E15: Andrew Meng

Transcript

Machine-generated (Whisper). Download: VTT · SRT

0:00 Okay, welcome to another episode of Frankly Speaking. Today we are joined by Andrew, who is the CEO of Yorbi AI. He has over 800,000 followers on TikTok and Instagram. He used to work in corporate, he became a content creator, and now he is the founder of a startup. Welcome, Andrew. Thanks for having me, Frank. Yeah, so I guess first I'd like to hear a bit more about, like, what made you jump from corporate to, like, content creation, or, like, why you even started in the first place, and then we can kind of go from there.

0:31 Yeah, so my story goes all the way back to college, actually. I studied electrical engineering in college, because growing up, I come from a family of engineers. It was mostly what I knew. So, did engineering, graduated with a degree in engineering, but even though I did graduate with a degree in engineering, I actually worked in investment banking after I graduated, because in my junior year of college, I did an internship in engineering at a defense company. And if you know anything about defense companies, it's, like, one of the most slow-moving industries out there, where I just felt the work I was doing wasn't making, it sounds cliche, but I was like, I'm not making impact.

1:11 I'm not doing anything special here. But I think that really shaped me as, like, a very ambitious 21-year-old person. I was like, holy crap, is this going to be the rest of my life? I don't want it to be this way. Looking back, it was just the company I was working at. You can get a very, very impactful, intense job in tech engineering as well. But that was what I knew at the moment. So, then I started looking into different career paths. And as an engineering major, I was like, I want to work in a job that pays me just as well as a tech role would, but also gives me some sort of, like, prestige, because I was, like, very up out there in my early, as a young male in your early 20s, you care a lot about those things.

1:52 So, the two career paths I ultimately landed on were investment banking and consulting. And how I ended up picking between the two, I was just like, which one sounds cooler? And looking back, I was just like, banking sounds cool. I watched the movie The Wolf of Wall Street, The Big Short. I was like, dude, this is definitely a cool career I want to go into. It sounds like a lot of fun. So, I was like, great, I'm going to try to break into investment banking. That's honestly a whole story in itself of how I was able to get into investment banking with, like, no experience, no financial internships or background whatsoever.

2:25 But ended up working in finance on Wall Street at Wells Fargo for two years. And, you know, people often ask me, it's like, oh, did you quit because it was, like, a very toxic work environment, like, the long hours and everything? And the answer is, like, no, not really. I think it just came to another conclusion where those two things. Number one, I never enjoyed finance, the work of finance. I was never the guy to wake up at, like, 7 a.m. on Monday to figure out what's happening on, like, the bond markets or whatever. Like, some of my coworkers did, I'm just like, you're kind of weird, dude, you know that?

3:05 But the second thing I noticed was, I took a look at my managers, people who are 30, 40, VPs, MDs. And I saw a pattern where all the people who seem very happy with their lives and their jobs, and some people are in banking, surprisingly. I didn't relate to them at all. I didn't understand why they liked what they did. I didn't understand why they enjoyed the game of finance and investment banking. But I realized, everyone that got along with and saw myself in, they all seemed very miserable doing what they were doing.

3:47 They made a lot of money, but they seemed very miserable. And I was, like, 24 at this time. So I was like, okay, I know if I stay on this path, I'm going to be probably very well off. You know, investing in banking is one of those careers where it doesn't matter what you do afterwards. You can stay in banking, go to the buy side, or, like, you know, pivot out into, like, corporate finance or something. You're going to do what in life. But I was just like, I think I'm just going to be miserable for the rest of my life doing this.

4:17 So I took a step back, and I thought about, like, what I always wanted to do. And it was two things. Number one, it was to become a content creator. And I think a huge reason for that is because I grew up during the golden age of, like, YouTube, where a bunch of big Asian YouTubers like David Tso, Ryan Higa, Kev Jumba, I grew up watching those guys. So I think I was very inspired by that. I also used to make, like, video game commentary back when I was in middle school. Like, did you play Team Fortress 2 by any chance?

4:49 No, I didn't. Do you know that game? Do you know that game? Yeah, of course. Yeah, so I used to make, like, video game commentary where I'll, like, I'll record myself playing around, and then I'll just do, like, a voiceover afterwards. So I was like, okay, I want to be a content creator, and I also want to start a business, whatever that meant at that point. So I quit my job, cold turkey. Nothing lined up. This was, like, January of 2021. And it was actually kind of funny, because a lot of my coworkers and my manager were asking me what I was doing afterwards, because they all assumed I got some offer at some private equity firm or hedge fund or whatsoever.

5:25 And I just remember, in the moment, I couldn't tell them that I was quitting my investment banking job to go make TikToks. Like, it didn't sound good. It didn't sound like it would roll off the tongue very well. So I didn't say anything. I just told them that I was taking a career break for the most part. And then right after I quit my job, the next week, I posted my first TikTok. And it got, like, 16,000 views. And the rest was kind of history from that point forward. Nice. So you kind of knew what you were planning to do. And, like, there weren't any real pivots. You just, like, thought, okay, I'm going to click this, do this thing, and I kind of took off.

6:03 I mean, hindsight's 20-20. I'm pretty sure if you asked me in the moment, like, it wasn't like that. But looking back, I'm able to tell this, like, grand story of, like, I knew exactly what I was trying to do. But it's, like, in the moment, no. I was just, like, I think I want to make content. So I just start making content. Yeah, I think that's a jump that a lot of people don't do. It's, like, they say, oh, I want to do X, Y, Z, or I want to quit my job. But, like, they don't really mean it. Like, I know people who are still at, my first job was at IBM.

6:33 And, of course, you know, you shoot the shit with your co-workers. Be like, oh, this place sucks. I'm going to quit. Like, no one actually does it. Like, no one actually, like, goes to do the thing that they say, typically, in my experience. Yeah. I learned that later on as well, where I realized most people just talk shit about their jobs because it's the thing to do, like you said. But I realized nowadays, now looking back, I'm like, oh, I had a very extreme hate for what I was doing. Like, it was, like, a deep, ingrained, like, I cannot do this for the rest of my life.

7:07 But for most people, I feel like it's just kind of like, yeah, I just want, like, better pay, better managers, better hours. For me, it's just like, dude, I don't want to, like, work in a nine-to-five. And I want to be very clear about this, actually, because I feel like there's so much rhetoric going around, like, trashing on, like, nine-to-five jobs. And I really hate that. I actually really hate that rhetoric. Because I think people want different things out of life. And just because you picked one path doesn't give you the right to say the other one was wrong.

7:39 Because they both have pros and cons to them. You work a nine-to-five. Okay. You get stable salary, stable pay. I haven't stable pay in the past five years. And also, in a nine-to-five, you get benefits, health insurance. Not to mention, you get off work at five. For most jobs, dude. Maybe not investing in banking consulting. It's a lot of jobs. You get off at five. You have your weekends to yourself. And you get PTO. PTO is a wild concept, Frank. Did you know that? Did you know that, like, every year, you can take two weeks off from work and just not think about it? That's a wild concept.

8:17 But you get that. In a nine-to-five. You don't get that when you be an entrepreneur or run your own business. Now, the upside of entrepreneurship is you can potentially make a lot of money. But then you sacrifice, like, work-life balance. And you also sacrifice, a lot of times, earning potential in your early years. Especially, like, I'm assuming a lot of your audience are probably people working in tech or corporate. I'll tell you what, nine out of ten times you quit your job to go become an entrepreneur, and you're probably not making as much as you were in your corporate job for the first couple of years.

8:48 Yeah. Yeah, I agree with that. And, like, a lot of people, especially content creators, it's hard to really make a nuanced video because you need a really, like, loud opinion or strong hook. And it really kind of, like, dumbs down the argument in a lot of cases. And if you make the video, like, too long, then no one watches it. Like, no one cares. And I've, like, kind of A-B tested this, where it's, like, if you cover all the angles, it's, like, too tame. It's, like, too reasonable. People don't care. But if you say, like, you know, nine-to-five is trash, or it's, like, oh, start up for trash, and then you go on to say whatever, it just performs a lot better.

9:22 Whereas, like, the more extreme views usually do a lot better. But, like you said, it's, like, it's all trade-offs. And it's the same for, like, architecting a software system versus, like, choosing your lifestyle. It's, like, it's not this is strictly better in every single case. Mm-hmm. Yeah. I mean, your brand for the longest time was you were the fire guy. Yeah. Like, was that intentional, or did you just happen to fall into that, and then you were just kind of like, okay, I guess this is my brand. I should talk more about it.

9:54 More fell into it because, so the first time I created content, I never thought I was going to create content. Like, I didn't really, I wasn't even really active on social media prior to five years ago. But I'm very competitive in general. And my wife, she's always been very into it. Like, she has been trying to go viral on Instagram, like, basically since I met her. But I think she had around, like, 4,000 followers or something. And then she had some, like, early success with TikTok, early in the sense of, like, maybe a couple thousand followers.

10:26 And she, like, kept on bragging about it, rubbing it in my face. And I was like, I can't take this. So I'm like, okay, you know what? In the next month, like, I'm going to surpass you. And, like, I just talked about the stuff that I knew. So I talked about, like, finance, career, tech. And those videos did really well. And I just kind of continued from there. But my original thought, because, like, tech was still super hot at the time. I think this is, like, I don't know, 2019, 2020, maybe 2021. I don't really remember the timeline anymore.

10:56 But basically, it's, like, my next job, I probably could have made, like, 600 or 700K. Because I was at staff level. And all the companies, including startups, were, like, paying a ton of money. I mean, I think they still are. It's just hard to get jobs these days. But at the time, also, I, like, didn't really want to work anymore. I mean, I think similar to you, like, I never enjoyed my job by any means. But I was, like, I was decent at it. And I made good money. But I, like, we had a, I mean, we have more money now.

11:27 But we had a lot of money, enough back then to retire. And my goal was always to, like, retire by 30. Like, I've been telling everyone around me, like, yeah, screw working, right? Like, this sucks.

11:38 And I actually had a lot of interviews, but eventually I just stopped taking interviews. Because I'm, like, okay, I, like, looked at our finances. Like, we're actually good. Like, we don't need any more money.

11:50 And then after a couple of years, like, my kids were quite young back then. So one was just born. Another one was, like, two. But now they're a bit older. So they're in school now. And I have a lot more free time. And then AI has really blown up over the past couple of years. And I just thought, like, oh, well, I'll probably regret it a lot if I, like, don't try something now. Like, I could continue to kind of just do nothing all day every day. But I feel like I would definitely regret it when I'm older. Because I really feel like now it's, like, it's, like, our version of the dot-com bubble.

12:23 Like, potentially even bigger, right? Like, this could be the last technological revolution. I mean, there might be, like, robots and whatever. But they're all kind of, like, offshoots of the current AI boom. So, I mean, to answer your question, it's, like, I never wanted to be, like, the fire guy. It's mostly just, like, that content did well. And, I don't know. I just, like you said before, I just kind of talk about whatever. It was never meant to be, like, a money-making scheme or anything. It just started off because, like, I'm a very sore loser type of guy.

12:58 Got it. Got it. Okay. Interesting. Interesting. I do have a follow-up question to that. Should I ask that now? Sure. Yeah, make sure. I'm, my, okay.

13:11 I'm very curious about your take on fire because here's my take on fire. And this is why I never understood the fire of community. Generally, the people who go after fire, from my knowledge, are highly ambitious individuals who are actually, like, borderline workaholics, from my experience. Yeah. People who want to crush it in life and work hard. So, my hot take is fire actually doesn't make a lot of sense for those people because, yes, I'm not saying, yes, it can be achieved. But once they do achieve it, I think the idea of fire, everyone talks about, I can retire.

13:50 I can do whatever I want. But I think nine out of ten times, you can tell me I'm wrong. They end up going back to their job or just find something else to, like, take up their time and stress them out again. Like, am I, am I right? Am I, am I along the lines of that? Like, that was my biggest, like, thing about fire with the people I knew who were going after. I'm just, like, you cannot not work, my guy. Yeah. I think it's, like, it goes back more to human nature than anything where it's, like, you think once you attain a goal that you're, like, done.

14:22 But no one's ever done. Like, I know a lot of people who are much, much, much wealthier than I am. And it's, like, you know, they'll kind of retire for a couple of years. But then, gradually, they, like, start another business or they maybe do some consulting. Or, like, it's really hard to sit still because it's, like, your success is kind of like a snowball. And it's, like, once you accumulate a certain amount of wealth or status or what have you, it's, like, opportunities almost just roll in just because of your prior success.

14:53 So, it's, like, I know exited founders. And it's, like, it's really easy for them to become advisors and then get some board positions or do some consulting here and there. And it's, like, their ROI is so high because it's, like, they already have this past experience. And they have the time, right? Because, and they have the money. So, it's, like, having money and, I guess, status in a way allows you to just continue to do it without too much effort. But I think it mostly goes back to, like, once you achieve a goal, there's always another goal.

15:24 It's, like, once I hit fire, I'm, like, well, what do I do now? It's, like, I'm raising my kids. And then we're quite busy for the first couple of years, like, bad sleep and everything. But it's, like, now they're in school. Like, what do I do with all my time? Just keep on making content. Like, I'm starting to get kind of bored of that. And also, I think your desires kind of go up. Before, I was, like, okay, our current house is pretty good. You know, our school district's good. But, like, it'd be pretty cool to own, like, a penthouse in China or something, right?

15:53 Or I also think that it's going to be much harder for my kids to make much money. Like, I'd be very surprised if they even go to college at this point. I think the entire education system is, like, wrecked. Whoa, okay. I mean, like, so by the time they go to college, it'll be, like, 15 plus years from now. And it's, like, it's hard to imagine a world where human-level intelligence can actually outpace the growth of AI in any capacity. Maybe, like, the very, very, very smartest gen alpha kids right now are, you know, going to be able to do something.

16:24 But, like, I kind of expect, like, there's going to be billions of robots in the next 10 years and, you know, other crazy stuff we can't even imagine.

16:34 But, like, I also do know people who have stayed fired. Like, I joined a fire community and then a lot of people, they just, but they're usually a bit older. Like, they're usually in their 40s or 50s. So, I mean, I'm still in my 30s. I feel like I have a bit more energy. What are you? 34. 34? Oh, okay. Wow. So, I'm older than most of the people on TikTok, but, like, younger than most of the fire community.

17:05 But, yeah, let's talk more about your journey of content creation. So, now you started killing it at content creation. You have a lot of followers. You're starting to get brand deals. At what point did you kind of decide, like, okay, I want to start, like, a more full-fledged tech or software business and kind of transition away from, like, purely content creation and brand deals, et cetera? Yeah. So, when I was making content, and I started making content in 2021, so it's been, rough, like, five years of me making content.

17:37 For the first two years, I made almost no money from content creation. So, I also was running a dropshipping business, like, e-commerce. That's a whole story in itself. But, mate, no money from that as well. That's a rabbit hole. But, because I was making no money, I was like, okay, I should probably figure something out. And brand deals weren't really flowing in because I was making content on TikTok. And if you know a thing about TikTok, it's very inconsistent when it comes to, like, views. So, brands are generally very scared of that because how the brand works is they're going to look at your past 10 videos, and they're going to be like, okay, which one is the lowest performing video or the lowest 25th percentile?

18:21 And they're just going to assume that's what they're going to get from you in terms of ROI of impressions. So, didn't do too many brand deals. And around middle of 2022, I was actually reached out to by a college buddy of mine, which you actually know, David Park. And he DM'd me on Instagram and was like, hey, I'm actually starting an AI startup called Jenny AI. We're really small. We're at 2K monthly recurring revenue at the moment. I'm trying to do some marketing on social media. Don't know too much about it.

18:54 You think you can help out? And I was like, yeah, sure. I can help out. I don't make any money right now. Why don't I just join your company and I will start your social media accounts for you and make content? So, I did that. I flew out to Malaysia, actually, because their entire team was in Malaysia at this point, or just Asia in general, because they want to save money instead of living in the States. It's not cheaper to live overseas. Met the team, made content for them for roughly three to four months, and was able to go pretty viral.

19:25 So, I had a couple of videos hit in the millions of impressions. And during my time there, because I was actually at the company for a pretty short amount of time, only for like four months, roughly, the company went from $2,000 in monthly recurring revenue to $35,000 in monthly recurring revenue. And all of that was driven with organic social media. And they also had like a couple of posts on like Twitter go like really, really viral. So, they never spent much money on like paid out to influence the marketing.

19:53 So, that really opened up like both my eyes and also David's eyes to like, oh, the power of like social media marketing, especially organic social media marketing. And, of course, the end of like 2022, I basically told David, I was like, hey, I want to go like explore my own thing. Wasn't really trying to like work at a startup at this point. I still wanted to like, you know, run my own business. And I was like, okay, well, how about this? You're obviously very good at this whole social media thing. I have a lot of connections within the startup ecosystem.

20:24 Why don't we just partner up? We can start a marketing agency. And we can basically help other startups go viral on social media. And I was like, sure, why not? That sounds like a good idea. So, we ended up doing that. We began onboarding clients. And, you know, basically ran the agency for the next three years. And was able to work with some pretty cool brands along the way. Like we, I don't know if you're big on Twitter, indie hacking Twitter, but we worked with PDF AI, which is like one of Damon Chen's AI company.

20:57 We ended up working on Coppin's Bagel, the dating app, which we still work with to this day. Actually, they're like one of like the longest standing clients I've worked with. So, I was able to work with a variety of customers and clients. Most of them within like the software and AI space, obviously. And around the middle of last year, and okay, I don't want to get the timeline too mixed up here because things are like overlapping a lot right now. But I started the marketing agency in the beginning of 2023, okay?

21:31 In the beginning of 2024, I partnered up with my current co-founder, not David, my current co-founder, Thomas. Because Thomas is a software engineer and we began working on different software apps together. And we began launching a couple like micro SaaS app. We launched like an AI interviewing tool. We launched a B2B organic database. And then we launched like an AI meeting recorder. But then in 2025, we're like, okay, let's get very serious about what we're building here. And let's see like what's actually like an idea we can turn into something bigger.

22:05 And the idea at that point was Yorbi, which we did pivot around a little bit. So, this is like the third iteration of Yorbi. But what Yorbi is, is accumulation of all of our knowledge of social media from both making content as a content creator in the past five years. My co-founder, Thomas, is also a content creator. He has like 300,000, 400,000 followers on social media as well. So, it's accumulation of our knowledge of content creation and also my knowledge of running my marketing agency where I worked with a handful of clients to help them go viral on social media.

22:39 And we put that all into an AI software that basically the goal of it is to help creators and marketers go viral on social media. As of today, the tool, mostly what it is right now, is the AI content script writer. So, we can upload any viral video from TikTok or Instagram. Like, you can take a frank new video that you really like and be like, okay, why did this video go viral? What elements about it made it like good for the audience and algorithm? And then you can tell the AI about the topic that you want your video to be about.

23:10 You know, basically take the structure, the format of that video and generate an entire new content script for your topic, audience, or niche. Very cool. I guess my first kind of follow-up question is, what made you decide to drop the previous ideas? Because you mentioned you went through several of them. Like, at what point did you think, like, okay, this is not good enough, moved on to the next one. And then what made the latest iteration, like, stick? Like, what made you think, okay, this is the one we should focus more on and continue on with compared to previous ideas?

23:45 Yeah, I think there's both, like, a very personal, this is, like, a very personal question and also, like, a very, like, tactical question as well. Or, sorry, answer. The tactical answer to this is, like, we didn't see, like, the results you want to see in the first couple of months of launching these different software products. That we just ended up just pivoting away, like, way too fast with each of these software. So, actually, like, looking back, I think if we just stuck to something, it probably would have eventually worked.

24:22 But I think we were so drunk on the idea of, like, things need to happen fast, especially with, like, AI SaaS. Because all you see on social media is, like, how I got to a million ARR in 30 days. You know, and you're like, oh, shit, this is, like, how it's supposed to happen. I'm supposed to hit a million ARR in 30 days. And if I don't, then it means the product was a failure. And I think that's a lesson I learned, like, pretty recently, actually. It's, like, no, just pick one thing, stick to it, don't just work out for three months.

24:50 You should probably wait until, like, wait until at least six months or something, I don't know, before you kind of take a look at it and be like, okay, this is something you need to pivot away from. But I think, like, we pivoted way too fast because we didn't see results. The more personal reason is some of the tools we're making, I don't want to say I have to be passionate about it, but I really didn't like some of them. For example, we made an AI mock interviewing tool to help people land jobs. I really didn't enjoy building that product because I think a huge part of me is I just don't care about the job search industry.

25:27 I'm not passionate about job search. I'm not saying you have to be passionate about what you're building, but you also shouldn't not want to think about it or do it. And I just didn't care about, like, how I can build the best AI mock interview tool. It just didn't excite me. Well, I'm thinking about, like, okay, how can I help creators and marketers make better content? Look, Frank, do I go to bed at night being like, yo, this is, like, my life mission and goal? Well, if I'm an investor call, then yes, it is what I say.

25:57 But it's not. But I get excited about it. I do enjoy thinking about social media. I do enjoy thinking about why videos perform well. And I do enjoy building a product that, oh, my God, this is going to sound like such a, like, I feel like I'm in, like, an MSNBC interview right now. I like serving my customers. It brings me a lot of joy when I hop on customer calls. And they tell me how the product has impacted their workflow and helped their content. And it brings me a lot of joy.

26:30 Yeah. I think that all makes sense. It's, like, you can't work on something that you don't care about. And if you, like, Annie is, like, getting some stuff behind me right now. No worries. I can't see her. But, yeah, I think what you mentioned earlier about, like, if you don't care about it, it's really hard to work on it every day. And I think that's one kind of underrated thing that people don't really talk about. Like, it's actually not just about the money because it's, like, sure, you can try to force yourself to work on something every day.

26:58 But then it's, like, the product suffers. And it's, like, if you're not thinking about it, if you're not making it the best it can be, like, you're just hurting everyone all around. Like, your customers are not going to get the best product. You're not going to enjoy working on it. And then you'll probably get burnt out a lot faster. Yeah. I know you also pivoted around a little bit. Like, what was your experience and how did you land on, like, your language learning app? I mean, it's pretty similar to you.

27:24 Of course, like, I was working on a couple different ideas solo, but none of them were that serious. They were more just ideas I thought, like, this could probably work out. Like, I see this product that does pretty well. I could probably do it better. Like, it would probably make money. And the train of thought was more, like, mercenary in the terms of, like, I think this is a decent idea. I think I'm a smart person. I think I could probably out-execute some of the competitors that I see. I mean, some of the ideas fail just because the margins weren't really there and the current competitors in the space are, like, just burning VC money.

27:58 So one idea was, like, I mean, a lot of ideas are centered around my kids because I'm around my kids a lot. It was, like, just shoving an LLM inside of a toy. And there's kind of two competitors in that space. One is, like, heavily VC-funded. One's, like, more scrappy. It's, like, a tiny startup. I think they have some funding. But, like, the unit economics are just terrible. Wait, wait, wait. Wait, what was this product? You just, like, shove an LLM inside of a toy. Because I noticed that, like, my kids...

28:25 So, like, what would it do? So your kids, I mean, once you have kids or if you're around little kids, it's, like, their follow-up question is always why. And it comes to a certain point where, like, you can't really explain it any more that, you know, they're just asking why because they're kids, right? But, like, an LLM is infinitely patient. It can, like, tell infinite stories and teach them stuff that you realistically don't even know as well at this point. And, like, I noticed my kids really enjoyed talking with, like, the voice feature in ChatGPT when that came out.

28:57 But then as far as, like, the terrorists went up a ton on China. Sourcing the parts and, like, building the thing per unit is roughly, like, $100. And it's, like, a lot of the competitors, they started off charging, like, $99. And they would raise their price to, like, $150. And then, you know, they try to make money on the subscription after the fact. So it's basically similar to, like, video game console model. You lose money on every hardware piece you sell. And you try to make it back with subscriptions.

29:26 And basically, I'm, like, man, like, maybe this will work eventually. But, like, this just seems like a bad business. It's, like, I'm not going to say as bad as, like, running a restaurant or something, which is, like, the worst business to run. But in terms of, like, pure numbers, it's, like, okay, I'll have to raise a bunch of money first. And then try to, like, scale. It was, like, a very, very, very long play that I felt like the payoff just wouldn't be worth it. And, like, my kids enjoy doing other stuff.

29:54 Like, they're in sports. You know, they read books. They play with Legos. It's, like, this thing is not, like, absolutely necessary. And then I also noticed that, like, I actually got sent free samples from some of my, from some of these other companies that already made the product. And they would get bored after, like, a couple days. It was just, like, it's missing a lot of features because, like, they're, like, trying to show these stuffed animal stuff. It's, like, can you see it? And it doesn't have any, like, touch input.

30:20 They're, like, punching and kicking it. Like, hiding underneath a blanket, stuff like that. So that was one that would, like, I love just because it's, like, not a great business. Not a great bootstrap business. You're going to have to, like, raise capital and, like, go bigger with it and stuff. Just, like, ah, it's not, like, a one-man show that can be running this. Yeah. And then I had a couple other ideas that I also pivoted away from that I don't even really remember. But then when I started to take it a bit more serious was when I asked this guy I met through boxing back when boxing was a thing.

30:56 I think I met him back when Critter Clash 1 was just getting started because that was when I was supposed to do that before I got cut because I wasn't popular. Oh, yeah. What was that Critter Clash thing? It's, like, creators for going to box each other. Yeah. So, like, iDubbbz. When I was, like, really big. Yeah. iDubbbz was the guy running it. He was huge back in the, like, YouTube. Similar to, like, the Niga Higa Kevjumba era. He was, like, the edge lord. Yeah. And he reached out to me, and I was supposed to box this guy named Leon, who's, like, a Pokemon YouTuber who, like, does Pokemon unboxings.

31:31 But he replaced me with this other YouTuber guy who's, like, much more popular. He has, like, 2 million subs on YouTube. I don't know who he is, but, you know, he's better than me. And that's really unfortunate because I would have destroyed the Leon guy. Like, that guy got creamed, and I'm pretty unfortunate that I did get to box him. But I did meet my co-founder there. I didn't actually ask him to join me until last year or sometime, though. Like, I met him two or three years ago at this point. But we mostly were just hanging out, like, playing tennis, board games, whatever.

32:04 And I knew he graduated from Waterloo. I knew he used to work at Google. And then I was thinking more, like, okay, all the advice is, like, go find a co-founder, go find a co-founder. And I felt like every day just, like, looking at a computer by myself was, like, not that productive. Because it's, like, you don't really have anyone to bounce ideas off of. It gets kind of boring, I want to say, a lot of the time. So I asked him, like, hey, do you want to, like, work on something? And then we came up with a bunch of ideas.

32:34 But, like, something that, so he's a white guy with a Chinese girlfriend. I think, I mean, I guess they're technically considered, like, common law. But he had already been learning Chinese by himself for the past couple of years. And I have to, you know, manually teach my kids Chinese every day. Because we're in Canada. And otherwise, they don't learn Chinese at all. So we're thinking, like, okay, here's, like, a personal problem that, you know, we both care about a bit. And it's something that with, like, the latest advancements in speech tech, whereas, like, before you speech-to-text, text-to-speech, all this stuff, like, just didn't really work that well.

33:10 And even now, it's, like, not perfect, but it's a lot better. And in this case, I kind of checked all the boxes where it's, like, here's a new technology. The unit economics are really good. It can be a SaaS product. It's something that we both, like, care about and that we would probably use in our day-to-day lives. So we just kind of ran with it. And, you know, we're still working on it. We're actually pretty much at the starting point where you mentioned where David first reached out to you. Like, we're at, like, 2K MRR now.

33:38 So, you know, just starting out. But, like you said, I think it's more, it will work out. Just continue to make the product better and keep going. Yeah, yeah. So are you guys starting with just Mandarin and eventually branch out to, like, other languages, I'm assuming? Yep. So starting with Mandarin, then eventually other languages, then eventually other subjects. So, like, math, sciences. And then kind of the end goal is that it'll be, like, a more generic tutor. Because right now I think the vision models are still kind of crappy slash very expensive.

34:09 But, like, if you use, like, a tutoring app and you have, like, a real tutor that's digital, you know, they're able to see your screen and they're able to see you and they're able to do a bit more. I think that's kind of, like, the end goal. But, you know, that's quite a ways away. Got it. Got it. Is he also technical, your co-founder? Yeah. So we're both technical. I'm probably going to take care of most of the marketing stuff, though. I mean, he's a better coder than I am, for sure. I mean, with, like, Cloud Code, it kind of levels the playing field a lot.

34:42 But I feel like when I consider someone, like, who's really good, I never considered myself a truly exceptional, like, programmer or by any means. Okay. And I consider myself, like, a smart person, but, like, not the best, like, coder. And then, so he's mainly taking care of the technical stuff. Like, I'll still do bug fixes and some, like, small features here and there. But, like, he's mostly taking care of the product. Got it. Got it. Do you think you have that perception because you worked at, like, very top companies in the past?

35:22 Surprisingly, no. I feel like, funny enough, most of the top companies I worked at, people were quite average. Like, it's, like, such a myth that, like, thing engineers are that much better. Like, the best engineers I worked with, like, you would not expect it from their resume at all. There'd be, like, some random local college, you know, kind of, like, no-name companies. And then, but then when you work with them, it's, like, it's really, really obvious who the best people are. And they just get, like, I mean, these days it's hard to measure because, like, with cloud code, everyone can, like, pump out insane amounts of code.

35:56 But, like, prior to the days of AI, like, you would meet people who would literally get 10 to, like, 100 times more work done. Like, they're just so efficient. Like, they knew the syntax so well. And they're, like, very solid fundamentals. And also, just, like, speed of execution. Which, I think speed of execution, that has leveled out a bunch recently. But as far as, like, knowing what to build and architecture and stuff, you still kind of have to know a bit of that. And AI is not perfect. Like, there's still bugs you still have to manually debug.

36:28 But, like, yeah, I think the outward perception is, like, thing companies have the best engineers is a myth. Like, of course, there are pretty good ones there. But, like, funny enough, the best engineers I worked with were at, like, smaller companies. And you wouldn't know it by their resume. Yeah. I think, like, a good comparison in that is when I was studying engineering as well. I always find it very interesting. Number one, okay, I'm not saying to show off, but I graduated cum laude as a W grad. So, I was, like, top of my class.

36:58 I would not consider myself, by any means, like, the best engineer I knew in college. While I knew so many people who had, like, 2.0s, 2.8 GPAs, that were just phenomenal engineers. Because instead of studying equations, they were actually just building stuff. Yeah, exactly. And understanding how engineering worked. Well, hey, man, you want me to write forward transforms? I could do that for you. What does it do? I don't know. Yeah, I have no idea. But I know something, something, something Delta function, right?

37:31 It's, like, I was very good at school. But when it came to, like, practical stuff, I realized all the best engineers were the ones that are, like, working on, like, IEEE projects or something like that. They're the hackers. They're the ones who are, like, building projects on the weekends for fun. It's, like, I was never one of those dudes. Like, I mean, similar to you. I got good grades and everything. And, like, I can do the work. But it's not, like, I'm spending my free time coding up projects. It's, like, so one of those dudes, when Pokemon Go first came out, he, like, reverse engineered the Pokemon Go API.

38:01 And he built a live map that showed all the Pokemon and, like, when they would spawn the timers and everything. He did that in, like, a day. And he's just, like, yeah, whatever. I just did this, you know, after work in 30 minutes. And it's, like, sure, today you could probably spin that up yourself with AI. But, like, this is pre-AI days when, like, you actually didn't know your shit. But, yeah, exact same experience. It's, like, the people who are more hands-on are, like, it doesn't translate to your resume.

38:30 So I got a question for you, Frank. How does it feel to be the villain that everyone blames for saturating the CS market? I think, I mean, I actually do feel a little bad because when I was promoting it, like, it was the right thing to do. And, obviously, it did not pan out because I used to, I truly did think that before, everyone had a chance. Like, everyone could do it. And I don't think that anymore because, like, even two months ago, people had a better chance than they do now. Like, the models that came out in December, like, starting with Opus 4.5, it, like, it literally made, like, not even just juniors, but I would say, like, half of all engineers almost at every level.

39:16 Because, like, when you actually work at these companies, most people kind of suck, honestly. Like, they're not that good. But then I think there's, like, if you're good, there has never been a better time than today. But that's the asterisk. It's, like, most people are not good. So, like, most of the audience who followed that advice, like, there are some who are wildly successful, like, way more successful than me at this point. Because I first started creating content, like, four or five years ago. Some of them have, like, made companies, exited, and are, like, stupidly wealthy now in that time.

39:49 Like, before, they didn't even know how to code at all. So, it's, like, the top 1%, they are, like, killing it. They're doing way better than ever before. But I think, like, the bottom half of people who used to be fine, like, prior to COVID, like, all these companies just hire, like, everyone. They needed test engineers, QA engineers, they needed drones to, like, do things. Yeah. Do all this, like, BS work, and you kind of slowly gain experience. But now it's, like, companies don't need that anymore. Not really.

40:18 Why would you get, like, three juniors when you could just have one senior? Who's going to do better work and probably do it faster, too? And I think it's only going to get worse. Because now even seniors are starting to get afraid. They're, like, oh, shit. Like, AI is pretty good now. Like, and again, this was not even true three months ago. Because, like, I've been using AI this whole time. And in the beginning, it was quite bad. Like, you couldn't really get stuff to work. But now it's, like, really, really good.

40:45 It can get you, like, 90%, 95% of the way there. And maybe the last 5% you need to manually fix. But, like, to answer your question, it's, like, I think a lot of this was unavoidable. Whereas, like, before it was obviously the right path. And I think even now it is the right path if you are good. And I hate the word agency because I feel like it's overloaded. But it's, like, if you're a go-getter, like, if you're a killer, like, you're still going to be successful regardless of AI is there or not. Because you're going to leverage it.

41:14 You're going to build a business. You're going to sell a business. You're going to be getting your postdoc in, like, machine learning. You're going to be making millions at OpenAI Anthropic. Like, if you're among the best, then you will be paid. And, like, very, very handsomely. Like, there are people my age who are, like, VPs at OpenAI. They're, like, you know, making $100 million pay packages. But, like, the delta between, like, the very top, like, if you're top 0.01% versus 0.1% versus 1% versus 10%, like, each one is, like, a step function of, like, how much you make.

41:45 But, like, unfortunately, most people are, like, you know, their average or their below average. In which case, like, not even just for comp sci. I was kind of, like, taking a look at all white-collar jobs. Like, everyone's in danger. Like, accounting, lawyers, as long as you have some sort of work that deals with a computer and text and eventually even voice. Eventually, even vision. Like, graphic design, every job is really at risk. So, it's, yeah, I don't know. Like, I think I, obviously, looking back, like you said, it's, like, kind of 2020 now.

42:17 Like, looking back, it was not the right recommendation because I did not anticipate AI advancing so quickly. I thought, like, oh, you know, we'd have, like, five more years before we get to this point. And I also thought, like, I truly believe before, like, it was the best way to make money, even if you were kind of dumb. Because I knew plenty of dumb people in engineering and who would make a decent amount. But, like, those people are screwed nowadays. Like, there's just no shot for them. Like, why would you hire these people now when you can hire a smart person that will do, you know, not just 2x, but, like, 10x, literally 10x the amount of work compared to, like, 10 dumb people.

42:54 So I think it's, like, for my audience, there's a lot of people who are very grateful. But I also think that those same people, they probably would have been successful regardless. Like, if they didn't study comp sci, they could have studied something else, they probably would have done well. And it's all the people who, you know, I don't want to say dumb, but, like, they're, that are not top, like, 30. It's hard to put exact percentage on it because it's unclear to me, like, how much companies actually know what's going on.

43:22 Because, you know, some companies might limit you to, like, let's say, like, a crappy model. Like, they don't let you use cloud code or maybe they only let you use, like, GitHub Copilot or, like, one of these shitty AIs, right? Then they don't realize, like, how powerful it is. But as far as, like, I think the bifurcation and, like, the term people are using is, like, permanent underclass. Like, I think that's a kind of, like, an exaggerated term. But I definitely think there's a hint of truth to it where I kind of spoke to it earlier where, like, I just don't really accept my kids to amount to much, honestly.

43:57 Because, like, what the hell are they going to do in the future? It's, like, it's up to, it's literally up to our generation. If you're, like, working age now, you need to make your money now. Like, this is, like, the greatest technological change since, like, the internet. Like, possibly more so. This is, like, the industrial revolution of our lifetime. So it's, like, and most rich people that I've talked to pretty much agree. It's, like, if you don't have assets now, like, you better get some assets. Or else you're just screwed.

44:29 Wait, so, okay, number one. Cat, clip that. Frank New officially apologizes. Live today.

44:38 But, okay. So you also, so you do believe in some sort of, like, AI doomsday. Where you lean more towards, like, AI. Society is kind of hooked with AI. I guess it depends. Because, like, I think we'll have an abundance of everything. But it's really hard to, like, see the people who control all their resources. Like, how generous are they willing to be? Because, like, even 10, 15 years ago, people wrote books saying, like, you know, assets outpace earnings potential in, like, a salaried role. And that's always been the case.

45:19 But that's only, like, that gap is getting bigger and bigger. It's, like, the amount of assets that top 0.1% controls, like, it's growing exponentially. Whereas, like, middle class and below, like, they're kept on getting pushed down in terms of, like, the amount of wealth they have. And no, like, people, in some cases, can't even afford groceries, right? It's, like, inflation is just crazy high in Western countries. And so, you know, ideal case is, like, we get some form of UBI. You know, we have all these robots who are building houses for everyone.

45:51 Everyone has food because, you know, robots are harvesting the food and we're distributing the food. But then it's, like, the opposite case is where, sure, you have the ability to do all these things. But then the people who actually control the compute or the farmland, you know, they're hoarding it for themselves. And it's, like, I don't know if you've seen the movie Elysium. But it's, like, you know, an exaggerated form of this where, like, the rich people, they literally live on, like, this cloud city where everything is perfect.

46:17 They've solved hunger, disease. It's just, you know, perfect everything. But then most people live on Earth. It's, like, super polluted. You don't really have any resources. You're doing, like, menial work. Or even, like, that. It's like Wall-E. Um, Wall-E's kind of like that. Except I think Wall-E, they, um, it's more of a paradise. Just, like, everyone has robots, you know, waiting on them. But, um, I think the other, have you seen the Black Mirror episode where everyone's just, like, on the bikes generating electricity for credits?

46:51 And that's... Yes, yes. First one I've seen, actually. That was the first Black Mirror episode. I was like, whoa. This is a cool show. That could be the future. Like, people are just, like, so useless that the only, like, benefit that they can provide humanity is, like, generating electricity. It's like the Rick and Morty episode where, like, you're in a pocket universe. You just do the step pedals because that's all you're good for. Like... Yes. Um, and it's, like, it's not too hard to imagine a future where that's true.

47:18 Where, like, okay, now we have robots that are smarter than everyone. Like, frankly speaking, robots are probably... The AIs of today are already smarter. If you aggregate their, like, intelligence across all these subjects, it's, like, it's very hard-pressed to meet any, like, individual person who has as much general knowledge. Like, sure, you might have more specific knowledge in, like, one area. Maybe one, even, if you're good. Um, so it's, like, okay, it's gonna be more intelligent. And then they haven't figured out robots yet.

47:46 But, like, if you look at robots today versus even last year. Like, last year, they're kind of, like, awkward and, you know, they fall down. Now they're, like, doing flips, like, doing kung fu, this, like, coordinated dance. Like, it's crazy, the rate of advancement. And I think it's only gonna get crazier in the future, right? So, I wouldn't say I'm a doomer. Like, there's definitely a scenario where it's, like, yeah, we have an abundance of everything. Everything's cheap. People can just focus on whatever.

48:12 But then, like, there's a other, I feel like, more likely scenario where it's, like, sure, we'll have all these things. But the people who own the resources, like, who own the farmland, like Bill Gates, he's the single largest farmland owner in the United States, like, private individual. He owns a crap ton of farmland. Or the people who own the people who control compute, like, the people who are working on these models, you know, like, the anthropic open AI. I actually think, like, the highest ED play is probably to work at one of those two companies, honestly.

48:46 Like, like, given the failure mode for, like, startups, it's, like, is more unlikely for you to, like, out-earn just how much those companies grow. So, just, like, you know, they're determining the future of humanity, effectively, at this point. Can you, so my, so I think about this a lot, and my question is, like, can't you make the same argument in the 1920s or late 1800s when the Industrial Revolution was happening? And everyone's, like, oh, yeah, robots and machineries are going to take over, like, factories.

49:22 We don't need humans to operate this machinery anymore. Is that the 1920s? I don't know. But you know what I'm talking about? When robots first took over, and then that kind of sprouted the entire industry of, like, software engineers. Like, software engineers did not exist back then. So, is there, like, a new industry that comes out of this? Granted, you know, there's going to be some industry that just naturally die out, and that's just how it goes. But it's, like, I don't know. You can make the argument, like, oh, prompt engineers are going to come out of this.

49:52 Or everyone talks about this thing, like, the tastemakers. I don't know what that means exactly. But I think about this a lot, where I'm just, like, well, it can't all be, like, doomsday. Like, human beings, we're going to find problems with our lives that we're unsatisfied with. Like, you can't AI automate away in-person events. Like, there's going to be, YC is going to invest in some, like, fucking immunity company that's going to host the biggest in-person event in the world in, like, 10 years or something.

50:24 You know, what's kind of funny is that, you remember the WeWork founder? Yeah, Adam Neumann. So he started a new company that's more focused on, like, kind of in-person living type thing. It's, like, a real, it's still another real estate play, but it's more like, you know, he might get some more funding now that you kind of mentioned this thread. But I think it's, like, the key difference, and I hear that, like, counter-argument a lot, where it's, like, hey, you know, this happened many times before with the Industrial Revolution, with, like, you know, calculators and writing and et cetera, et cetera.

50:54 But I think the key difference is, like, those were not straight-up replacements. Like, those were still tools, in a way, where, like, they required some sort of supervision. Whereas, like, we are literally building something better than ourselves. It's, like, at all aspects of intelligence, right? Which is why it's, like, people are like, oh, you know, go to the trades. That's going to save you. Like, no, it's not. Are you kidding me? Like, that's such a joke. Like, yeah, I just, I don't know what people are thinking.

51:20 I mean, I guess their thought is, like, oh, you know, if white-collar is screwed, we're going to blue-collar. It's, like, no, once white-collar gets screwed very rapidly, like, you think they won't be able to create a machine that, like, does your laundry or, like, you know, cooks food? It's, like, those are very rapidly gone next. And those are already lower paid. So, it's, like, I can pay $15,000 for this robot to be my maid or I can employ a person for, like, I don't know, $50,000 within nanny. But the robot is cheaper to manufacture on a per-unit basis than cars.

51:53 Like, a car has more material inside of it. And the entire reason why, like, Tesla is not tanking is they're, like, oh, you know, even though our car sales suck now, we're going to be a robot company. So, you know, you should invest in us. Like, we're going to make a billion robots because that's what we're going to do because they are now getting wrecked by BYD on the electric car market. So, it's, like, I just don't see it. Like, I can't envision all these jobs being created because, like, one person is able to do so much more by themselves now.

52:22 Like, before you needed maybe, like, a team of three or four engineers over the course of several months to build, like, an MVP. Now you can do it in a weekend or, like, an afternoon with one person and, like, one cloud code subscription. So, it's, like, I just, it's hard. I think a lot of the AI bosses, some of them are more upfront than others. Like, the Anthropics CEO, he's, like, yeah, white collar's fucked. Like, they're going to lose, like, probably 30% to 40% of all workers, like, at least junior workers in the next three to five years.

52:52 And he's, like, telling it like it is. It's, like, yeah, this is, you know, Urbano's will be smarter and better. Whereas, like, other people, I think they probably think that privately, but publicly, they don't want to, like, make it too doomsday. Like, they're, like, oh, you know, there'll be new jobs. You can be, like, a space data center operator. It's, like, why the hell would you have a person doing that over a robot? Like, it doesn't make any sense. Yeah. But, yeah, I don't know. I think that we are at, like, the perfect age in terms of, like, well, basically anyone who's, like, 16 and above, like, you have access to tools now that no one in the previous, like, humanity never had access to these tools.

53:37 Like, now is the time to make something. And if you can't make something, try to get into, like, opening AI or Anthropic because they're the two juggernauts right now. Yeah, yeah. Or start a personal brand and use Yorbi AI to grow your personal brand. Yeah, there you go. Make viral content. You want to market your startup, check out Yorbi. Yeah. Okay, okay. Interesting. Oh, but, yeah, talking more about Yorbi. So, presumably, you dogfood your own content to create content. I know you recently hired someone to take over a lot of the marketing.

54:12 Could you talk about, like, what avenues you used for marketing? Like, I'm sure you did organic. Did you, like, look at UGC? Did you look at paid ads? Can you kind of go through the process of, like, how you started, well, one, deciding, like, Yorbi was the play to pursue more thoroughly? Was there anything more than just, like, okay, we both have a background. We both are pretty good at it. And then also, too, the various marketing strategies you used to grow the company. Okay. Let me see if I can explain this in a very coherent way.

54:45 So, last year in 2025, around August, we got into Startup Accelerator, Launch Accelerator, which is Jason Calacanis' Startup Accelerator. And we were building Yorbi, to answer your first question, because we both had a background in it. So, it just naturally made sense for us to go down our product within the social media realm. So, we launched it in mid-November of last year. But I'd say December, because even after we launched it, nothing happened for, like, two weeks. We're still, like, pushing our products.

55:19 Like, our Hero product didn't even come out until, like, December. But since December, we launched our Hero product, which is the content remixture, the one I talked about earlier, where you can remix any viral video and get, like, a viral optimized script. And we began marketing it on social media. And actually, I used AI to market Yorbi, like, an AI-created avatar, which leads to another funny conversation we can talk about in a bit about how AI is going to change content creation. But marketed that with my AI avatar for pretty much the past, like, three months, was able to go pretty viral and grew to over 20,000 users in that time span.

56:00 So, because of that, we only have done organic social media marketing up until this point. We, at least, I personally haven't promoted Yorbi to my own personal brand. Like, yes, I told a story about our user growth, but I never marketed it to my personal brand. Mostly because, like, my personal brand is, like, within the corporate startup ecosystem. So, it's not as relevant to creators. But we are building out more, like, U2B aspects to it. So, it will be eventually. That's down the line. So, all the growth has come from, like, pretty much organic social media marketing from, like, two accounts.

56:35 Mostly AI. And the reason why I am now going a lot harder on social media marketing, I made a LinkedIn post about this the other day, where I'm going to be spending 60K in the next three months to try to make our brand go viral, as viral as possible, is for two reasons. Well, number one, I'm going through a fundraising round right now. So, as I'm fundraising, I want to be able to show how much traction as I can. But the second thing is, we are already seeing validation with social media content. So, we know we can go viral on social media.

57:08 So, all right, great. This is a validated channel. I think it's time to, like, pour gasoline or fuel on a fire that we kind of have created. So, I made a post last week about, like I said, I'm spending $60,000 in the next three months to market my product. But the point of that post was to try to hire someone to come onto my team to basically, you know, manage all the marketing because I'm going to be very busy with fundraising and utter CEO duties, like, I don't know, taxes or something. But I end up hiring two individuals.

57:42 Well, I mean, two entities. One individual, one entity. So, I onboarded a person. He's in college right now. He's working for us part-time. And I also onboarded a marketing agency. And basically running both in parallel. Initially, I was just going to do one or the other. I wanted to lean more towards hiring, bringing someone in-house. But I ultimately opted for both just because I want to see how they both run these campaigns and maybe just, like, learnings that both of them can, you know, cross-reference from one another.

58:18 Because they're both, like, both the agency, also in Visual Onboarded, are both very experienced with, like, running campaigns for, like, other brands already. Cool. Yeah, let's dig more into, you talked earlier, how you did all of it with AI Influencer and how it's all organic. Can you feel more in-depth into, I guess, what made it take off? Or was it just kind of, like, viral from the start? So, the account that I created, at first I was making content on it, like me. I was making videos on it. Eventually, I don't know why I started using AI Avatar.

58:56 I think I just got curious or I got lazy. But essentially, I was just, like, you know, this Sora launched in, like, September of last year, I think, Sora 2. And it was, like, one of, okay, when Sora 2 launched, that was the moment where I was, like, oh, yes, AI beta generation is going to take over. Before that, I was, like, a non-believer in AI beta gen. I was, like, this is never going to be better than human creators. The moment Sora came out, my beliefs switched in that moment. I was, like, this is the future.

59:27 So, I was always very curious about the potential of, like, AI video generation, especially when it comes in the concept of marketing a product. Because my background is running my agency where I'll go out, I'll hire real human creators to make content for these brands. And you work with these creators, you'll train them, you'll show them how to make good content, how to have a good hook, all that stuff. Very manual, very human, very exhausting. You're dealing with people. You're basically running, like, a services business at that point.

59:58 So, I was, like, okay, can I do this with AI and achieve the same results? Is it going to lead to conversions? Is it going to lead to lower trust? All that stuff. And the conclusion was it's the same. It's almost the same. Some will argue it's better because, like, the cost is, like, pretty much no cost. It's free. Like, Sora 2 is free to generate content. So, what I do is I go into Sora, I generate a piece of content with, like, an AI avatar, and I use that three-second video clip as, like, the hook of the video.

1:00:29 And then with that hook, I also have, like, a text over the screen that basically outlines whether that's a dream outcome or a deep pain point that my target audiences face. And then after that initial hook, I'll have, like, a demo portion of how my tool works and how my tool solves the pain point or achieves the dream outcome that I promised in the hook. I basically repeated that video two to three times a day with, like, just different hooks, different text hooks, different generate AI avatar. Exact same demo.

1:01:00 Like, I didn't change anything at all. I just took the same video clip, ripped that for, like, two, three times a day for, like, two weeks in a row. And then some videos just started getting over, like, 100K, 200K, 300K views and impressions. It has since then kind of died out a little bit. Just, like, eventually you kind of burned through an account and you also burned through a format. But that was, like, really, really effective. And I kind of proved to myself, like, yes, you can go viral with AI-generated content.

1:01:25 It's not fully AI-generated. Like I said, I personally filmed the demo portion of the video, but, like, I'm not in it. But, like, the hook and everything that made every single video different, which, granted, is probably, like, the more important part, that part is all AI-generated. Now, I will say this. The question now is, like, okay, you were so successful with AI-generation. I don't know if I mentioned this earlier, but now I'm working with human creators again. Which, like, this is just a wild thing for me to say right now.

1:01:54 It's like, oh, I'm working with humans, not robots, you know? Let me tell you why I'm working with humans again. But I am hiring human UGC creators again. And the obvious question is, like, why? You were so successful with AI-generation. For two reasons. Number one, AI still has limitations of what it can and can't do. Like, you just don't have as much flexibility in terms of, like, the type of content you want to be producing. Like, you can tell humans, in theory, to make anything you want them to do within reason.

1:02:22 But with AI, you have to prompt it, right? And even when you prompt it, there's certain things that, like, AI just struggles to do with video generation to this day. The second reason is because these platforms, like Meta, TikTok, they have teams out there trying to stop you. And let me explain. Let me be very clear right now. These platforms are okay with AI content. They are. They are actually, if anything, I'll say Meta is very AI positive because they have Lama, right? They have to be friendly to AI-generate content.

1:02:58 TikTok, not so much. But even then, they still allow it. What these platforms don't want is spam and content farms. And the problem of AI-generate content right now is in order to scale, you need to start renting out phones or you need to start renting out, like, servers or whatever, right? To basically, like, open more accounts. And the problem with that is if you open up 50 accounts, let's just say, like, behind me, I have a wall of, like, 50 phones. If Meta sees 50 phones, each with seven accounts on them, posting on across, I don't know what that meant, 350 accounts, all posting three content a day, they're going back.

1:03:39 If this IP address is a content form, let's ban it. And I witnessed that firsthand because on my own phone that I have here, I had, like, three accounts I was running that were just fully AI-generated. And it was fine for a bit. But then one day I woke up and I was trying to post a piece of content. Two of my accounts I couldn't post anymore. Wow. No warning. No whatsoever. I just tried to post a piece of content to Instagram and it just says, your account, like, doesn't follow some community guidelines.

1:04:13 I checked the account status, everything. It was perfectly fine. Didn't matter. So, was I banned? I wasn't banned. But I think that was, like, a warning sign from Zuck being, like, hey, hey, hey, hey, hey. I know what you're doing. Don't do that. Don't do that. My account has been spend memory stores. I'm posting again. I'm a lot more careful of it now. Because before, I was, like, just spamming three bills a day. Like, literally, like, the same hook, same body, everything. Now I'm trying to switch things up a little bit, try and make the account more, like, human-like.

1:04:43 Almost to say. But I know these platforms have teams out there trying to catch us and stop exactly what we're trying to do with, like, content farms. And I don't know about you. I don't really want to bet against Zuck. I don't know. You know? I don't think that's a crazy statement to say. So, that's why I'm going back to, like, that's why I'm going back to hiring humans again. But that's why I'm going back to hiring humans again. Because, number one, flexibility. What the content they can make. And number two, it reduces the risk that my startup gets nuked off the platform.

1:05:19 Because, eventually, there's going to be, like, anything even remotely related to this thing called Yorbi AI, you don't exist on my platform anymore. And I don't want to burn that bridge. Because I've been, my Facebook account is permanently banned from marketing paid ads on Facebook because I used to do dropshipping. But that's, again, different rabbit hole. Different rabbit hole. But I will say this, though. I'm going to end it with this. AI content generation is advancing so fast. Like, I think what we have seen with, like, cloud code and AI vibe coding, just in general, in the past, like, two years.

1:05:53 I think right now, the next year or two, we're going to see the same thing with, like, AI video generation. Some of the latest models, like, Kling 3.0, I can make the same level of quality video. I can't tell. I can't tell anymore that it's AI. I can't either. I can't. I can tell up to, like, a month ago. Now, going forward, I can't tell anymore. I was like, this video, can I show you? No, I've been using Kling. I've been playing it. I completely agree. I can't tell anymore. It got to the point where I'm just like, okay, I can't tell now.

1:06:32 I can't tell. You show me this? I don't even, I wouldn't even have assumed it. Before, Sora has, like, this, like, vibe to it. Yeah, yeah, yeah. Yeah, it's kind of no. It's AI. Kling, Kling 3.0, I'm just like, no. This is actually legit. Nah. Yeah. And I think kind of, like, two key insights you talked about in there, like, the videos are so good now that, one, you can't tell, and two, they're able to convert just as well as normal people and also still establish a trust because no one knows that it's AI. Like, it's already past the level of our human eyes.

1:07:08 Like, I'm sure if you fed it through some algorithm, it would be like, oh, yeah, this is AI because of X, Y, Z. But, like, normal people cannot tell. Even if you study it for, like, a minute or two, you cannot tell. And then also, too, which is kind of an interesting problem, which I think there's definitely people who work on it where, like, they don't want you spamming from one phone. So, like, that's why you see the content farms in, like, China or, like, South Korea or, like, these, like, they have, like, a wall of phones and, like, each of them has, like, a dedicated IP.

1:07:36 And, like, they're kind of ways to get around it. And there are some services that provide you ways to get around it. But, like, you know, naturally, they don't want their platforms to be spanned with a bunch of AI spam, which it's funny that you mentioned, like, they have to kind of also be okay with it at the same time because they're pushing their own AI products. They're trying to say, like, hey, you know, we're an AI company, but so they can't outright ban all AI content. But there are already influencers that have, like, hundreds of thousands of followers, like, many, many of them, probably, like, in the hundreds, if not thousands, easily, of, like, purely AI at this point.

1:08:12 And, like, sometimes you'll see it called out in the comments, but, like, like you said, you really cannot tell probably, like, 95% of the time anymore. Yeah. I will say this, though. I don't think this is, like, doomsday for creators. If anything, I think this is a good thing for creators because I think, like, the creators that leverage AI are going to be even better off. Kind of, like, to your point of, like, coding, where it's, like, the good creators are going to flourish from this, but the ones aren't.

1:08:43 Yeah, that was a wreck. I think that's a common thread. It's probably across every industry where it's, like, right now, if you're exceptional. Exceptional, you're eating. Like, you're doing quite well. But if you're average, like, you're done. You're done-zo.

1:09:00 But, yeah, you answered my follow-up perfectly of, like, okay, this, like, is working out so well. Like, why not continue it? But then to, like, add on to your comment earlier, I think you're absolutely right that, like, AI UGC is going to be the path forward. And, like, what's kind of funny is, like, contrary to this trend, the biggest, one of the biggest AI UGC startups just blew up. Or they've been dead for a while. It's called, like, ArcAds. And their whole cell was, like, oh, you know, we're going to do AI influencers.

1:09:28 They bought a domain name for $12 million. It had a lot of big investors. And now it's, like, completely shut down. Like, their website is gone. The guy, like, changed it. Wait, wait, wait, wait, what? Oh, you're talking about Icon. Icon. Sorry, Icon. Yeah, ArcAds is the latest one that came out, like, blew up, like, a couple months ago. Like, they raised a big round. Yeah, I got that mixed up. Like, before last year. Icon's the one with, like, the founder that went on a tirade on Twitter. Yeah, yeah. He, like, blew up his company.

1:10:00 Like, the website's password protected, his LinkedIn's are, he changed his last name. I don't know if that's an official name change or, like, just on Twitter. But, like, it'd be pretty wild if he legally changed his name to try and distance himself from Icon. But, yeah, ArcAds is actually the one that's doing well. Yeah. Icon is the one that is now defunct. Icon is that story of, you know, whatever, founder aside. They were, like, what's that saying? Were they too early? Yeah. Like, I know what you're talking about.

1:10:34 The head of their time? Yeah. It's, like, when Apple came out with the iPad in, like, 1980s. But then it's, like, no one wants to use it because it just wasn't that good. It was, like, Blackberry with their Palm Pilot. Or, like, Microsoft also came out with a tablet. Like, that was just trash. No one wanted it. Yeah, because I remember when Icon came out, it was, like, 2023, 2024-ish. Like, AI content generation just wasn't there. But it's, like, it came out today and raised that much. It's, like, nah, dude, the tools are ready for you.

1:11:04 Yeah. Like, they literally came out in the last month. Like, they would be one of the hottest startups ever. Like, what's another interesting thing is that, like, I actually think that the whole, literally, someone could just make another arc ads. Like, there's so much opportunity. I feel like a lot of the problems that you've been discussed, like, there are ways around them. Like, you can buy physical devices. You can buy dedicated IPs from VPNs. You can even make your own VPN if you wanted to go that far with it.

1:11:36 And you can sell services, be like, hey, you want to spin up X amount of videos and just sell, like, a credits-based platform like everyone else is doing. Like, it even sounds like something that Yorby could do. It's, like, right now, it sounds like you train people. You write scripts. You are able to copy. Eventually, imagine it's, like, I just give you a video. I'm, like, hey, Yorby, just make me another video. Change the influencer to, like, you know, XYZ stats and just pump out the exact video. Like, you're probably not that far away from it.

1:12:06 Jeez, right. Did you read my pitch text or something? Like, what's going on right now? Let's get a pitch.

1:12:13 Yeah, yeah. That is on the roadmap as well. But did you say, like, getting to, like, content farms as well, like, having phones and everything? I was actually planning to do that because I was thinking, like, as far, I mean, similar to you, my main audience is not really, like, language learning, right? They're more, like, finance and tech and, like, working people who, like, maybe they'll randomly want to learn Chinese or maybe randomly learn another language, but it's not, like, targeted. So then I was thinking, like, okay, I can hire UGC.

1:12:46 But similar to you, I don't want to deal with people. Like, it seems like a pain. Like, I left corporate so I could not deal with a bunch of people. People, yeah. So I was, like, okay, I can go buy a bunch of phones, like, cheap phones from the States. I can, like, surely people already know how to run successful content farms. And, like, the only difference really is instead of hiring a bunch of, like, random people locally and having them and teaching them and training them, I'll just, like, make AI UGC.

1:13:14 Because I was playing with the Kling models literally, like, this past week. And I was, like, this is insane. Like, how, and it's fast, too. Like, before video generation used to take, you know, order of, like, tens of minutes. Now it's, like, one minute or less. Like, 30 seconds and boom. You can have, like, a 15-second video. And like you say, you don't even need 15 seconds. You only need, like, the three to five seconds for the hook. And the rest of it will be whatever you want. Yeah. Be careful, though.

1:13:45 It's, you can do a small scale. You try to scale up to, like, 50 phones. Ooh. Ooh. Yeah, I mean, I feel like surely there's a technological way to get around that. Because, like. I think you're just always going to be playing cat and mouse with, like, these platforms. And like you said, it's, like, maybe you still have UGC creators, but now you're literally teaching them, like, hey, like, use this AI thing. Like, I want you to make, like, instead of, you know, two or three videos today, I want you to make 50 videos today.

1:14:13 And I want you to, like, pick the best ones. I want you to have, like, four accounts. So it's, like, you're still, you know, you're pushing the AI abstraction one human layer down. So one of my UGC creators does do that right now. He makes, I taught him how to make my AI UGC. And he just posts for me, like, AI UGC accounts, too. I'm, like, just testing that to be, like, okay, I did it for myself. Does it work for him? It did. Actually, funny enough, our tool kind of, like, I don't want to say it went viral, but, like, was getting a lot of traction in Korea.

1:14:46 So I made him make, like, an AI Korean UGC account that's in Korean. And that will end up, like, getting a lot of traction with, like, hundreds of thousands of views as well. Kind of a funny story. Like, 20% of our user base is in Korea. Dude, you are literally on the bleeding edge. This is, like, you know, stuff that wasn't even possible a month ago. And now it's, like, you're not only doing it yourself. You taught someone else to do it. They're repeating their process. It's, like, dude, your picture, your picture early.

1:15:13 I have to. I have to, right? Yeah, I don't know. For my job. I think, like, when you said earlier, like, you don't think influencers are cooked now. I kind of feel like they are cooked. Because, like, I tried a bunch of different scenarios. I tried, like, dancing videos. I tried, like, comedy skits. I tried fighting stuff. I tried, like, some, like, cinema-level stuff. It all looks so real. Like, I don't know. Yeah. My mind is just blown. Well, I mean, because you already have Frank New as the brand. So that's actually good for you.

1:15:44 Because now you can be like, oh, my latest video I posted on Instagram, it wasn't trying to hide this. But, like, the second half of the video was AI-generated. It was, I was, like, I'm, like, doing, like, a dance. Like, you know Jump Style, by any chance? It's, like, shuffling. Okay. Basically, if you go to my profile on Instagram, you go look at my latest posted video. The second half of that video, it's AI-generated. I didn't do that. I can't dance like that. But I wanted to because I knew that it would have looked so cool.

1:16:14 It would have fit the video I wanted to make. Yeah. And I'm able to make that video now because of AI. Wow. I mean, I also think, like, if you think about the most popular, some of the most popular streamers of today, like, they're literally just talking heads, right? Like, they're doing React videos of actually viral content. And, like, as I was, like, looking at some Kling resources, I noticed that the dude I was watching was literally an AI avatar. He was an AI avatar teaching people how to use AI. And, like, no one in the comments called it out.

1:16:46 I'm, like, how does no one recognize this dude is not real? Like, the channel. Dude, I know what you're talking about. It's, like, the YouTube channels, right? Yeah. Where it's, like, just AI. It's, like, kind of, like, this is how you do this now. And you do this. And then, like, it pops with him, like, sharing his screen. It looks like he's looking down on his keyboard. Yeah. It's all AI. It's, like, and that's, they have, like, hundreds of thousands of subs. And their videos are getting a lot of views. It's, like.

1:17:12 I mean, I don't think people care. I don't think people care. I think. So, my take on this whole situation is if you're a creator and what you're selling is, like, connection of your audience and community, people care if you're human. They do. But if all you're selling is, like, education, I don't think it matters. Because just, like, okay, just teach me. You know, I know someone made this. As long as the education is good, why do I mind? Which is why I say, like, ads. Like, running paid ads on, like, Facebook, TikTok, whatever.

1:17:42 I think that's going to get completely replaced by AI. Because just, like, you don't care who's selling you something. But I don't think you can replace, like, influencer marketing. Like, we're forever going to know who Frank New is, ideally. Ideally, like, I continue my personal brand to some extent as well. And, like, other creators can leverage that, too. So it's, like, having your name and likeness. And the fact that you're a real human being, you can't replace that. But, like, just, like, running advertisements.

1:18:09 Yeah, it doesn't matter. People, I don't think people care. Yeah. It's like you said. Like, right now, we still have the story. And, like, I think Drake even was saying, like, oh, there are plenty of musicians who are, like, extremely talented, better than me. But they don't have the story. Like, they don't have the narrative, right? They will never get famous. And then, even, like, celebrities, like, I was surprised that Sidney Sweeney was able to move a stock by doing an ad. I was thinking, like, okay, sure, if you made an ad, it went viral.

1:18:36 Or Gap? Or, no, not Gap. It was for, um, a cat company. I forget which one. American Eagle. Yeah, yeah, yeah. Yeah, yeah, yeah. They literally pumped their stock up a couple percent. It was, like, you know, not entirely because of the ad, but it was a large part of it. And I was just, I was honestly shocked because I was, like, okay, sure, she's a famous actress or whatever. But, like, who cares? This is just some random ad, right? But, like, no. It, like, helped their sales a lot. And I don't know. I think that, like, I mean, at this point, a lot of brands have figured out, like, influencer marketing works.

1:19:09 That's why there's a lot of, that's why there's marketing agencies in the first place, right? They know how to try to, like, sell stuff. But I think the surprising part to me is, like, how much power a lot of influencers have. Like, at this point, Speed is probably one of the most famous people on the entire Earth. Like, when he's, like, around the world tour. Like, that's wild to me. Like, they're more popular than real celebrities. I mean, I guess they are the real celebrities now. Like, it's hilarious.

1:19:36 I don't know if you've played the latest Pokemon, ZA, Legends, whatever game. I have not, no. It's not worth playing. But the point I'm making is, one of the gym leaders is a streamer. Like, a live streamer. I was like, what is this, dude? And, like, one of your, like, buddies, one of your, like, friend characters, they're, like, really obsessed with this streamer Pokemon leader. I'm just thinking, like, holy shit, dude. Like, this is, this is actually Gen Alpha type shit. Like, dude, they're with the times, man.

1:20:06 That's, they know their audience. It's crazy. But, yeah, I don't know. I think I could probably replace myself. Like, not, I tried with Sora. But, like you said, it's not quite there. Like, because I'm pretty deadpan. Like, quite, it's, you can tell it's AI me because it's always too animated. Even when I tell it to be deadpan, it's, like, it looks real-ish. But I couldn't even, like, trick my kids with it. And they're, you know, little kids. But I feel like with Kling, it probably is possible. Like, I feel like they don't even work on Sora anymore.

1:20:36 Like, it's just kind of there. Yeah. Yeah. It was, like, a marketing gimmick that they did. I heard that they, like, lose a dollar per generation. So, it's, like, a huge category. Right here. I'm using it every single day still to generate my videos. It's, so you're saying the Sora level is still good enough? Like, you don't need to use Kling 3.0 or, like, the mobile control stuff. Oh, yeah. Oh, yeah. Um, so all my AI avatar account that I grew to, it has, like, it's crazy. It has, like, 7,000 followers on Instagram.

1:21:13 Um, all Sora. All Sora generated. And I think half the people know it's AI. Um, the other half. The, and then, like, the people that know it's AI either don't care. Or, yeah. You know. So, it's good enough. It's good enough. That's what I'm trying to say. But with Kling 3.0, yeah. That, that was the one where I was, like, I don't know anymore. This is, this is, this is, this is it. This is it. Like, this, I think Kling 3.0 was a turning point. Sora was, like, oh, it's going to happen. Kling 3.0 was, like, it's happened.

1:21:47 Yeah. And I think the other really insightful piece you said earlier, it's, like, now you can create whatever brand ambassador you want for whatever specific nationality. It's, like, hey, oh, I have a lot of customers in South Korea. Let me make this South Korean influencer. Oh, I have a lot of, like, customers in, I don't know, Bangladesh or Australia. Like, literally pick any country, any ethnicity, any, like, demographic. And I think this is going to go really, really far. Not even just, like, creating brand ambassadors, but also, like, in movies, you'll be able to hot swap people.

1:22:20 So, like, say you want to watch, like, The Matrix. It's, like, who do you want playing Neo? You want yourself in Neo? You want to watch this whole entire Matrix and you're Neo? Or you want, like, you know, Will Smith instead? Or you want whoever? Like, I think it's going to get, oh, what's his name? Ben Affleck. He had an AI production studio that he started back in, like, 24 or 23 or something. And Netflix just acquired them. Hmm. Wow. Okay. Didn't know that. How much? Do you know? I don't know the details, but I watched a couple interviews where Ben Affleck talks about the whole, like, AI Hollywood scene.

1:23:00 And I always think, like, wow, I'm always surprised that this guy is actually quite intelligent. I thought it was just, like, a rando actor. It's like, now he knows what's going on. Yeah. Got it. Got it. Yeah. You know, kind of going back to the point, the whole Korean creators, I think, like, the whole UGC economy and being UGC thing is, like, very American and Western, like, European as well. I don't think, like, people in Asia know what this is, where you create a separate account and then you post videos on this separate account to market a product.

1:23:35 And I got so sick of trying to explain to people in Korea how it works. I'm just like, I just made it up AI instead. Oh, man. That's, I think that it's, like, kind of funny also because, like... They don't get it. They don't get it. I'm trying to explain to them. Like, okay, so this is UGC. It's a little different to influence on marketing. So, like, okay, okay. So, it's influence on marketing. Well, I mean, kind of. So, what type of creators? I can do it. But generally, like, smaller micro creators make this type of content.

1:24:06 Because, like, if you're a big influencer, you're probably making enough money from brand deals. You'll need to do, like, I don't know, 500 to, like, a couple thousand dollars a month retainers, right? And it's, like, they just didn't understand what I was trying to do. And it's, like, oh, okay, it's, the meta game in Asia is not there yet. Oh, man, that's surprising that they're so far behind. Like, you would think... Well, it's just people I talked to. I don't think I spoke to everyone. Like, David knows about UGC.

1:24:33 He lives in Korea. So, obviously, like, someone there knows what it is. Oh, man. I mean, yeah, it sounds like your whole pitch, like, it's definitely happening. It's, like, it just now depends, like, who's going to build this. And what's funny is, like, even last year, I was thinking about, like, okay, definitely AI influencers are going to be a thing. Like, it was one of the ideas I considered pursuing. But, like, I didn't really have, like, a feasible way of building anything good enough. Because I'm, like, man, all these things suck right now.

1:25:04 Like, but now it is good enough. Like, now's the perfect time. Yeah, yeah, yeah. That's the idea here. Very cool. We're already over time by, like, half an hour. Or is there anything that you wanted to kind of wrap up with in terms of, like, promoting or just, like, chatting about? Anything you want to shout out at the end here? Um, let me think about this.

1:25:33 Well, if you're a content creator or marketer, and you're looking for the best AI script writer you ever used in your life, go check out Yorbi AI. I guess I want to give a quick shout out to all my friends and family for potentially going to watch this. I got to say, Frank, this is actually an honor. It's actually really funny. I think I'm going to show, like, people that I spoke to you and they're like, because I think you're one of the most well-known figures within, like, the Asian American corporate ecosystem.

1:26:01 Like, everyone knows who you are. My older brother knows who you are. I'm a nobody. Okay, everyone, I'll say it at the end. Thanks a lot to Andrew. He's clearly going to be more successful with me than me. If you're an investor, you know, listen to his pitch deck. It's definitely where the industry is headed. Yeah. Email me, andrew at yorbi.ai, Y-O-R-B-Y dot A-I, if you're interested in hearing my pitch of the future of AI and content. Amazing. All right. Well, thanks, everyone, for listening. This has been another episode of the Frankly Speaking podcast.