you2idea@video:~$ watch a2JBWwASzUU [38:29]
// transcript — 1081 segments
0:03 This is not your standard SIP episode. I brought on this guy who went from zero
0:09 to 250 million mobile app installs using some open-source AI models. He tells the
0:15 whole story how he did it. You know, it definitely fired me up around this whole
0:19 new wave of creating, you know, multi-million dollar uh mobile apps in
0:27 under 7 months. [Music] Ben, you've created these AI apps. You've got hundreds of millions of
0:39 downloads. You're backed by Nvidia. I want you to teach us, well, actually,
0:44 what do you want to teach us before what I want you to teach us?
0:47 >> Yeah. Uh, I thought I'd just come on here and talk about a strategy that's
0:51 worked well for me in many different domains in life. Uh, and it's sort of a
0:59 basic fundamental of mimisus, which is if something is working, then probably
1:03 you can copy it and put your own spin on it and it'll continue to work.
1:07 >> Okay? And by the end of this episode, like what are the concrete things that
1:11 people are going to learn? >> I'm going to show you how I've applied
1:16 that strategy to the few different uh successful things in my life. how I used
1:22 it to come up with a concept for an app, my first mobile app ever that got over
1:27 100 million downloads. How we use that concept to consistently go viral on
1:32 Instagram and have reels that get millions of views and that are on one
1:37 hand a remix of something else that already exists, but on the other hand
1:42 something that we put our own unique uh flavor on. >> Okay. I mean, how to create apps that
1:47 get millions of downloads, how to get how to create content that gets millions
1:51 of views, how to raise, you know, millions of dollars from from people
1:56 like Nvidia. You you have my attention. >> Okay, let's do it. It's actually pretty
1:59 [ __ ] easy if you uh don't overthink it too much. Um, so, okay, first very
2:04 brief anecdote. I think I picked up this principle from video games. I was a
2:08 jungle main in League of Legends. I was sort of like a platinum gold level
2:13 player which is like 60th 70th percentile. I wanted to get better. So I
2:18 started watching streams of players that were better than me that were playing
2:22 the same role and I would just copy them and then put my own spin. So that's
2:25 where this sort of entered my head for the first time. Uh in the summer of 2020,
2:33 I applied this to conceptualize and eventually launch uh one of the most
2:38 viral apps ever, Aombo. Uh so there were a few signals that I picked up on. First
2:43 was Reface. Uh Reface in the summer of 2020 was the number one app on the app store and Dre
2:52 invested and led a $5.5 million round. And what got my attention about this was
2:56 it was an app that I fully understood. Uh, so what did Reface do? Reface did
3:01 face swaps. So you could put in like a Cardi B music video, >> take a selfie, it would put your selfie
3:07 into the Cardi B music video. Hundreds of millions of downloads, memes that,
3:11 you know, my friends were sending each other. Um, I knew which model they were
3:15 using and how they were accomplishing this. And the the design of the app was
3:21 also very simple and intuitive and kind of made sense to me. So that was the
3:24 first seed. Uh, the second seed was this PewDiePie video >> where he covered some very strange
3:35 memes. And we can see the date here, 4 years ago, August 2020.
3:40 >> It's beautiful. It's beautiful. And of course, this got 7 million views because
3:45 it's PewDiePie. Um, and this got my attention as well because this was
3:49 another open source AI model that I was familiar with called the first order
3:54 motion model. And uh it was starting to go viral in sort of the AI circles and
3:59 also just like the hardcore like Discord autist circles like people on their
4:04 computers for like 14 hours a day. Um and they were Yeah, people like us.
4:08 >> And they were using a Google Collab notebook uh to basically stitch this
4:13 output together. You gave it a selfie, you gave it a driving video, which is
4:17 what kind of creates this animated choreography. It's actually someone else
4:21 moving their head around. And so this one driving video of a guy
4:26 singing uh Baka Mita Thai went viral over and over and over again. And so one
4:31 night I'm uh smoking a joint with my roommate on the roof. We're looking at
4:34 these memes. Someone made a Telegram bot where you could do something similar.
4:38 We're like, man, like why doesn't someone just make an app like where anyone can do this where it's
4:46 simple? Uh, and so immediately like it kind of starts off as a joke, but the
4:50 circuits start firing like, "Oh wait, like Reface just did the same thing.
4:53 They have hundreds of millions of downloads. Here's a new open source AI
4:57 model that like no one's done this for yet." Like easy. Like we're going to be
5:02 billionaires. We'll get this done in a week, you know. Um, did not take a week.
5:09 Took us seven months. But in uh March of 2021, we launched Wombo. And maybe I can
5:14 find a quick uh walkthrough on YouTube. >> When when you had this idea and you saw
5:22 that A16Z reace company, like a lot of people would look at that and be like,
5:26 "Okay, they're going to win this space." Like, you know, it's already going to
5:29 get competitive. But you kind of said, "No, no, no. I'm going to create an I'm
5:33 going to compete with them." >> Yeah. I wasn't thinking that I'm going
5:36 to compete with them. I was just thinking like it's early like there's
5:41 lots of space to play like I don't know like there's where my mind goes to is like uh like
5:50 like movie directors or or musical artists like we're never going to run
5:55 out of like new movie directors or artists like we're going to continuously
6:01 want new interesting [ __ ] So I I don't think anything is like too crowded.
6:03 >> Mhm. >> Yeah. >> Abundance mindset. Yeah,
6:09 >> Sam Alman, the co-founder of OpenAI, just said that it is the era of the idea
6:15 guy, and he is not wrong. I think that right now is an incredible time to be
6:19 building a startup. And if you listen to this podcast, chances are you think so,
6:22 too. Now, I think that you can look at trends uh to basically figure out uh
6:28 what are the startup ideas you should be building. So, that's exactly why I built
6:33 ideabser.com. Every single day you're going to get a free startup idea in your
6:39 inbox and it's all backed by high quality data trends. How we do it?
6:44 People always ask. We use AI agents to go and search what are people looking
6:49 for and what are they screaming for in terms of products that you should be
6:53 building and then we hand it on a, you know, silver platter for you to go check
6:58 out. Um, we do have a few paid plans that, you know, take it to the next
7:02 level. uh give you more ideas, give you more AI agents and more almost like a
7:07 chat GBT for ideas with it, but you can start for free ideabrowser.com. And if
7:11 you're listening to this, I highly recommend it. Um, so I'll just kind of
7:16 briefly walk through this. You can actually see that in this video, this
7:21 lady shows this old Google Collab notebook for the first order motion
7:24 model. So like my grandma could never use this [ __ ] Like I mean she could if
7:28 I like taught her Python. >> Yeah. but like she could never use that.
7:34 Um, so here we go. Uh, you would take a selfie. That's the first thing you do. Then you pick a song
7:41 and then you sit through a little loading screen like this. And then
7:48 finally, you get your output. And uh, we'll just see see that at the end.
7:53 obviously looks super basic by 2025 standards, but literally those three
8:01 screens of input, song selection, output, oh, sorry, loading screen, and then
8:06 output, four screens. That's all we needed to get 100 million downloads. And
8:12 uh we intentionally sort of designed something that would be extremely simple
8:18 to use. um sort of uh the whole thing is engineered for verality in a way because
8:24 the whole expectation is you as a user come in, you make a piece of content,
8:28 you make that piece of content extremely easily and extremely quickly and then
8:33 your natural inclination is to share or to make another one where like your dad
8:37 is the one singing or the pope or a politician or whatever. And so we we
8:45 spent no money on marketing and we spent all our money on inference and uh that's
8:51 that's how the thing went viral. Now maybe the only other thing I'll say
8:55 here and I'll kind of kiss my own ass a little bit here is the song selection.
9:00 >> Mhm. Uh we selected 15 songs that we launched with and these are some of the most
9:07 recognizable and iconic songs of all time or there were songs there were kind
9:13 of meme songs and so like you know one example of something that would make a
9:17 song a meme song is you don't need to know English to vibe with it.
9:21 >> Uh for example uh boom boom boom boom by the Wenger boys.
9:24 >> You don't need to know English. No >> sounds great.
9:25 >> Yeah. Absolutely. >> Wherever you are. >> Yeah. And so we we really deliberately
9:31 picked those 15 songs and then we really deliberately made the 15 driving videos
9:39 for those songs. And uh that was I think key >> and and he made it like stupid simple.
9:42 >> Stupid simple >> stupid like this is you can give this to
9:46 a 5-year-old and they'll know what to do. >> Exactly. >> A buddy of mine has a an app that's
9:51 going viral right now. Maybe you've seen it. Mixie. >> Mixie.
9:54 >> Yeah. Have you seen that? >> I think so. >> Yeah. So, it's like a mashup. Um, like
10:01 go to just just Google Mixi >> Mixie app. >> I think it's with an I. There it is.
10:06 >> Oh, yeah. I've seen this. >> Yeah. So, it's like two guys.
10:12 >> Um, you know, you put two beats together, it mashes it up. Like, simple idea, right?
10:21 But the UI is really simple. Um, and they're getting like >> I mean, it's one of the top music apps.
10:24 Yeah. >> Um, I don't know the monetization of this, but
10:31 I think there's this whole next there's this whole new generation of apps that
10:35 are just like really stupid simple. >> Um, that uh I mean it's a similar
10:41 concept, right? Like I'm sure his songs are like he's doing a queue of songs
10:45 that you know people actually want to choose, right? Um, it's simple UI. Um,
10:52 and it's viral because you want to share that on IG, etc. >> Yeah. I I personally
10:59 like starting with stupid simple >> and maybe from stupid simple you can do
11:04 something more advanced and interesting. >> Yeah. >> Um, but
11:07 >> with with respect to mobile apps, right? >> With respect to mobile apps.
11:10 >> And I mean it it all kind of boils down to attention at the end of the day. Like
11:15 if if you're trying to make a product, uh it's not just someone's money that
11:19 you're after. Like even before that, it's their attention. And there's so
11:22 many [ __ ] distractions. There's so many different things that are pulling
11:25 you in a million different directions at all times that I find the best way to
11:32 sort of introduce myself and get someone's attention for the first time
11:35 is to do something stupid, do something funny, do something simple, not try to
11:39 [ __ ] explain Socrates to them or something. Um, but just basically tell a
11:45 small joke or another analogy could be like give them a small piece of candy,
11:47 >> right? >> As opposed to like, you know, here's a
11:49 [ __ ] steak, >> right? Yeah. Because what I'm trying to
11:52 figure out, what I'm trying to reverse engineer is how do you create a mobile
11:57 app that gets millions of downloads? So, stupid simple one. >> What else? Uh so uh in our case there's
12:06 always been some kind of viral content creation loop meaning someone comes in
12:12 makes a piece of content shares that content on other platforms and that's
12:17 what drives your user acquisition. Um and so that whole uh that whole
12:23 component has a lot of extremely important details. Um so first is the
12:27 content creation. How are you going to help someone make a good piece of
12:31 content? Also, the game has changed since 5 years ago uh with how far
12:36 generative AI has come um and the kind of content that people are interested in
12:39 making, but still I think it's applicable today. Um so here's how I
12:45 would boil it down. How do you how do you help someone make a piece of content
12:48 that either they're going to post on Instagram and post on Tik Tok for the
12:53 algorithm and the algorithm is like that's a great piece of content. let's
12:57 give it millions of views or more for their own like personal networks and the
13:01 people in the personal network are going to be like wow that's really awesome I
13:05 want to make that thing too >> so you're basically saying you start
13:09 with the content like the end content which is kind of what my friend Joe is
13:13 doing with Mixie right cuz like he knows that people want to share these mixes
13:18 >> absolutely and so like you know one thing maybe we could touch on briefly is
13:23 just the algorithmic feeds of today like the amount of human attention that is
13:27 just being sucked into the full screen infinite scroll format whether it's Tik
13:32 Tok or Instagram or YouTube shorts or you know X has one like that is just
13:37 like the battlefield of attention and so I think like a great source of product
13:42 inspiration is to look at that and look at where are people's attention going
13:45 what kind of content are people making can I make it easier for them to make
13:49 some some piece of content that is consistently going viral over and over
13:52 again >> and then and then put my watermark in the bottom right.
13:57 >> Does this only work for like BTOC fun apps or could this also work for like
14:04 B2B or or proumer type apps? >> I think it could work for everything.
14:09 like you probably need to adjust your uh like principles a little bit and like
14:13 you know maybe I'm going to stop paying attention to the Tik Tok algorithm and
14:17 start paying attention to like I don't know the conference that all the people
14:22 in a particular niche are going to but it's it's the same thing like what do
14:26 the people you care about pay attention to >> right >> and where right
14:34 >> and and sort of the key the key point is like look at what other you you know,
14:38 look at the content, look at what's going viral, and then apply that to
14:41 whatever it is you're doing. >> Yes. So, in the case of Wombo in the
14:48 summer of 2020, it was, well, Reef is just got a number one app and hundreds
14:52 of millions of downloads, and hey, there's this new model now, which is
14:56 still very hard to use. This is another very good signal, by the way. If
15:01 something is going viral inside of like some technical circle, but it's still
15:05 it's still difficult for a lay person to use, then uh if you can help the lay
15:11 person get access to that same technology and make it easier for them,
15:15 that's typically like a really good recipe for verality or a really good
15:21 sign that something might go viral. Um, so I'll just wrap up the the Wombo
15:24 story. Y >> from kind of that moment of inspiration, it took us seven months to actually get
15:30 the app on the app store. Uh, but we finally launched it in March of 2021. It
15:36 went insanely viral, got like 50 million downloads in its first month, uh, in
15:42 first 3 months. And um you know after sort of some just like stabilizing and
15:46 taking it all in and raising a little bit of money and building out our team
15:52 uh the obvious like next question was okay well we just did this with one AI
15:57 model. What other AI models can we do this with? And now this is a really
16:02 obvious question to ask in 2025 in the age of AI. Um again 2020 was still a
16:07 little bit different earlier. Generative AI was not even a buzz word yet
16:11 >> and generative AI is already a buzz word that we've moved past now which
16:15 >> what's the latest now >> I feel like now it just like AGI super
16:19 intelligence like agents >> you know uh so the next concept was dream which
16:27 was an artwork generator um for the nerds it predates midjourney and stable
16:33 diffusion and deli and was basically the first time that image generation went
16:37 viral now of course image generation is everywhere. Um and uh we saw something
16:43 very similar happening again almost the exact same thing we saw happening with
16:47 Wombo. Uh people were playing with a combination of two models VQ GAN
16:55 [ __ ] Vugan and Pashan I need your help. >> What's up?
16:57 >> I forgot what were the two models. >> Huh? >> Diffusion models.
17:00 >> It was Vugan diffusion. >> Vugan clip. >> Vugan clip. That's right. Vugan clip. So
17:07 you combine VQGan with clip, you could generate these crazy dreamlike images.
17:12 There was like maybe 20 or 30 people on Twitter um and really like five at the
17:17 beginning who were like messing around with this stuff. And those were like the
17:22 earliest incantations of like this era's neural networks for image synthesis. And
17:28 we did the same thing. Uh we wrapped that up into a fun and easy to use
17:33 mobile app. We made an an extremely optimized version of the model which had
17:38 fast generation times and uh we we we did some templating to
17:45 make it easier to create high quality images. Basically just adding words to
17:51 the prompt that the user gave us. And uh exactly the same thing happened another
17:56 number one app. And many of the lessons that we learned from Wombo about give the user as simple
18:04 to provide input. In the case of Wombo is a selfie. In the case of Dream was a
18:08 text prompt. Pre-process that user's input with your own input that kind of like structures
18:16 it and makes it work well. So in the case of Wombo, it was the driving video.
18:20 In the case of Dream, it was the kind of stylistic keywords that we were adding
18:25 to the user's prompt. uh make the inference really really fast and free
18:32 for the user. Um did that in both cases put all that together in a fun and easy
18:36 to use interface and you have something with like all the ingredients for
18:38 verality. >> Okay. And um like Okay. So it goes number one. What next?
18:49 >> Yeah. So just in terms of the trajectory of the company,
18:52 listen, I had no idea what the [ __ ] I was doing. Like I'm not like one of
18:56 these like Silicon Valley kids who like watched every Paul Graham video. I was
19:01 just like smoking weed in my bedroom playing League of Legends and like
19:06 really really interested in AI. Um so made every mistake we could. Uh but uh
19:13 the at that time now we're sort of at the end of 2021. We had a team of about
19:18 20 people all working here in Toronto. Um got a few apartments that we were
19:25 using as offices. Um, in 2022, we're like, let's raise money, cuz
19:29 that's what you're supposed to do. Uh, ended up putting together a series A uh
19:36 where the lead investor kept us in due diligence for 2 months and then pulled
19:41 out of the deal. and they pulled out of the deal partially because our original
19:47 Wombo app used music and there were some sort of unknown liabilities and risks
19:52 around our use of music. I don't know if I should bring it up with these mixie
19:56 guys, but [ __ ] the music licensing people are vicious, annoying
20:00 [ __ ] but also, you know, it is what it is. Um, and also partially because FTX crashed
20:09 and uh all this crazy [ __ ] was happening in crypto and
20:11 >> Zer was over. >> Yeah, there was like a market spook.
20:16 >> So, we went from having [ __ ] number one app thinking we're about to get $15
20:22 million to having negative runway and being like, "What the [ __ ] just
20:25 happened?" Um, >> negative runway meaning like you're out
20:27 of money. >> Yeah. Like >> or you're like really out of money. like
20:34 really out of money where and it really happened because that month Dream went
20:38 viral again and we had a million dollar server bill on Amazon for that month for
20:43 all the influence that we were doing and I had like maybe not the perfect
20:49 philosophy around this which which was like money is fake let's just make an
20:52 amazing [ __ ] product let's get all the users we'll figure it out later um
20:59 yeah didn't work at that point So, uh, we went into like a year and a half
21:05 of cockroach mode basically where I wasn't paying any of my bills unless I
21:09 absolutely had to. We were extremely focused on monetizing our product which
21:13 previously we had never given a [ __ ] about and basically getting to like
21:18 profitability and along the way we you know still raised some money from
21:22 investors who believed in us. Uh, so yeah about a year and a half of
21:27 cockroach mode got to profitability. And how did you monetize the apps? Because I
21:30 think >> because I think that's a question that a lot of people have, which is like, okay,
21:34 great. Now I've built this mobile app. It's starting to get some traction.
21:36 >> Yeah. >> What, you know, how do you think about
21:41 monetizing? Like what what's the sauce around monetizing mobile apps?
21:45 >> Just copy people who are better than you. >> So, who's the best? Who's the best? And
21:49 what are they doing? >> [ __ ] Nikita, man. >> Nikita. That guy is the
22:03 >> Shout out Nikita Beer. Yeah, go or I guess you can't do it anymore, but go
22:07 like pay $10,000 to talk to him on intro and I'm sure he'll give you some good
22:09 tips. >> Okay, so how did you how did you monetize it?
22:15 >> Uh we introduced a subscription and we introduced ads. The subscription did
22:18 things like, well, if you want to generate an image, uh, you could
22:23 generate four at a time instead if you're a subscriber, or you could act
22:27 have access to some premium styles if you're a subscriber. Um, and that took
22:33 us from zero to I think at our peak, Dream was doing over $500,000 a month.
22:39 Uh, and that got us to the profitability and sort of stability that we needed.
22:43 And and from a sub like what are we talking like $5 a month, $5 a week? I
22:48 know a lot of people are doing weekly subscriptions now. Like what was it?
22:54 >> We've done everything from $5 a week to $5 a month. We started really cheap and
23:01 got more expensive over time. And um there's also like all the like dark
23:05 patterns that you could do with like hidden trials and all that kind of
23:09 stuff, but we never went to too deep into that. >> Yeah. Okay. Have you I'm sure you
23:15 studied like Cal AI and what they do. >> Yeah. Yeah. Yeah.
23:16 >> What do you think? >> It's a new generation of killers, man.
23:22 >> I love to see it. It seems like every few months now there's like a new
23:26 18-year-old kid making $2 million a month on some simple stupid [ __ ] And uh
23:33 I think it's only going to accelerate. >> Mhm. Yeah. And and those apps are
23:38 monetizing via subscriptions, right? >> Yeah. Weekly, monthly?
23:42 >> I'm not sure. We'd have to double check. >> I think weekly. Yeah. I think a What I
23:47 That's a trend I'm seeing a lot is like weekly subscriptions anywhere from like
23:53 $3 to $20 a week. >> Mhm. Makes sense. >> Yeah. >> Cool. Okay. So, going back, so you're
24:02 you start making money basically. >> And then what next?
24:06 >> Yeah. So we slowly emerge out of cockroach mode. I guess bringing us into
24:13 2023 and 2024. The most notable things that happened uh were one, we had a
24:18 relationship with Nvidia. Uh they approached us very early on uh because
24:23 they saw that our apps were going viral and that we were doing this thing that
24:26 was going to be called generative AI and they want to kind of learn more about it
24:30 and see how they could support it. Of course, it's evolved like far far beyond
24:34 that now. Um and that culminated in uh us meeting them more and more and them making an
24:44 investment into Wombo. Uh the second thing that happened was I
24:50 sort of I lost some interest in making apps and I gained more interest in you know where
24:59 all of this was going, where AI was going. Oh man, this is going to be kind of a
25:04 convoluted point, but it'll make sense at the end. Okay, so one thing that
25:09 we've always cared about is inference efficiency. How do we make generation
25:15 times extremely fast and how do we spend the least amount of money that we can on
25:20 powering this inference? Uh so initially that meant we are going to host our
25:28 models on GPUs. We're going to sign a three-year contract with Amazon where
25:33 they give us some better pricing and we're going to pay uh specialists
25:41 $3,000 a day to make our model produce the same output but in a fraction of the
25:45 time and it's just worth it to find a specialist like that. That was sort of
25:49 the first solution. And and and then we're going to run all the inference on
25:53 the cloud. Our users never have to worry about it and it's consistent. It's high
25:58 quality. It's reliable. Uh then there was a motivation to move some of
26:07 this inference on device especially as your user base has devices in it that
26:13 are powerful and perfectly capable of running the model that you care about.
26:17 And so we started doing a little bit of that and sort of the natural extension of
26:23 that is well if there's someone in my user base with an iPhone 16 Pro or a
26:30 computer with a 4090 on it um that can run their own workloads,
26:34 what if they could run workloads for other users as well? And so that idea of
26:41 sort of peer-to-peer computing and using our users compute was born out of our
26:46 continuous push towards inference efficiency. Um hopefully all of that is
26:52 making sense so far. Uh and and so the other thing that happened was at the end of 2023 at the
26:59 start of 2024 I got really interested in how AI and crypto would intersect. And
27:06 so I looked at every project in the space, studied them closely, again
27:11 coming back to that same principle of copy what other people are doing or at
27:14 least understand what other people are doing >> before you do your own thing. And so we
27:18 looked at, you know, Bit Tensor and Jensen and Render and Akos and Golem and
27:26 the Internet Computer and Folding at Home and all these other sorts of
27:31 examples of decentralized AI or decentralized computing happening.
27:36 And I just ended up finding that extremely interesting. Um, and to make a
27:42 long story short, aside from learning about those projects and even building
27:48 in some of their spaces, uh, the conclusion we came to was to build our
27:53 own thing called W.ai, which would essentially allow anyone to turn their
27:57 MacBook, their gaming computer, and in the future their PlayStation or iPhone
28:01 or whatever into a compute contributor. And so, it's almost like Airbnb for that
28:05 device. when you're not using it, you can make it available to the world. It's
28:10 going to run some hopefully valuable workload and earn you a reward.
28:15 >> And reward being money. >> Reward today being points
28:19 >> which will later turn into a token >> which will later turn into money.
28:24 >> Uh I'm not like media trained so I don't know what what I'm
28:26 >> not media trained. I had no idea. >> No, I'm not [ __ ] media trained. Uh
28:30 but >> the idea is you are contributing to the network. Like if I just let if I let W.I
28:39 generate 10,000 images for somebody else on my machine, >> then I should be able to generate at
28:45 least 10,000 and1 images in return. And that seems fair. >> Yeah.
28:49 >> Totally cool. And that's the journey. >> That's the journey. A lot of crazy [ __ ]
28:55 has happened in between. Um cycled through the team many many times. We
29:00 started with six co-founders. Only one remains. Three of them have started new
29:04 companies that are very [ __ ] cool and interesting. Uh so check out Omega,
29:10 check out Dippy. That's the Wombo Mafia. And uh now we all live in a big ass
29:15 house in LA. There's like 12 of us who live and work there. Uh there's a few of
29:18 us who don't live and work there, but are still core parts of the team. Um but
2:04 brief anecdote. I think I picked up this principle from video games. I was a
2:08 jungle main in League of Legends. I was sort of like a platinum gold level
2:13 player which is like 60th 70th percentile. I wanted to get better. So I
2:18 started watching streams of players that were better than me that were playing
2:22 the same role and I would just copy them and then put my own spin. So that's
2:25 where this sort of entered my head for the first time. Uh in the summer of 2020,
2:33 I applied this to conceptualize and eventually launch uh one of the most
2:38 viral apps ever, Aombo. Uh so there were a few signals that I picked up on. First
2:43 was Reface. Uh Reface in the summer of 2020 was the number one app on the app store and Dre
2:52 invested and led a $5.5 million round. And what got my attention about this was
2:56 it was an app that I fully understood. Uh, so what did Reface do? Reface did
3:01 face swaps. So you could put in like a Cardi B music video, >> take a selfie, it would put your selfie
3:07 into the Cardi B music video. Hundreds of millions of downloads, memes that,
3:11 you know, my friends were sending each other. Um, I knew which model they were
3:15 using and how they were accomplishing this. And the the design of the app was
3:21 also very simple and intuitive and kind of made sense to me. So that was the
3:24 first seed. Uh, the second seed was this PewDiePie video >> where he covered some very strange
3:35 memes. And we can see the date here, 4 years ago, August 2020.
3:40 >> It's beautiful. It's beautiful. And of course, this got 7 million views because
3:45 it's PewDiePie. Um, and this got my attention as well because this was
3:49 another open source AI model that I was familiar with called the first order
3:54 motion model. And uh it was starting to go viral in sort of the AI circles and
3:59 also just like the hardcore like Discord autist circles like people on their
4:04 computers for like 14 hours a day. Um and they were Yeah, people like us.
4:08 >> And they were using a Google Collab notebook uh to basically stitch this
4:13 output together. You gave it a selfie, you gave it a driving video, which is
4:17 what kind of creates this animated choreography. It's actually someone else
4:21 moving their head around. And so this one driving video of a guy
4:26 singing uh Baka Mita Thai went viral over and over and over again. And so one
4:31 night I'm uh smoking a joint with my roommate on the roof. We're looking at
4:34 these memes. Someone made a Telegram bot where you could do something similar.
4:38 We're like, man, like why doesn't someone just make an app like where anyone can do this where it's
4:46 simple? Uh, and so immediately like it kind of starts off as a joke, but the
4:50 circuits start firing like, "Oh wait, like Reface just did the same thing.
4:53 They have hundreds of millions of downloads. Here's a new open source AI
4:57 model that like no one's done this for yet." Like easy. Like we're going to be
5:02 billionaires. We'll get this done in a week, you know. Um, did not take a week.
5:09 Took us seven months. But in uh March of 2021, we launched Wombo. And maybe I can
5:14 find a quick uh walkthrough on YouTube. >> When when you had this idea and you saw
5:22 that A16Z reace company, like a lot of people would look at that and be like,
5:26 "Okay, they're going to win this space." Like, you know, it's already going to
5:29 get competitive. But you kind of said, "No, no, no. I'm going to create an I'm
5:33 going to compete with them." >> Yeah. I wasn't thinking that I'm going
5:36 to compete with them. I was just thinking like it's early like there's
5:41 lots of space to play like I don't know like there's where my mind goes to is like uh like
5:50 like movie directors or or musical artists like we're never going to run
5:55 out of like new movie directors or artists like we're going to continuously
6:01 want new interesting [ __ ] So I I don't think anything is like too crowded.
6:03 >> Mhm. >> Yeah. >> Abundance mindset. Yeah,
6:09 >> Sam Alman, the co-founder of OpenAI, just said that it is the era of the idea
6:15 guy, and he is not wrong. I think that right now is an incredible time to be
6:19 building a startup. And if you listen to this podcast, chances are you think so,
6:22 too. Now, I think that you can look at trends uh to basically figure out uh
6:28 what are the startup ideas you should be building. So, that's exactly why I built
6:33 ideabser.com. Every single day you're going to get a free startup idea in your
6:39 inbox and it's all backed by high quality data trends. How we do it?
6:44 People always ask. We use AI agents to go and search what are people looking
6:49 for and what are they screaming for in terms of products that you should be
6:53 building and then we hand it on a, you know, silver platter for you to go check
6:58 out. Um, we do have a few paid plans that, you know, take it to the next
7:02 level. uh give you more ideas, give you more AI agents and more almost like a
7:07 chat GBT for ideas with it, but you can start for free ideabrowser.com. And if
7:11 you're listening to this, I highly recommend it. Um, so I'll just kind of
7:16 briefly walk through this. You can actually see that in this video, this
7:21 lady shows this old Google Collab notebook for the first order motion
7:24 model. So like my grandma could never use this [ __ ] Like I mean she could if
7:28 I like taught her Python. >> Yeah. but like she could never use that.
7:34 Um, so here we go. Uh, you would take a selfie. That's the first thing you do. Then you pick a song
7:41 and then you sit through a little loading screen like this. And then
7:48 finally, you get your output. And uh, we'll just see see that at the end.
7:53 obviously looks super basic by 2025 standards, but literally those three
8:01 screens of input, song selection, output, oh, sorry, loading screen, and then
8:06 output, four screens. That's all we needed to get 100 million downloads. And
8:12 uh we intentionally sort of designed something that would be extremely simple
8:18 to use. um sort of uh the whole thing is engineered for verality in a way because
8:24 the whole expectation is you as a user come in, you make a piece of content,
8:28 you make that piece of content extremely easily and extremely quickly and then
8:33 your natural inclination is to share or to make another one where like your dad
8:37 is the one singing or the pope or a politician or whatever. And so we we
8:45 spent no money on marketing and we spent all our money on inference and uh that's
8:51 that's how the thing went viral. Now maybe the only other thing I'll say
8:55 here and I'll kind of kiss my own ass a little bit here is the song selection.
9:00 >> Mhm. Uh we selected 15 songs that we launched with and these are some of the most
9:07 recognizable and iconic songs of all time or there were songs there were kind
9:13 of meme songs and so like you know one example of something that would make a
9:17 song a meme song is you don't need to know English to vibe with it.
9:21 >> Uh for example uh boom boom boom boom by the Wenger boys.
9:24 >> You don't need to know English. No >> sounds great.
9:25 >> Yeah. Absolutely. >> Wherever you are. >> Yeah. And so we we really deliberately
9:31 picked those 15 songs and then we really deliberately made the 15 driving videos
9:39 for those songs. And uh that was I think key >> and and he made it like stupid simple.
9:42 >> Stupid simple >> stupid like this is you can give this to
9:46 a 5-year-old and they'll know what to do. >> Exactly. >> A buddy of mine has a an app that's
9:51 going viral right now. Maybe you've seen it. Mixie. >> Mixie.
9:54 >> Yeah. Have you seen that? >> I think so. >> Yeah. So, it's like a mashup. Um, like
10:01 go to just just Google Mixi >> Mixie app. >> I think it's with an I. There it is.
10:06 >> Oh, yeah. I've seen this. >> Yeah. So, it's like two guys.
10:12 >> Um, you know, you put two beats together, it mashes it up. Like, simple idea, right?
10:21 But the UI is really simple. Um, and they're getting like >> I mean, it's one of the top music apps.
10:24 Yeah. >> Um, I don't know the monetization of this, but
10:31 I think there's this whole next there's this whole new generation of apps that
10:35 are just like really stupid simple. >> Um, that uh I mean it's a similar
10:41 concept, right? Like I'm sure his songs are like he's doing a queue of songs
10:45 that you know people actually want to choose, right? Um, it's simple UI. Um,
10:52 and it's viral because you want to share that on IG, etc. >> Yeah. I I personally
10:59 like starting with stupid simple >> and maybe from stupid simple you can do
11:04 something more advanced and interesting. >> Yeah. >> Um, but
11:07 >> with with respect to mobile apps, right? >> With respect to mobile apps.
11:10 >> And I mean it it all kind of boils down to attention at the end of the day. Like
11:15 if if you're trying to make a product, uh it's not just someone's money that
11:19 you're after. Like even before that, it's their attention. And there's so
11:22 many [ __ ] distractions. There's so many different things that are pulling
11:25 you in a million different directions at all times that I find the best way to
11:32 sort of introduce myself and get someone's attention for the first time
11:35 is to do something stupid, do something funny, do something simple, not try to
11:39 [ __ ] explain Socrates to them or something. Um, but just basically tell a
11:45 small joke or another analogy could be like give them a small piece of candy,
11:47 >> right? >> As opposed to like, you know, here's a
11:49 [ __ ] steak, >> right? Yeah. Because what I'm trying to
11:52 figure out, what I'm trying to reverse engineer is how do you create a mobile
11:57 app that gets millions of downloads? So, stupid simple one. >> What else? Uh so uh in our case there's
12:06 always been some kind of viral content creation loop meaning someone comes in
12:12 makes a piece of content shares that content on other platforms and that's
12:17 what drives your user acquisition. Um and so that whole uh that whole
12:23 component has a lot of extremely important details. Um so first is the
12:27 content creation. How are you going to help someone make a good piece of
12:31 content? Also, the game has changed since 5 years ago uh with how far
12:36 generative AI has come um and the kind of content that people are interested in
12:39 making, but still I think it's applicable today. Um so here's how I
12:45 would boil it down. How do you how do you help someone make a piece of content
12:48 that either they're going to post on Instagram and post on Tik Tok for the
12:53 algorithm and the algorithm is like that's a great piece of content. let's
12:57 give it millions of views or more for their own like personal networks and the
13:01 people in the personal network are going to be like wow that's really awesome I
13:05 want to make that thing too >> so you're basically saying you start
13:09 with the content like the end content which is kind of what my friend Joe is
13:13 doing with Mixie right cuz like he knows that people want to share these mixes
13:18 >> absolutely and so like you know one thing maybe we could touch on briefly is
13:23 just the algorithmic feeds of today like the amount of human attention that is
13:27 just being sucked into the full screen infinite scroll format whether it's Tik
13:32 Tok or Instagram or YouTube shorts or you know X has one like that is just
13:37 like the battlefield of attention and so I think like a great source of product
13:42 inspiration is to look at that and look at where are people's attention going
13:45 what kind of content are people making can I make it easier for them to make
13:49 some some piece of content that is consistently going viral over and over
13:52 again >> and then and then put my watermark in the bottom right.
13:57 >> Does this only work for like BTOC fun apps or could this also work for like
14:04 B2B or or proumer type apps? >> I think it could work for everything.
14:09 like you probably need to adjust your uh like principles a little bit and like
14:13 you know maybe I'm going to stop paying attention to the Tik Tok algorithm and
14:17 start paying attention to like I don't know the conference that all the people
14:22 in a particular niche are going to but it's it's the same thing like what do
14:26 the people you care about pay attention to >> right >> and where right
14:34 >> and and sort of the key the key point is like look at what other you you know,
14:38 look at the content, look at what's going viral, and then apply that to
14:41 whatever it is you're doing. >> Yes. So, in the case of Wombo in the
14:48 summer of 2020, it was, well, Reef is just got a number one app and hundreds
14:52 of millions of downloads, and hey, there's this new model now, which is
14:56 still very hard to use. This is another very good signal, by the way. If
15:01 something is going viral inside of like some technical circle, but it's still
15:05 it's still difficult for a lay person to use, then uh if you can help the lay
15:11 person get access to that same technology and make it easier for them,
15:15 that's typically like a really good recipe for verality or a really good
15:21 sign that something might go viral. Um, so I'll just wrap up the the Wombo
15:24 story. Y >> from kind of that moment of inspiration, it took us seven months to actually get
15:30 the app on the app store. Uh, but we finally launched it in March of 2021. It
15:36 went insanely viral, got like 50 million downloads in its first month, uh, in
15:42 first 3 months. And um you know after sort of some just like stabilizing and
15:46 taking it all in and raising a little bit of money and building out our team
15:52 uh the obvious like next question was okay well we just did this with one AI
15:57 model. What other AI models can we do this with? And now this is a really
16:02 obvious question to ask in 2025 in the age of AI. Um again 2020 was still a
16:07 little bit different earlier. Generative AI was not even a buzz word yet
16:11 >> and generative AI is already a buzz word that we've moved past now which
16:15 >> what's the latest now >> I feel like now it just like AGI super
16:19 intelligence like agents >> you know uh so the next concept was dream which
16:27 was an artwork generator um for the nerds it predates midjourney and stable
16:33 diffusion and deli and was basically the first time that image generation went
16:37 viral now of course image generation is everywhere. Um and uh we saw something
16:43 very similar happening again almost the exact same thing we saw happening with
16:47 Wombo. Uh people were playing with a combination of two models VQ GAN
16:55 [ __ ] Vugan and Pashan I need your help. >> What's up?
16:57 >> I forgot what were the two models. >> Huh? >> Diffusion models.
17:00 >> It was Vugan diffusion. >> Vugan clip. >> Vugan clip. That's right. Vugan clip. So
17:07 you combine VQGan with clip, you could generate these crazy dreamlike images.
17:12 There was like maybe 20 or 30 people on Twitter um and really like five at the
17:17 beginning who were like messing around with this stuff. And those were like the
17:22 earliest incantations of like this era's neural networks for image synthesis. And
17:28 we did the same thing. Uh we wrapped that up into a fun and easy to use
17:33 mobile app. We made an an extremely optimized version of the model which had
17:38 fast generation times and uh we we we did some templating to
17:45 make it easier to create high quality images. Basically just adding words to
17:51 the prompt that the user gave us. And uh exactly the same thing happened another
17:56 number one app. And many of the lessons that we learned from Wombo about give the user as simple
18:04 to provide input. In the case of Wombo is a selfie. In the case of Dream was a
18:08 text prompt. Pre-process that user's input with your own input that kind of like structures
18:16 it and makes it work well. So in the case of Wombo, it was the driving video.
18:20 In the case of Dream, it was the kind of stylistic keywords that we were adding
18:25 to the user's prompt. uh make the inference really really fast and free
18:32 for the user. Um did that in both cases put all that together in a fun and easy
18:36 to use interface and you have something with like all the ingredients for
18:38 verality. >> Okay. And um like Okay. So it goes number one. What next?
18:49 >> Yeah. So just in terms of the trajectory of the company,
18:52 listen, I had no idea what the [ __ ] I was doing. Like I'm not like one of
18:56 these like Silicon Valley kids who like watched every Paul Graham video. I was
19:01 just like smoking weed in my bedroom playing League of Legends and like
19:06 really really interested in AI. Um so made every mistake we could. Uh but uh
19:13 the at that time now we're sort of at the end of 2021. We had a team of about
19:18 20 people all working here in Toronto. Um got a few apartments that we were
19:25 using as offices. Um, in 2022, we're like, let's raise money, cuz
19:29 that's what you're supposed to do. Uh, ended up putting together a series A uh
19:36 where the lead investor kept us in due diligence for 2 months and then pulled
19:41 out of the deal. and they pulled out of the deal partially because our original
19:47 Wombo app used music and there were some sort of unknown liabilities and risks
19:52 around our use of music. I don't know if I should bring it up with these mixie
19:56 guys, but [ __ ] the music licensing people are vicious, annoying
20:00 [ __ ] but also, you know, it is what it is. Um, and also partially because FTX crashed
20:09 and uh all this crazy [ __ ] was happening in crypto and
20:11 >> Zer was over. >> Yeah, there was like a market spook.
20:16 >> So, we went from having [ __ ] number one app thinking we're about to get $15
20:22 million to having negative runway and being like, "What the [ __ ] just
20:25 happened?" Um, >> negative runway meaning like you're out
20:27 of money. >> Yeah. Like >> or you're like really out of money. like
20:34 really out of money where and it really happened because that month Dream went
20:38 viral again and we had a million dollar server bill on Amazon for that month for
20:43 all the influence that we were doing and I had like maybe not the perfect
20:49 philosophy around this which which was like money is fake let's just make an
20:52 amazing [ __ ] product let's get all the users we'll figure it out later um
20:59 yeah didn't work at that point So, uh, we went into like a year and a half
21:05 of cockroach mode basically where I wasn't paying any of my bills unless I
21:09 absolutely had to. We were extremely focused on monetizing our product which
21:13 previously we had never given a [ __ ] about and basically getting to like
21:18 profitability and along the way we you know still raised some money from
21:22 investors who believed in us. Uh, so yeah about a year and a half of
21:27 cockroach mode got to profitability. And how did you monetize the apps? Because I
21:30 think >> because I think that's a question that a lot of people have, which is like, okay,
21:34 great. Now I've built this mobile app. It's starting to get some traction.
21:36 >> Yeah. >> What, you know, how do you think about
21:41 monetizing? Like what what's the sauce around monetizing mobile apps?
21:45 >> Just copy people who are better than you. >> So, who's the best? Who's the best? And
21:49 what are they doing? >> [ __ ] Nikita, man. >> Nikita. That guy is the
22:03 >> Shout out Nikita Beer. Yeah, go or I guess you can't do it anymore, but go
22:07 like pay $10,000 to talk to him on intro and I'm sure he'll give you some good
22:09 tips. >> Okay, so how did you how did you monetize it?
22:15 >> Uh we introduced a subscription and we introduced ads. The subscription did
22:18 things like, well, if you want to generate an image, uh, you could
22:23 generate four at a time instead if you're a subscriber, or you could act
22:27 have access to some premium styles if you're a subscriber. Um, and that took
22:33 us from zero to I think at our peak, Dream was doing over $500,000 a month.
22:39 Uh, and that got us to the profitability and sort of stability that we needed.
22:43 And and from a sub like what are we talking like $5 a month, $5 a week? I
22:48 know a lot of people are doing weekly subscriptions now. Like what was it?
22:54 >> We've done everything from $5 a week to $5 a month. We started really cheap and
23:01 got more expensive over time. And um there's also like all the like dark
23:05 patterns that you could do with like hidden trials and all that kind of
23:09 stuff, but we never went to too deep into that. >> Yeah. Okay. Have you I'm sure you
23:15 studied like Cal AI and what they do. >> Yeah. Yeah. Yeah.
23:16 >> What do you think? >> It's a new generation of killers, man.
23:22 >> I love to see it. It seems like every few months now there's like a new
23:26 18-year-old kid making $2 million a month on some simple stupid [ __ ] And uh
23:33 I think it's only going to accelerate. >> Mhm. Yeah. And and those apps are
23:38 monetizing via subscriptions, right? >> Yeah. Weekly, monthly?
23:42 >> I'm not sure. We'd have to double check. >> I think weekly. Yeah. I think a What I
23:47 That's a trend I'm seeing a lot is like weekly subscriptions anywhere from like
23:53 $3 to $20 a week. >> Mhm. Makes sense. >> Yeah. >> Cool. Okay. So, going back, so you're
24:02 you start making money basically. >> And then what next?
24:06 >> Yeah. So we slowly emerge out of cockroach mode. I guess bringing us into
24:13 2023 and 2024. The most notable things that happened uh were one, we had a
24:18 relationship with Nvidia. Uh they approached us very early on uh because
24:23 they saw that our apps were going viral and that we were doing this thing that
24:26 was going to be called generative AI and they want to kind of learn more about it
24:30 and see how they could support it. Of course, it's evolved like far far beyond
24:34 that now. Um and that culminated in uh us meeting them more and more and them making an
24:44 investment into Wombo. Uh the second thing that happened was I
24:50 sort of I lost some interest in making apps and I gained more interest in you know where
24:59 all of this was going, where AI was going. Oh man, this is going to be kind of a
25:04 convoluted point, but it'll make sense at the end. Okay, so one thing that
25:09 we've always cared about is inference efficiency. How do we make generation
25:15 times extremely fast and how do we spend the least amount of money that we can on
25:20 powering this inference? Uh so initially that meant we are going to host our
25:28 models on GPUs. We're going to sign a three-year contract with Amazon where
25:33 they give us some better pricing and we're going to pay uh specialists
25:41 $3,000 a day to make our model produce the same output but in a fraction of the
25:45 time and it's just worth it to find a specialist like that. That was sort of
25:49 the first solution. And and and then we're going to run all the inference on
25:53 the cloud. Our users never have to worry about it and it's consistent. It's high
25:58 quality. It's reliable. Uh then there was a motivation to move some of
26:07 this inference on device especially as your user base has devices in it that
26:13 are powerful and perfectly capable of running the model that you care about.
26:17 And so we started doing a little bit of that and sort of the natural extension of
26:23 that is well if there's someone in my user base with an iPhone 16 Pro or a
26:30 computer with a 4090 on it um that can run their own workloads,
26:34 what if they could run workloads for other users as well? And so that idea of
26:41 sort of peer-to-peer computing and using our users compute was born out of our
26:46 continuous push towards inference efficiency. Um hopefully all of that is
26:52 making sense so far. Uh and and so the other thing that happened was at the end of 2023 at the
26:59 start of 2024 I got really interested in how AI and crypto would intersect. And
27:06 so I looked at every project in the space, studied them closely, again
27:11 coming back to that same principle of copy what other people are doing or at
27:14 least understand what other people are doing >> before you do your own thing. And so we
27:18 looked at, you know, Bit Tensor and Jensen and Render and Akos and Golem and
27:26 the Internet Computer and Folding at Home and all these other sorts of
27:31 examples of decentralized AI or decentralized computing happening.
27:36 And I just ended up finding that extremely interesting. Um, and to make a
27:42 long story short, aside from learning about those projects and even building
27:48 in some of their spaces, uh, the conclusion we came to was to build our
27:53 own thing called W.ai, which would essentially allow anyone to turn their
27:57 MacBook, their gaming computer, and in the future their PlayStation or iPhone
28:01 or whatever into a compute contributor. And so, it's almost like Airbnb for that
28:05 device. when you're not using it, you can make it available to the world. It's
28:10 going to run some hopefully valuable workload and earn you a reward.
28:15 >> And reward being money. >> Reward today being points
28:19 >> which will later turn into a token >> which will later turn into money.
28:24 >> Uh I'm not like media trained so I don't know what what I'm
28:26 >> not media trained. I had no idea. >> No, I'm not [ __ ] media trained. Uh
28:30 but >> the idea is you are contributing to the network. Like if I just let if I let W.I
28:39 generate 10,000 images for somebody else on my machine, >> then I should be able to generate at
28:45 least 10,000 and1 images in return. And that seems fair. >> Yeah.
28:49 >> Totally cool. And that's the journey. >> That's the journey. A lot of crazy [ __ ]
28:55 has happened in between. Um cycled through the team many many times. We
29:00 started with six co-founders. Only one remains. Three of them have started new
29:04 companies that are very [ __ ] cool and interesting. Uh so check out Omega,
29:10 check out Dippy. That's the Wombo Mafia. And uh now we all live in a big ass
29:15 house in LA. There's like 12 of us who live and work there. Uh there's a few of
29:18 us who don't live and work there, but are still core parts of the team. Um but
29:22 yeah >> and and just to tie a bow on it. Okay. So I think that there's a huge
29:27 opportunity to build and I know you agree too build mobile apps that have an
29:31 AI component to it. >> Yes. >> A lot of you know like every startup
29:37 most fail. >> If you want to create a mobile app with some AI component to it
29:46 >> that gets millions of downloads. The way to do it is step one see how other
29:53 people see see like understand your landscape. What are the what are the
29:55 other apps? >> Go look at Caloni. Go look at Rayon. Go
29:59 look at all those apps. >> Right. Study them. >> Study
30:02 >> like understand that every single screen. Yes. Why they're doing it? How
30:05 they're doing it? What colors they're using? What typography? Everything.
5:22 that A16Z reace company, like a lot of people would look at that and be like,
5:26 "Okay, they're going to win this space." Like, you know, it's already going to
5:29 get competitive. But you kind of said, "No, no, no. I'm going to create an I'm
5:33 going to compete with them." >> Yeah. I wasn't thinking that I'm going
5:36 to compete with them. I was just thinking like it's early like there's
5:41 lots of space to play like I don't know like there's where my mind goes to is like uh like
5:50 like movie directors or or musical artists like we're never going to run
5:55 out of like new movie directors or artists like we're going to continuously
6:01 want new interesting [ __ ] So I I don't think anything is like too crowded.
6:03 >> Mhm. >> Yeah. >> Abundance mindset. Yeah,
6:09 >> Sam Alman, the co-founder of OpenAI, just said that it is the era of the idea
6:15 guy, and he is not wrong. I think that right now is an incredible time to be
6:19 building a startup. And if you listen to this podcast, chances are you think so,
6:22 too. Now, I think that you can look at trends uh to basically figure out uh
6:28 what are the startup ideas you should be building. So, that's exactly why I built
6:33 ideabser.com. Every single day you're going to get a free startup idea in your
6:39 inbox and it's all backed by high quality data trends. How we do it?
6:44 People always ask. We use AI agents to go and search what are people looking
6:49 for and what are they screaming for in terms of products that you should be
6:53 building and then we hand it on a, you know, silver platter for you to go check
6:58 out. Um, we do have a few paid plans that, you know, take it to the next
7:02 level. uh give you more ideas, give you more AI agents and more almost like a
7:07 chat GBT for ideas with it, but you can start for free ideabrowser.com. And if
7:11 you're listening to this, I highly recommend it. Um, so I'll just kind of
7:16 briefly walk through this. You can actually see that in this video, this
7:21 lady shows this old Google Collab notebook for the first order motion
7:24 model. So like my grandma could never use this [ __ ] Like I mean she could if
7:28 I like taught her Python. >> Yeah. but like she could never use that.
7:34 Um, so here we go. Uh, you would take a selfie. That's the first thing you do. Then you pick a song
7:41 and then you sit through a little loading screen like this. And then
7:48 finally, you get your output. And uh, we'll just see see that at the end.
7:53 obviously looks super basic by 2025 standards, but literally those three
8:01 screens of input, song selection, output, oh, sorry, loading screen, and then
8:06 output, four screens. That's all we needed to get 100 million downloads. And
8:12 uh we intentionally sort of designed something that would be extremely simple
8:18 to use. um sort of uh the whole thing is engineered for verality in a way because
8:24 the whole expectation is you as a user come in, you make a piece of content,
8:28 you make that piece of content extremely easily and extremely quickly and then
8:33 your natural inclination is to share or to make another one where like your dad
8:37 is the one singing or the pope or a politician or whatever. And so we we
8:45 spent no money on marketing and we spent all our money on inference and uh that's
8:51 that's how the thing went viral. Now maybe the only other thing I'll say
8:55 here and I'll kind of kiss my own ass a little bit here is the song selection.
9:00 >> Mhm. Uh we selected 15 songs that we launched with and these are some of the most
9:07 recognizable and iconic songs of all time or there were songs there were kind
9:13 of meme songs and so like you know one example of something that would make a
9:17 song a meme song is you don't need to know English to vibe with it.
9:21 >> Uh for example uh boom boom boom boom by the Wenger boys.
9:24 >> You don't need to know English. No >> sounds great.
9:25 >> Yeah. Absolutely. >> Wherever you are. >> Yeah. And so we we really deliberately
9:31 picked those 15 songs and then we really deliberately made the 15 driving videos
9:39 for those songs. And uh that was I think key >> and and he made it like stupid simple.
9:42 >> Stupid simple >> stupid like this is you can give this to
9:46 a 5-year-old and they'll know what to do. >> Exactly. >> A buddy of mine has a an app that's
9:51 going viral right now. Maybe you've seen it. Mixie. >> Mixie.
9:54 >> Yeah. Have you seen that? >> I think so. >> Yeah. So, it's like a mashup. Um, like
10:01 go to just just Google Mixi >> Mixie app. >> I think it's with an I. There it is.
10:06 >> Oh, yeah. I've seen this. >> Yeah. So, it's like two guys.
10:12 >> Um, you know, you put two beats together, it mashes it up. Like, simple idea, right?
10:21 But the UI is really simple. Um, and they're getting like >> I mean, it's one of the top music apps.
10:24 Yeah. >> Um, I don't know the monetization of this, but
10:31 I think there's this whole next there's this whole new generation of apps that
10:35 are just like really stupid simple. >> Um, that uh I mean it's a similar
10:41 concept, right? Like I'm sure his songs are like he's doing a queue of songs
10:45 that you know people actually want to choose, right? Um, it's simple UI. Um,
10:52 and it's viral because you want to share that on IG, etc. >> Yeah. I I personally
10:59 like starting with stupid simple >> and maybe from stupid simple you can do
11:04 something more advanced and interesting. >> Yeah. >> Um, but
11:07 >> with with respect to mobile apps, right? >> With respect to mobile apps.
11:10 >> And I mean it it all kind of boils down to attention at the end of the day. Like
11:15 if if you're trying to make a product, uh it's not just someone's money that
11:19 you're after. Like even before that, it's their attention. And there's so
11:22 many [ __ ] distractions. There's so many different things that are pulling
11:25 you in a million different directions at all times that I find the best way to
11:32 sort of introduce myself and get someone's attention for the first time
11:35 is to do something stupid, do something funny, do something simple, not try to
11:39 [ __ ] explain Socrates to them or something. Um, but just basically tell a
11:45 small joke or another analogy could be like give them a small piece of candy,
11:47 >> right? >> As opposed to like, you know, here's a
11:49 [ __ ] steak, >> right? Yeah. Because what I'm trying to
11:52 figure out, what I'm trying to reverse engineer is how do you create a mobile
11:57 app that gets millions of downloads? So, stupid simple one. >> What else? Uh so uh in our case there's
12:06 always been some kind of viral content creation loop meaning someone comes in
12:12 makes a piece of content shares that content on other platforms and that's
12:17 what drives your user acquisition. Um and so that whole uh that whole
12:23 component has a lot of extremely important details. Um so first is the
12:27 content creation. How are you going to help someone make a good piece of
12:31 content? Also, the game has changed since 5 years ago uh with how far
12:36 generative AI has come um and the kind of content that people are interested in
12:39 making, but still I think it's applicable today. Um so here's how I
12:45 would boil it down. How do you how do you help someone make a piece of content
12:48 that either they're going to post on Instagram and post on Tik Tok for the
12:53 algorithm and the algorithm is like that's a great piece of content. let's
12:57 give it millions of views or more for their own like personal networks and the
13:01 people in the personal network are going to be like wow that's really awesome I
13:05 want to make that thing too >> so you're basically saying you start
13:09 with the content like the end content which is kind of what my friend Joe is
13:13 doing with Mixie right cuz like he knows that people want to share these mixes
13:18 >> absolutely and so like you know one thing maybe we could touch on briefly is
13:23 just the algorithmic feeds of today like the amount of human attention that is
13:27 just being sucked into the full screen infinite scroll format whether it's Tik
13:32 Tok or Instagram or YouTube shorts or you know X has one like that is just
13:37 like the battlefield of attention and so I think like a great source of product
13:42 inspiration is to look at that and look at where are people's attention going
13:45 what kind of content are people making can I make it easier for them to make
13:49 some some piece of content that is consistently going viral over and over
13:52 again >> and then and then put my watermark in the bottom right.
13:57 >> Does this only work for like BTOC fun apps or could this also work for like
14:04 B2B or or proumer type apps? >> I think it could work for everything.
14:09 like you probably need to adjust your uh like principles a little bit and like
14:13 you know maybe I'm going to stop paying attention to the Tik Tok algorithm and
14:17 start paying attention to like I don't know the conference that all the people
14:22 in a particular niche are going to but it's it's the same thing like what do
14:26 the people you care about pay attention to >> right >> and where right
14:34 >> and and sort of the key the key point is like look at what other you you know,
14:38 look at the content, look at what's going viral, and then apply that to
14:41 whatever it is you're doing. >> Yes. So, in the case of Wombo in the
14:48 summer of 2020, it was, well, Reef is just got a number one app and hundreds
14:52 of millions of downloads, and hey, there's this new model now, which is
14:56 still very hard to use. This is another very good signal, by the way. If
15:01 something is going viral inside of like some technical circle, but it's still
15:05 it's still difficult for a lay person to use, then uh if you can help the lay
15:11 person get access to that same technology and make it easier for them,
15:15 that's typically like a really good recipe for verality or a really good
15:21 sign that something might go viral. Um, so I'll just wrap up the the Wombo
15:24 story. Y >> from kind of that moment of inspiration, it took us seven months to actually get
15:30 the app on the app store. Uh, but we finally launched it in March of 2021. It
15:36 went insanely viral, got like 50 million downloads in its first month, uh, in
15:42 first 3 months. And um you know after sort of some just like stabilizing and
15:46 taking it all in and raising a little bit of money and building out our team
15:52 uh the obvious like next question was okay well we just did this with one AI
15:57 model. What other AI models can we do this with? And now this is a really
16:02 obvious question to ask in 2025 in the age of AI. Um again 2020 was still a
16:07 little bit different earlier. Generative AI was not even a buzz word yet
16:11 >> and generative AI is already a buzz word that we've moved past now which
16:15 >> what's the latest now >> I feel like now it just like AGI super
16:19 intelligence like agents >> you know uh so the next concept was dream which
16:27 was an artwork generator um for the nerds it predates midjourney and stable
16:33 diffusion and deli and was basically the first time that image generation went
16:37 viral now of course image generation is everywhere. Um and uh we saw something
16:43 very similar happening again almost the exact same thing we saw happening with
16:47 Wombo. Uh people were playing with a combination of two models VQ GAN
16:55 [ __ ] Vugan and Pashan I need your help. >> What's up?
16:57 >> I forgot what were the two models. >> Huh? >> Diffusion models.
17:00 >> It was Vugan diffusion. >> Vugan clip. >> Vugan clip. That's right. Vugan clip. So
17:07 you combine VQGan with clip, you could generate these crazy dreamlike images.
17:12 There was like maybe 20 or 30 people on Twitter um and really like five at the
17:17 beginning who were like messing around with this stuff. And those were like the
17:22 earliest incantations of like this era's neural networks for image synthesis. And
17:28 we did the same thing. Uh we wrapped that up into a fun and easy to use
17:33 mobile app. We made an an extremely optimized version of the model which had
17:38 fast generation times and uh we we we did some templating to
17:45 make it easier to create high quality images. Basically just adding words to
17:51 the prompt that the user gave us. And uh exactly the same thing happened another
17:56 number one app. And many of the lessons that we learned from Wombo about give the user as simple
18:04 to provide input. In the case of Wombo is a selfie. In the case of Dream was a
18:08 text prompt. Pre-process that user's input with your own input that kind of like structures
18:16 it and makes it work well. So in the case of Wombo, it was the driving video.
18:20 In the case of Dream, it was the kind of stylistic keywords that we were adding
18:25 to the user's prompt. uh make the inference really really fast and free
18:32 for the user. Um did that in both cases put all that together in a fun and easy
18:36 to use interface and you have something with like all the ingredients for
18:38 verality. >> Okay. And um like Okay. So it goes number one. What next?
18:49 >> Yeah. So just in terms of the trajectory of the company,
18:52 listen, I had no idea what the [ __ ] I was doing. Like I'm not like one of
18:56 these like Silicon Valley kids who like watched every Paul Graham video. I was
19:01 just like smoking weed in my bedroom playing League of Legends and like
19:06 really really interested in AI. Um so made every mistake we could. Uh but uh
19:13 the at that time now we're sort of at the end of 2021. We had a team of about
19:18 20 people all working here in Toronto. Um got a few apartments that we were
19:25 using as offices. Um, in 2022, we're like, let's raise money, cuz
19:29 that's what you're supposed to do. Uh, ended up putting together a series A uh
19:36 where the lead investor kept us in due diligence for 2 months and then pulled
19:41 out of the deal. and they pulled out of the deal partially because our original
19:47 Wombo app used music and there were some sort of unknown liabilities and risks
19:52 around our use of music. I don't know if I should bring it up with these mixie
19:56 guys, but [ __ ] the music licensing people are vicious, annoying
20:00 [ __ ] but also, you know, it is what it is. Um, and also partially because FTX crashed
20:09 and uh all this crazy [ __ ] was happening in crypto and
20:11 >> Zer was over. >> Yeah, there was like a market spook.
20:16 >> So, we went from having [ __ ] number one app thinking we're about to get $15
20:22 million to having negative runway and being like, "What the [ __ ] just
20:25 happened?" Um, >> negative runway meaning like you're out
20:27 of money. >> Yeah. Like >> or you're like really out of money. like
20:34 really out of money where and it really happened because that month Dream went
20:38 viral again and we had a million dollar server bill on Amazon for that month for
20:43 all the influence that we were doing and I had like maybe not the perfect
20:49 philosophy around this which which was like money is fake let's just make an
20:52 amazing [ __ ] product let's get all the users we'll figure it out later um
20:59 yeah didn't work at that point So, uh, we went into like a year and a half
21:05 of cockroach mode basically where I wasn't paying any of my bills unless I
21:09 absolutely had to. We were extremely focused on monetizing our product which
21:13 previously we had never given a [ __ ] about and basically getting to like
21:18 profitability and along the way we you know still raised some money from
21:22 investors who believed in us. Uh, so yeah about a year and a half of
21:27 cockroach mode got to profitability. And how did you monetize the apps? Because I
21:30 think >> because I think that's a question that a lot of people have, which is like, okay,
21:34 great. Now I've built this mobile app. It's starting to get some traction.
21:36 >> Yeah. >> What, you know, how do you think about
21:41 monetizing? Like what what's the sauce around monetizing mobile apps?
21:45 >> Just copy people who are better than you. >> So, who's the best? Who's the best? And
21:49 what are they doing? >> [ __ ] Nikita, man. >> Nikita. That guy is the
22:03 >> Shout out Nikita Beer. Yeah, go or I guess you can't do it anymore, but go
22:07 like pay $10,000 to talk to him on intro and I'm sure he'll give you some good
22:09 tips. >> Okay, so how did you how did you monetize it?
22:15 >> Uh we introduced a subscription and we introduced ads. The subscription did
22:18 things like, well, if you want to generate an image, uh, you could
22:23 generate four at a time instead if you're a subscriber, or you could act
22:27 have access to some premium styles if you're a subscriber. Um, and that took
22:33 us from zero to I think at our peak, Dream was doing over $500,000 a month.
22:39 Uh, and that got us to the profitability and sort of stability that we needed.
22:43 And and from a sub like what are we talking like $5 a month, $5 a week? I
22:48 know a lot of people are doing weekly subscriptions now. Like what was it?
22:54 >> We've done everything from $5 a week to $5 a month. We started really cheap and
23:01 got more expensive over time. And um there's also like all the like dark
23:05 patterns that you could do with like hidden trials and all that kind of
23:09 stuff, but we never went to too deep into that. >> Yeah. Okay. Have you I'm sure you
23:15 studied like Cal AI and what they do. >> Yeah. Yeah. Yeah.
23:16 >> What do you think? >> It's a new generation of killers, man.
23:22 >> I love to see it. It seems like every few months now there's like a new
23:26 18-year-old kid making $2 million a month on some simple stupid [ __ ] And uh
23:33 I think it's only going to accelerate. >> Mhm. Yeah. And and those apps are
23:38 monetizing via subscriptions, right? >> Yeah. Weekly, monthly?
23:42 >> I'm not sure. We'd have to double check. >> I think weekly. Yeah. I think a What I
23:47 That's a trend I'm seeing a lot is like weekly subscriptions anywhere from like
23:53 $3 to $20 a week. >> Mhm. Makes sense. >> Yeah. >> Cool. Okay. So, going back, so you're
24:02 you start making money basically. >> And then what next?
24:06 >> Yeah. So we slowly emerge out of cockroach mode. I guess bringing us into
24:13 2023 and 2024. The most notable things that happened uh were one, we had a
24:18 relationship with Nvidia. Uh they approached us very early on uh because
24:23 they saw that our apps were going viral and that we were doing this thing that
24:26 was going to be called generative AI and they want to kind of learn more about it
24:30 and see how they could support it. Of course, it's evolved like far far beyond
24:34 that now. Um and that culminated in uh us meeting them more and more and them making an
24:44 investment into Wombo. Uh the second thing that happened was I
24:50 sort of I lost some interest in making apps and I gained more interest in you know where
24:59 all of this was going, where AI was going. Oh man, this is going to be kind of a
25:04 convoluted point, but it'll make sense at the end. Okay, so one thing that
25:09 we've always cared about is inference efficiency. How do we make generation
25:15 times extremely fast and how do we spend the least amount of money that we can on
25:20 powering this inference? Uh so initially that meant we are going to host our
25:28 models on GPUs. We're going to sign a three-year contract with Amazon where
25:33 they give us some better pricing and we're going to pay uh specialists
25:41 $3,000 a day to make our model produce the same output but in a fraction of the
25:45 time and it's just worth it to find a specialist like that. That was sort of
25:49 the first solution. And and and then we're going to run all the inference on
25:53 the cloud. Our users never have to worry about it and it's consistent. It's high
25:58 quality. It's reliable. Uh then there was a motivation to move some of
26:07 this inference on device especially as your user base has devices in it that
26:13 are powerful and perfectly capable of running the model that you care about.
26:17 And so we started doing a little bit of that and sort of the natural extension of
26:23 that is well if there's someone in my user base with an iPhone 16 Pro or a
26:30 computer with a 4090 on it um that can run their own workloads,
26:34 what if they could run workloads for other users as well? And so that idea of
26:41 sort of peer-to-peer computing and using our users compute was born out of our
26:46 continuous push towards inference efficiency. Um hopefully all of that is
26:52 making sense so far. Uh and and so the other thing that happened was at the end of 2023 at the
26:59 start of 2024 I got really interested in how AI and crypto would intersect. And
27:06 so I looked at every project in the space, studied them closely, again
27:11 coming back to that same principle of copy what other people are doing or at
27:14 least understand what other people are doing >> before you do your own thing. And so we
27:18 looked at, you know, Bit Tensor and Jensen and Render and Akos and Golem and
27:26 the Internet Computer and Folding at Home and all these other sorts of
27:31 examples of decentralized AI or decentralized computing happening.
27:36 And I just ended up finding that extremely interesting. Um, and to make a
27:42 long story short, aside from learning about those projects and even building
27:48 in some of their spaces, uh, the conclusion we came to was to build our
27:53 own thing called W.ai, which would essentially allow anyone to turn their
27:57 MacBook, their gaming computer, and in the future their PlayStation or iPhone
28:01 or whatever into a compute contributor. And so, it's almost like Airbnb for that
28:05 device. when you're not using it, you can make it available to the world. It's
28:10 going to run some hopefully valuable workload and earn you a reward.
28:15 >> And reward being money. >> Reward today being points
28:19 >> which will later turn into a token >> which will later turn into money.
28:24 >> Uh I'm not like media trained so I don't know what what I'm
28:26 >> not media trained. I had no idea. >> No, I'm not [ __ ] media trained. Uh
28:30 but >> the idea is you are contributing to the network. Like if I just let if I let W.I
28:39 generate 10,000 images for somebody else on my machine, >> then I should be able to generate at
28:45 least 10,000 and1 images in return. And that seems fair. >> Yeah.
28:49 >> Totally cool. And that's the journey. >> That's the journey. A lot of crazy [ __ ]
28:55 has happened in between. Um cycled through the team many many times. We
29:00 started with six co-founders. Only one remains. Three of them have started new
29:04 companies that are very [ __ ] cool and interesting. Uh so check out Omega,
29:10 check out Dippy. That's the Wombo Mafia. And uh now we all live in a big ass
29:15 house in LA. There's like 12 of us who live and work there. Uh there's a few of
29:18 us who don't live and work there, but are still core parts of the team. Um but
29:22 yeah >> and and just to tie a bow on it. Okay. So I think that there's a huge
29:27 opportunity to build and I know you agree too build mobile apps that have an
29:31 AI component to it. >> Yes. >> A lot of you know like every startup
29:37 most fail. >> If you want to create a mobile app with some AI component to it
29:46 >> that gets millions of downloads. The way to do it is step one see how other
29:53 people see see like understand your landscape. What are the what are the
29:55 other apps? >> Go look at Caloni. Go look at Rayon. Go
29:59 look at all those apps. >> Right. Study them. >> Study
30:02 >> like understand that every single screen. Yes. Why they're doing it? How
30:05 they're doing it? What colors they're using? What typography? Everything.
30:11 >> Step one. Step two is from a content perspective, understand what go is going
30:17 viral in a bunch of different niches, right? >> And and we can we can stay on this point
30:22 a little bit. I have two collections that I think of interest here. This
30:28 group chat on Instagram >> and this collection, which I can
30:31 actually make public and share with you if you'd like. Yeah,
30:36 >> this has been me sort of cataloging AI generated content or content that could
30:41 have been generated by AI which has gone viral on these respective platforms for
30:47 the last like two years. Uh so for example, let's go. There was like a a
30:51 week or two where I was getting a bunch of these baby videos.
30:55 >> Yeah. >> Like the von Yeah. I I've seen these
31:00 like the baby This is with Hedra, right? >> Yeah, this is with Hedra. And I think
31:04 they were using chat GPT to turn the original image into like a babyfied
31:05 image. >> Okay. >> And then Hedra to animate it.
31:10 >> Okay. Hedra.ai for folks who haven't checked it out. ChatGBT to create the
31:14 reference image and feed that to Hedra. Okay. >> Yeah. So I'll upload like a picture of
31:18 Heisenberg be like turn him into a baby and then take that baby picture bring it to
31:26 Hedra. So maybe someone, you know, for the week or two that this was going
31:30 viral or even now can make an app that just [ __ ] abstracts that away from
31:33 for people. They don't have to worry about chat GPT. They don't have to worry
31:37 about hedra. It's just, you know, take a selfie, choose a scenario, done,
31:43 >> right? So it's basically productizing the content format is what you're
31:45 saying. >> Yes, exactly. Yeah. >> Um, it's also just useful to study how this
31:55 stuff changes over time cuz it's the AI is not just transforming or creating
31:59 content in one way. It's doing a bunch of different things. >> Let's see.
32:03 >> So, is is your advice to people just like scroll and then just save?
32:08 >> Yes. >> If you want to brain rot then you must
32:11 be brain rotted first, >> right? True. So lately, we've been having a lot of
32:17 success with some really stupid [ __ ] Uh, and it's
32:22 turning celebrities, extremely recognizable celebrities, into pregnant
32:27 cats. >> Okay, like the Trump. Let's go open the Trump one.
32:33 >> Hey, Elon, it's time for our litter to Hey, Donald. Why do these kittens look
32:51 >> So, like, why does this work? It works because you have celebrities who are
32:55 recognizable. Works because it's kind of controversial. It works because
33:01 pregnancy is a concept that everybody can evolutionarily like understand. And
33:07 then male pregnancy is just something that shocks your mind a little bit. Um,
33:12 >> and it's top of mind, right? This is right around when Elon and Trump were
33:16 fighting. >> Yes. >> Right. >> Now, the other thing about that meme I
33:21 just showed you is we've repeated that exact same scenario, that exact same
33:26 story line like six different times and it goes viral each time. Um, including
33:31 this one that has almost 20 million views. Hey, Messi. It's time for our
33:42 [Music] >> Hey, Ronaldo. Why does this baby look >> So, literally the exact same thing, just
33:53 changing the characters, changing the settings a little bit,
33:55 >> right? >> And as a result, like we can pump that
33:58 out in like an hour. >> And so, that that got 18.8 8 million
34:02 views on Instagram. >> Yeah. >> And your goal is to get, you know, how
34:07 many downloads from that? >> Uh well, hopefully like even 1 or 2% is pretty good.
34:12 >> Yeah. >> Um right now there it's a weird time for
34:17 us where we don't care at all about our mobile apps and we care 100% about W.I.
34:23 >> And this content is not really about driving downloads to W.I. It's more
34:31 about keeping us sharp and uh on top of the algorithm. >> Yeah.
34:33 >> Like I think that's really important. >> Yeah. The reason I asked is just for
34:37 people who are creating content like this, what like what sort of conversion
34:42 rate should they expect? >> Yeah. I think 1 to 2% is like a win.
34:47 >> Yeah. And uh I think also a lot of the downloads are not going to like it's
34:51 possible that your downloads won't come from the video getting lots of views on
34:55 the algorithm, but from people sharing some piece of personalized content with
35:00 their friends. So like for example, let's say we turned this thing like
35:07 whatever this recipe is into an app. Like if I turn myself into a pregnant
35:13 cat, I might not post that on Instagram, but I might post that into like a
35:18 WhatsApp group chat and it'll go like that'll lead to some reality.
35:21 >> Someone someone in the comment section, I can hear it already now, is is going
35:26 to say, "Okay, I'm going to go create an app that turns you into a pregnant cat."
35:31 >> Yes. I'm on 10%. >> No, no, no. They're that's not even what
35:34 they're going to say. They're going to say, "And you think that you're going to
35:39 get millions of downloads and then you're going to be able to charge people
35:46 $3 a week for the privilege to do that? That's crazy. You are you guys are out
35:50 to lunch." That's that's what someone in the comment section is saying right now.
35:55 >> Yeah, I think uh I think that person is not entirely wrong, but they are wrong.
35:59 Like if you make an app and it goes viral, even if it goes viral once and
36:05 people never use it again, like that's a eight, nine, potentially nine figure
36:10 opportunity. Um, think about it like you're making a movie. Like if you make
36:15 a movie, people go to the movie theater and spend 20 bucks to watch the movie,
36:19 another 30 bucks to get a bunch of [ __ ] popcorn and [ __ ] and uh then
36:23 they leave and they never think about that movie ever again. So, like you're
36:27 just giving someone a bite-sized version of that that they paid $3 for instead of
36:33 50. Um, so yeah, I think you know the other thing is with these apps to get millions of
36:39 downloads, 98% of people are going to use them for free and you're going to
36:43 make all your money on the 2% who don't. The 98% are going to drive your verality
36:48 and keep making your content and sharing it with their families and friends and
36:51 getting you to the top of the app store. And then the 2% of people who don't give
36:54 a [ __ ] and they're just clicking buttons and they want the extra thing that you
36:57 can't get for free. That's where you're actually going to make all your money.
37:02 >> Cool. And let's maybe let's end it on, you know, for people listening,
37:07 you know, do you really believe that there's a lot of opportunity to to
37:12 create mobile apps with AI right now? >> Infinite opportunity.
37:14 >> Yeah. >> Yes. >> And why do you believe that? Because
37:18 people are glued to their [ __ ] phones and they want to be entertained and they
37:23 want utility. So if you can be useful or entertaining, then you can make an app. Do do they
37:28 need to raise venture capital for it? >> [ __ ] no.
37:32 >> And you know, how big of apps are we talking? Like could they create?
37:37 >> I don't know. Like it's >> with no venture capital.
37:40 >> Like how if if I make a really good song, how many people will listen to it?
37:43 If I make a really good movie, how many people will watch it? If I make like a
37:49 really good recipe for some gum, like it that it it depends on a lot of factors,
37:53 but if you make something truly great and truly unique and then kind of truly
37:57 captures the the current moment, then I think you can get hundreds of millions
38:02 of downloads and make at least tens of millions of dollars just with popcorn
38:06 mobile apps. >> Popcorn mobile apps. >> Yeah. >> All right. I like it.
38:11 >> Yeah. And good luck to everyone. I hope you make it. It's a [ __ ] crazy time
38:15 to be alive. Thank you for coming on, sharing some sauce, and uh
38:21 you fired me up to to create a pregnancy uh cat mobile app. >> I want 10%.
38:27 >> Later, dude.
$

I quit my job to make $6M/year with AI apps

@GregIsenberg 38:29 11 chapters
[AI agents and automation][content creation and YouTube][developer tools and coding][solo founder and bootstrapping][marketing and growth hacking]
// chapters
// description

Join me as I chat with Ben Benkhin, Creator of WOMBO, about how he built mobile apps that achieved over 250 million downloads by identifying emerging AI technologies and making them accessible through simple, user-friendly interfaces. His first app, Wombo, allowed users to animate selfies to sing popular songs, while Dream was an early art generator that predated tools like Midjourney. Ben emphasizes the importance of studying viral content formats and creating apps that help users easily produc

now: 0:00
// related videos (5)
Two Ways To Come Up With Business Ideas
[community building and audience growth][market research and validation][revenue model and pricing strategy]
→
Episode 2 - Building a SaaS to $1k MRR
[community building and audience growth][market research and validation][revenue model and pricing strategy]
→
Building a SaaS to $1,000 MRR - episode 3
[community building and audience growth][market research and validation][revenue model and pricing strategy]
→
Someone Will Get Really Rich Doing This
[community building and audience growth][market research and validation][revenue model and pricing strategy]
→
Episode 1 - Building a SaaS to $1k MRR
[community building and audience growth][market research and validation][revenue model and pricing strategy]
→
// actions
← back to terminal open on youtube → @GregIsenberg channel →
// tags
[AI agents and automation][content creation and YouTube][developer tools and coding][solo founder and bootstrapping][marketing and growth hacking]