22:37So I recommend this. They should get into many different things,
- 1:22:40like get into painting.
- 1:22:41For example, I had an engineer friend in Germany
- 1:22:44He was employed as an engineer, then he started working out.
- 1:22:48He started doing sports professionally too, got certificates and all.
- 1:22:51Now he's got residency in Germany, became a sports instructor I mean,
- 1:22:55look how far he's come.
- 1:22:56I really recommend it to people.
- 1:22:58Don't stick to one profession, do multiple things, develop yourself, I say,
- 1:23:03but people take it the wrong way.
- 1:23:05Like, did we study this
- 1:23:06profession for nothing? You're actually adding something to yourself. You're
- 1:23:10not studying completely for nothing
- 1:23:13Yeah, I think it's very right not to be-- I mean, not being
- 1:23:16proud is important in these things. Don't let pride get in your way, you
- 1:23:19know? Uh, because you touched on something very nice. You said the old-timers
- 1:23:23Exactly. So
- 1:23:24Those old timers are exactly like my late father.
- 1:23:27My father was a primary school teacher.
- 1:23:29Uh, my father was already up there in age,
- 1:23:31but, uh, I mean, he was a very,
- 1:23:33uh, uh, very hardworking man, okay?
- 1:23:35He was an incredibly hardworking man, but he had to be
- 1:23:38I'm sorry for your loss
- 1:23:39That's how he could buy a house. Otherwise, he
- 1:23:40couldn't have bought a standard house.
- 1:23:42Look, I'll tell you what he did. Primary school teacher. We had morning
- 1:23:45and afternoon shifts. I mean, it was two things. There was no such thing as a full
- 1:23:49day school in our time. You were either in the morning shift or the afternoon.
- 1:23:52You start at seven in the morning, say it ends at twelve
- 1:23:56So if the shows, if the other one starts at eight or nine,
- 1:23:59it ends at three. Now, my father was, he was a teacher in the morning shift.
- 1:24:03He'd go to school in the morning, he'd leave school, he'd sell slippers
- 1:24:06to schools on installment.
- 1:24:08From Istanbul, in our childhood, there were those leather-soled slippers.
- 1:24:11And he'd sell those slippers, give them out, and wouldn't even take money. It
- 1:24:14really suited them to pay in installments. There were no credit cards or anything,
- 1:24:17you know? But on paydays, he'd be right there. He followed that very well
- 1:24:20Oh my, well
- 1:24:21He followed that very well. That was his second job.
- 1:24:23He was doing it at the same time. His third job was giving private lessons.
- 1:24:26These all happened within the same day.
- 1:24:28His fourth job back in the day in Izmir, there were the Maksimum Kasalar casinos.
- 1:24:33He was the manager on the night shift there. Look
- 1:24:35He'd also go at night and work at the casino.
- 1:24:39Thanks to that,
- 1:24:40back then he was very like, you know,
- 1:24:42I've like listened to everyone from the Bülent Ersoy's
- 1:24:45to the Ibrahim Tatlises's. I mean, I've listened to them
- 1:24:48because I went with my father like that. It was in Izmir
- 1:24:51and, uh, like he, he'd do four jobs.
- 1:24:54He'd sleep for two or three hours.
- 1:24:56At one point, they even gave him the restroom
- 1:24:59The, the restroom was very good money back then. You'd take money from whoever came
- 1:25:03to the restroom and that money would stay directly with you. I mean,
- 1:25:06the venue would give you the management of it. You're inside the restroom
- 1:25:09there are still places with that old school mindset now.
- 1:25:12Ah, there's someone standing there offering you cologne or something similar
- 1:25:15there are perfumes and all sorts of items. I mean,
- 1:25:18that was the only way he could really earn a living back in those days
- 1:25:21then eventually things got a bit more balanced, man.
- 1:25:24Then we,
- 1:25:26we return to where you said again.
- 1:25:28We return to the era of our childhood again. I agree with you, but whatever you
- 1:25:31do, work very hard for a while. I mean, have a house, like at least have a house
- 1:25:35you can live your life in, because nowhere in the world can people live
- 1:25:38on a pension, not just in Turkey. Look, in Turkey they can't live on it at all anyway
- 1:25:41Yeah, for sure
- 1:25:42Yeah, sure.
- 1:25:42Uh, in England, I don't know if you ever go abroad, but go to Europe,
- 1:25:46go to England man, all the cashiers are
- 1:25:48eighty years old why do you think this guy is working?
- 1:25:50Because if he didn't buy his house, he can't survive.
- 1:25:54For example, the pension in the UK is 800 pounds.
- 1:25:56With 800 pounds, rent in the UK starts from 1,500 pounds.
- 1:25:59So think about it
- 1:26:01I have been thinking about something lately. Uh,
- 1:26:04I received a job offer from somewhere in Europe.
- 1:26:06The salary looks higher than what I'd make in Turkey, but it really isn't
- 1:26:10As you said,
- 1:26:11life in Europe is more expensive.
- 1:26:12If I go there with that money, I definitely won't be able to save.
- 1:26:16Maybe I'll have to live in a house with two other people
- 1:26:19because shared houses are very common in Europe. You know, you share the house
- 1:26:23with people you don't know because the rents are very high. Now let's say I went,
- 1:26:27I won't be able to save money and it won't make sense I won't be able to buy my house
- 1:26:32either. I mean, of course, there's no problem with working until the age of 80,
- 1:26:36but it's also a problem, you know? After a certain age, people want
- 1:26:39to rest. They say, "Let me at least work while sitting." That's why it's a problem.
- 1:26:43You have to think about the future, too. Oh, I'm not thinking
- 1:26:46about retirement in Turkey, by the way. I've put retirement out of my mind.
- 1:26:50Um, I will still be working with agents when I am 80 years old.
- 1:26:53I want to make that perfectly clear
- 1:26:56Of course, I find myself in complete agreement with you
- 1:26:58Well, uh,
- 1:26:59a lot will change by the time you're 80.
- 1:27:02Uh, you still have 50- something years until you're 80.
- 1:27:05Uh, so, uh, I can't imagine you're 80 really, what kind of world it will be.
- 1:27:09I mean, for me, two or three years from now is,
- 1:27:12is vague.
- 1:27:13Uh, I really know.
- 1:27:14But I know this,
- 1:27:15uh, I know this clearly
- 1:27:18I mean, buying your own house, at least for example, I think you're
- 1:27:21doing this YouTube thing very well.
- 1:27:23I mean, YouTube at some point, uh, maybe we're
- 1:27:25not doing it as a main job right now.
- 1:27:27We're doing it a bit hobby-oriented,
- 1:27:29not income-oriented. But, uh,
- 1:27:31at some point, I think it also allows you to put your eggs in different baskets.
- 1:27:35I mean, in the future, for example,
- 1:27:37if you have time and produce more content,
- 1:27:39it will be watched more
- 1:27:42By the way,
- 1:27:44everyone underestimates social media,
- 1:27:46but I'm very surprised because we
- 1:27:48actually found the client I'm working with in America right now through social media.
- 1:27:53I mean, he saw our videos on YouTube and worked from there.
- 1:27:56And I'll give you an interesting piece of information.
- 1:27:59Uh, I've changed an average of eight companies so far.
- 1:28:02I only applied to two of these eight companies.
- 1:28:05I didn't apply to the rest They all found me through my projects on YouTube,
- 1:28:09and I had posted very simple projects on YouTube five years ago,
- 1:28:13HTML, CSS.
- 1:28:14I mean, it incredibly increased my brand awareness,
- 1:28:17you know, as a name, as a brand
- 1:28:20I recommend this to people too.
- 1:28:21I'm not saying everyone should go and shoot YouTube videos,
- 1:28:25but I'm saying show what you do.
- 1:28:27I mean, you can post it on LinkedIn, Instagram, Twitter.
- 1:28:30And if a friend of mine named Kerem is watching me,
- 1:28:32he's one of the best designers I've ever seen.
- 1:28:35Forbes saw the design he posted on Twitter.
- 1:28:37Forbes? Forbes?
- 1:28:40And they ended up working quite closely with him. On the American side,
- 1:28:43his design was displayed on those massive digital
- 1:28:45screens at the New York Times. They actually caught sight of this on Twitter.
- 1:28:48He was working directly with Twitter, I mean. It is all very strange indeed
- 1:28:52Definitely. I mean, the, uh, you have
- 1:28:54to overcome shyness somehow because, you know, some people are really very
- 1:28:58talented. I see very talented people, but, you know, they don't share it with anyone.
- 1:29:03They don't share it. They're hesitant. I mean, it's useful not to be
- 1:29:06hesitant because when you're hesitant, people can't see what you're doing.
- 1:29:09I definitely agree with that.
- 1:29:11YouTube, you know, for your personal brand
- 1:29:15Uh, it's very valuable for polishing.
- 1:29:18It's valuable for this.
- 1:29:19You know, it's not valuable if we go and talk nonsense on YouTube.
- 1:29:23But on YouTube,
- 1:29:24if you say what's what,
- 1:29:25an agent is used like this, not like that,
- 1:29:27this is hype, this isn't I mean,
- 1:29:29when you explain the job realistically, maybe a million people don't watch you.
- 1:29:33I mean, when you say that
- 1:29:34a million people don't watch you, but those 30, 40,000 people,
- 1:29:38100,000 people, 50,000 people,
- 1:29:40200,000 people, whatever, who watch you, they're already decision makers,
- 1:29:44sensible types. And, uh, you, I mean, you have to think of a person like
- 1:29:47this. When do you start trusting a person? You spend time with that person,
- 1:29:51right? You know them not when you see them once, but when you go on a holiday.
- 1:29:54I mean, because you spend a lot of time. You know their bad side, their good side.
- 1:29:57Now, for us, too, I think being on the screen has something like this.
- 1:30:00I mean, for example, there are people who are annoyed by me, by the way I'm fine
- 1:30:04with that. I don't like everyone either. Not everyone has to like me I mean,
- 1:30:07they might feel annoyed by me,
- 1:30:09but at the very least, you know,
- 1:30:11they watch and gather enough data to decide whether to be annoyed
- 1:30:14or to actually like me. I mean, look at this guy. You know,
- 1:30:17people can really get to know you. That is what I am trying to say.
- 1:30:21Uh, anyway, no one is ever loved by everyone in any way, shape, or form
- 1:30:26So now not everyone's gonna like you, you know?
- 1:30:28Someone actually messaged me the other day.
- 1:30:30Uh, it's usually positive.
- 1:30:32On YouTube may- maybe
- 1:30:3399% of the comments are positive. But someone said, "Man, I watched your video,
- 1:30:37but you really annoyed me." I said, "Well,
- 1:30:39we'll be here whenever you feel like being annoyed." I mean, what else can I say?
- 1:30:43So I, I, I mean, I get annoyed by some people too, but I really like others.
- 1:30:47For example, my wife might find someone I really like annoying. Uh, she'd go
- 1:30:51So, uh, it is, uh, quite personal. But above all else, you really get the chance
- 1:30:55to know them, don't you think? Now, if I have personally watched
- 1:30:5963 of your videos and you have never once done me wrong, which to be fair, you never
- 1:31:02have, then I say to myself, "This person is completely harmless. They are truly
- 1:31:06a decent human being." There is simply no better way to really get to know someone.
- 1:31:10That is exactly why investing in your personal brand on platforms like LinkedIn,
- 1:31:13Twitter, YouTube,
- 1:31:14or wherever you happen to be is so incredibly valuable.
- 1:31:18Honestly, it is really not that hard to do So ma--
- 1:31:21I think cheat, cheat, uh, just being natural is key, not trying to prove things
- 1:31:24as if you're someone else. I'm not a big fan of those o- o- overhyped video styles.
- 1:31:28I prefer things to be more grounded.
- 1:31:30I say we shouldn't die or be a Warner
- 1:31:33Uh, the crew, we be sure we didn't prepare anything. We just said we'd talk
- 1:31:36about these things beforehand. No script or anything, we just sit and talk.
- 1:31:40Uh, I learn a lot from this podcast too. Show
- 1:31:42the ga- uh it naturally goes in all sorts of directions afterwards
- 1:31:47Um,
- 1:31:48but by the way, what you said at the beginning was true.
- 1:31:51There are very few podcasts I listen to in Turkey.
- 1:31:54You're one of them. You're in my top five.
- 1:31:56There really aren't many, and I think what you're doing is very valuable.
- 1:32:00I got to know Ali Bey through you and other YouTubers too
- 1:32:04Um, I thought, um, wow, he's doing such
- 1:32:07great work talking to people, guiding them, and so on. For example, I'll have
- 1:32:11the podcast on in the background while I'm working. I'll laugh occasionally, stop
- 1:32:14what I'm doing, and look at the screen. I think these are really great things
- 1:32:18Um, uh,
- 1:32:19y- yeah, definitely.
- 1:32:20Uh, thanks so much for saying that, by the way.
- 1:32:23I'm also really happy about this.
- 1:32:25For instance, you and I wouldn't normally find, uh,
- 1:32:28a- an hour or an hour and a half
- 1:32:30just to sit and talk.
- 1:32:31We're talking, we're learning from each other,
- 1:32:34or maybe not, but I think we are, and everyone gets to watch it
- 1:32:38Honestly, that's what I find so rewarding about the whole podcasting thing.
- 1:32:42I've always been someone who loves to talk and, uh, more importantly, to listen.
- 1:32:46I truly enjoy the act of talking
- 1:32:48ultimately, I'm just doing what I love, man.
- 1:32:51Furthermore, in addition
- 1:32:52to all of that, we are helping people in many small, meaningful ways.
- 1:32:55Uh, and honestly, that is why I find this work so rewarding Uh, thanks a lot.
- 1:32:59I'm glad you like it.
- 1:33:01Uh, I'm actually filming a podcast with Ali tomorrow.
- 1:33:05We'll hopefully release it as soon as possible.
- 1:33:07Let me tell you the topic. We're going to talk about, uh, what do you think?
- 1:33:11I'm curious about your thoughts on this,
- 1:33:13and if there's anything you want us to discuss in that podcast, let me know
- 1:33:17The main topic we have here is this
- 1:33:21It is much easier to build things now, but selling them remains just as difficult as
- 1:33:25it was before. Like you said, there are so many people building apps today.
- 1:33:28Professional developers are building them too, and they are doing some truly amazing
- 1:33:32things.
- 1:33:34Since I know a lot of developers from the agency I used to work
- 1:33:37at, they all send me their stuff like, "Hey, I made this, I made that."
- 1:33:41Well, you made it, but selling it, the marketing, the meta, Google, Fa- uh,
- 1:33:45Facebook ads, creating content,
- 1:33:47UGC, the business plan,
- 1:33:49scaling, monetization, go,
- 1:33:51um
- 1:33:53ah issues are all separate issues separate issues. That's where most people get stuck
- 1:33:57most people can't even build something decent if they're not developers
- 1:34:01some can but even if they do they can't sell it.
- 1:34:04They, they,
- 1:34:05Uh, we're talking about this tomorrow. Like, you made it, but how are
- 1:34:09you going to sell it, buddy?
- 1:34:10Uh, what do you think about that?
- 1:34:13That is a truly fascinating topic to discuss.
- 1:34:16Right now, I am in the process of launching a new product.
- 1:34:19I do need to generate some sales after all.
- 1:34:21Developing a product usually takes about three or four years.
- 1:34:24It does not happen very quickly. Most major companies have stood
- 1:34:27by their products for years because they truly believed in them. Not all
- 1:34:31of them were instant hits, for example. Now I'm thinking about myself So, uh,
- 1:34:36Yıldız, you've been a white- collar worker until now.
- 1:34:40Now you're actually going to launch a product and sell it. Can you do it?
- 1:34:44I thought about it and decided this.
- 1:34:46I'm not training myself in just one area anymore
- 1:34:49because AI is an incredible teacher.
- 1:34:52I talk to Claude daily about things like, uh, the cloud economics,
- 1:34:56uh
- 1:34:58We're always chatting in other areas and so on.
- 1:35:01It's really starting to make me think like a CEO, and it even warns me.
- 1:35:05It started saying things like, "Don't think so simply.
- 1:35:08Act like a CEO."
- 1:35:10So thanks to it, it's taking my place
- 1:35:14People also focus on one thing when selling, the money issue.
- 1:35:18Yes, money is important, of course, to maintain our lives.
- 1:35:21But the broader your vision, the more money comes to you anyway.
- 1:35:25For example, most
- 1:35:26of the customers who come to me have worked with different developers,
- 1:35:30but there's always one sentence said to me, "Ms.
- 1:35:32Yildiz, we really loved your vision."
- 1:35:35Your design eye is great.
- 1:35:36You're different from the rest, they say.
- 1:35:39Now, in the AI age, some features will set you apart from others.
- 1:35:42It could be your design eye, your systemic thinking, or your marketing side.
- 1:35:46We can already access everything anyway
- 1:35:49Start differentiating yourself to sell the product.
- 1:35:52Because as you said, everyone is producing something now.
- 1:35:55So personal branding will stand out here.
- 1:35:57That's just my personal opinion completely
- 1:36:02Uh-huh. Okay, great. We've slowly reached an hour and a half.
- 1:36:05Of course, it wouldn't end. We could talk for five hours, but, uh,
- 1:36:09after all, it's evening and getting late into the night.
- 1:36:12I usually go to bed very early, but thanks to these podcasts, my sleep schedule
- 1:36:16is returning to-- I mean, it's slowly starting to be like everyone else's
- 1:36:21By the way, we've started doing live broadcasts every Wednesday at eight.
- 1:36:25That's also nice. It's interactive. Everyone can ask questions.
- 1:36:28I set it for that time because first I was doing it at six.
- 1:36:31No one was coming, of course.
- 1:36:33I did it at seven. Still no one
- 1:36:35comes. People can only get home by eight or so because of Istanbul and all that.
- 1:36:38Yes, that is right
- 1:36:40Well, I said I really need to train myself as well. I mean, just let me stay up a
- 1:36:43little bit later, man, since I usually go to sleep around nine o'clock or so. 9:00,
- 1:36:479:30, around those times. And I wake up around 3:00, 3:30 in the morning exactly
- 1:36:51It's really quite good indeed
- 1:36:52It's like, uh, you can focus so easily at those hours. There, there are
- 1:36:56no phone notifications, nothing at all. I have a bit of hyperactivity
- 1:36:58Yes, that's right
- 1:36:59Like Peggy, um, for example, I can't do what you do.
- 1:37:02Uh, I envy my wife in that regard. For example, she listens to something in the
- 1:37:05background and works at the same time. I mean,
- 1:37:07if I'm working on something, I can't even listen to-- I can't even turn on music.
- 1:37:11In fact, if you come and pass by me and ask something, I won't even notice
- 1:37:14you Of course, because once I concentrate on something, I can't look at anything
- 1:37:17else. I mean, I can't look at two things at the same time.
- 1:37:20Uh, that's why I work incredibly efficiently at those hours.
- 1:37:24Uh, but now let's slowly start moving to normal hours.
- 1:37:27Do you have anything final you'd like to say?
- 1:37:30No, thank you so much.
- 1:37:32I really appreciate it. By the way, this is
- 1:37:34actually the first time I've been a guest on another channel while having my own.
- 1:37:38So thank you for providing me with this wonderful experience.
- 1:37:42I truly hope we can have more conversations like this in the future.
- 1:37:45Yes, yes, yes, yes, yes
- 1:37:47You can um anything Is that so? Oh, the first guest, first to us
- 1:37:48Ooh well, maybe I'll come
- 1:37:50to your channel too. We can talk about marketing there in the future
- 1:37:53although it sounded like I was inviting myself like that
- 1:37:58Exactly so
- 1:38:00I was going to say that anyway.
- 1:38:02By the way, I have a thought like this.
- 1:38:05Since content is scarce now,
- 1:38:06I have game developers in my network and a director in the software field
- 1:38:11For example,
- 1:38:12I met you. Now you're one of them.
- 1:38:14I'll bring people like this to the channel
- 1:38:16How do you use AI and manage your professional life these days?
- 1:38:19We'll finally be able to ask truly insightful questions like that.
- 1:38:23I've started telling people about it slowly.
- 1:38:25They've really liked it so far. My goal is for them to hear directly
- 1:38:28from the experts Or from people who are truly dedicated to self-improvement,
- 1:38:32not from those who are just selling hype. That is really my main goal
- 1:38:36And one also needs to be careful about something.
- 1:38:38When I first started the podcast, now, for example, you can do podcasts very easily,
- 1:38:43especially with content creators, because more or less we're not professional
- 1:38:46content creators after all. I mean, we didn't get training for this.
- 1:38:50But if more or less a certain number of people are watching you,
- 1:38:53it means there are people who enjoy your talk.
- 1:38:55Now, uh, I'm looking at something
- 1:38:58For example, sometimes there's someone who's an expert in a subject,
- 1:39:01but man, he's such a boring person that he'd make you lose interest
- 1:39:04in the job. I mean, the guy knows the job very well but can't explain it.
- 1:39:08I mean, he can't explain. You'd get bored while listening
- 1:39:11Uh, for those kinds of people, I just talk on the phone beforehand
- 1:39:14if it's someone I don't know.
- 1:39:16I mean, finding an excuse because no matter how well
- 1:39:19a boring person explains a good subject, you know, there are teachers
- 1:39:22like that in school too. I mean, he's
- 1:39:24not a very good teacher, but he explains very well, you learn and all that.
- 1:39:27Yes, that's right
- 1:39:29The other one, for example,
- 1:39:31for example, is a great teacher,
- 1:39:32knows everything,
- 1:39:33but just can't explain it.
- 1:39:35So, uh,
- 1:39:37my only advice is to be careful with that. Uh, just because they know it well
- 1:39:41doesn't mean they can teach it well,
- 1:39:42uh, because this has happened to me before. Uh, and there's really
- 1:39:46a content gap in that world. I mean, you'd be filling a huge void,
- 1:39:49uh, especially on the game developer side Yeah, there's great money in the game
- 1:39:53business. I have friends in the game business too. I'll introduce you.
- 1:39:56I have friends doing mobile apps.
- 1:39:58I'll introduce you.
- 1:40:00Uh, people who know enough Turkish to be on a podcast and all
- 1:40:05And that is really great
- 1:40:08Oh, that is wonderful
- 1:40:09We'll make it happen, we'll collaborate
- 1:40:12That sounds absolutely wonderful. I will likely be
- 1:40:14extending an invitation to you quite soon, perhaps around the middle of November.
- 1:40:18You know, I usually maintain a very hectic work schedule. As I mentioned,
- 1:40:21I am a bit of a workaholic. I am certainly not stepping away from the field of AI,
- 1:40:25but I have recently received my very first invitations to visit several universities.
- 1:40:29I will be traveling there to serve as a guest speaker. I must say the young people
- 1:40:33were genuinely surprised when they had the chance to speak with me Yeah.
- 1:40:36"Oh, you're so warm, so sincere. You make jokes and all," they said. I said,
- 1:40:39"Aren't other developers like that?" "No way. They don't talk at all," they said
- 1:40:44Eh, by the way, when you say young, don't get me wrong, Yildiz is also quite young.
- 1:40:48Eh, she's almost the same age as me. Like, we're twenty two, twenty three.
- 1:40:51We're both from the same generation and all
- 1:40:55I always make myself out to be younger, but I mean, you're under 30, bless
- 1:40:58you. I mean, you're very young, too. But of course, you, you've experienced a lot.
- 1:41:02And even though you're young, you've done so many things at an early age.
- 1:41:05I think that's added value to you.
- 1:41:07I mean, uh, you've even worked as an assistant cook
- 1:41:11Eh, I mean, that teaches a person a lot.
- 1:41:13Every job you do teaches you something That's why, uh, whether
- 1:41:16you're in university or wherever,
- 1:41:17I think that's the beauty of being abroad. They kick you out of the house at 18 Uh,
- 1:41:22there's no such thing as, ah,
- 1:41:23I'm a student, I go to school, then I come home. You work, you
- 1:41:26have to. There's no money. They don't give you money. You earn your own money.
- 1:41:30I grew up like that too, by the way. I didn't take money from my family after 18.
- 1:41:34I earned all my own money.
- 1:41:35I started as a waiter at 19.
- 1:41:37I didn't just sit still as a waiter
- 1:41:40I said, "Uh, I can develop this business even further.
- 1:41:43I'm going to university at the same time."
- 1:41:46Later, they took me from waiting tables to being an assistant chef
- 1:41:49right next to the chef I learned how to make really good burgers there.
- 1:41:53I worked in specialty burger joints. After that,
- 1:41:55I started making money from painting. For a while, I started being a tattoo artist
- 1:41:59I was getting a tablet and trying to draw and so on.
- 1:42:02Later, I started getting into technical work.
- 1:42:04Uh, I started working directly as a technical officer at the university
- 1:42:09Well, that lasted a very short time.
- 1:42:11The pandemic hit and so on. Then I started software.
- 1:42:14In software, front end,
- 1:42:16now content creation along with front end. We'll see what happens in the future
- 1:42:19Well, in that case, I happen to have a job offer that I think you might like
- 1:42:23Yes, of course
- 1:42:25I have a job offer I'd like to give you. Uh,
- 1:42:26Selling donuts?
- 1:42:28So, uh, we are officially appointing you to the roles of lead developer, head chef,
- 1:42:32and primary waiter with full responsibility for all multitasking
- 1:42:35duties at the very
- 1:42:36top of the new Doner Burger restaurant chain that we are planning to launch
- 1:42:39together with Ellie. So my friends, please accept my sincere congratulations. I
- 1:42:42believe that this is a perfect and fitting way for us to conclude our video today Oh,
- 1:42:46thanks a lot to everyone listening.
- 1:42:48Uh, don't forget to subscribe to Yildiz's channel.
- 1:42:52If you've already forgotten to subscribe to my channel,
- 1:42:55I mean, ask yourself, why are you here?
- 1:42:57Why are you watching? Take a look.
- 1:42:58Turn on notifications
- 1:43:01I, I never used to say this, Yıldız.
- 1:43:03Without notifications, the stream doesn't reach folks.
- 1:43:05Please turn them on
- 1:43:07So that, that part
- 1:43:08is very important. If you don't, that's up to you folks. Take care of yourselves. Oh,
- 1:43:12thanks a lot again for joining.
- 1:43:14Uh, hope to see you in future podcasts.
- 1:43:16If there are other things you want to learn or if you want more podcasts,
- 1:43:20write in the comments, like hype it up, whatever there is.
- 1:43:23Hit all the buttons, folks. It'll happen one way or another. See you.
- 1:43:26Take very good care of yourselves
About this transcript
This page contains the full transcript of İşten ayrıldı, yapay zeka ajanlarıyla kendi işini kurdu. by Umut İLHAN ile Sonraki Adım, generated from the public captions YouTube serves with the video. The transcript has 20,538 words across 2,129 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.
What you can do with it
Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to choose a caption track available on the video.
Free YouTube transcript tool
YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, caption language selection, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.