The Startup Ideas Podcast - AI App Ideas that make Millions with Danny Postma

Episode Date: February 9, 2023

Today Greg is joined by Danny Postma, a full-stack developer whose build AI apps like This Model Does Not Exist, Profile Picture AI and Headlime which was acquired by Conversion.ai in 2020. In this episode, Greg and Danny talk about app ideas powered by AI that could make millions in the next five years. ►►Subscribe to Greg's weekly newsletter for insights on community,creators and commerce.You'll also find out when new and exclusiveepisodes come out from Where it Happens. And it's totally free.https://latecheckout.substack.comFIND ME ON SOCIAL:Twitter: https://twitter.com/gregisenbergInstagram: https://instagram.com/gregisenberg/TikTok: https://tiktok.com/@gregisenbergLINKS FOR THIS EPISODE:Production Team:https://www.bigoceanpodcasting.comDanny Postma:https://www.postcrafts.com/https://twitter.com/dannypostmaaSHOW NOTES:0:00 - Intro0:11 - AI Generated Influencers8:53 - Building AI tools as an indie hacker16:30 - Case Study: Profilepicture.ai18:46 - How to think about competition27:00 - More AI app ideas

All right. So let's start with, I think, one of the most interesting projects I've seen this year, which is this model does not exist.com. So I came across this. If you go to the website, it's this model generated by AI called A-LIS. Is that how I pronounce it? I guess so. Yeah. And it's this idea that it's an Instagram account with a new photo. of a virtual influencer but it's not like Lil Michaela if you've seen one of those that kind of look fake like it looks real you know she's got blonde hair she looks like I can touch her and she's a real person

I wonder how many guys have really fallen in love with her I'm sure many guys have fallen in love with her and one of the most interesting things about it is you have this section where it says help me pick tomorrow's photo so basically the way this product works is a new photo is generated by A-LIS every 15 minutes, you upvote your favorite photo, and the one that gets most upvoted goes on the Instagram.

Is that correct? Yeah, well, there was a plan, and then until people started uploading all the messed-up photos, so now it's basically like one of the most up-foted photos gets uploaded to the Instagram. Why? What happened exactly? People were just curious. The first photo is her with two hats on top of each other. which the AI likes to do sometimes.

Sometimes the eyes are a bit messed up, especially the hands. It's not good at hands. But I think there's only like 5% of the photos, 10%. The other 90% are pretty well off. But people like to upload stupid things in the internet, right? Think about Bodie McBode face. Right. And how did you create this product?

How long did it take you? Could you talk a little bit about the process of building this from concept to pressing publish? So building the product Like this product I did in I think a morning one afternoon But it's basically because I already built a tech Before it with ProVoPicture.aI

So I already had the upload functions I knew how to do the models So it was just basically building the website And basically what it does is I just grabbed a bunch of photos From StyleGun Which is a it's an AI That generates like millions of fake people

That don't exist It's pretty known, generated.photo. I think it's a website that uses the AI, for example. So just grabs a bunch of women that looked the same, uploaded them into my training tool, pick the ones that look like each other, trained it again, and basically then this one person that looks always the same,

just came out. And I've been using that to generate the photos with. Like, why did you do this project? Like, why is it exciting to you to build a virtual influence? there. So I don't like the typical way of doing marketing because it's quite boring.

I'm not good in it. I've tried advertisement. I suck an advertisement. The only thing I'm good in this SEO. So what I like to do is I like to create products, like launch mini products as a marketing. So I was thinking, all right, so my tool is profile pictures.

It's basically what this model product is, but then you can do it for your own face for other people. I was like, okay, what is a great way to bring some attention to it? And what is a fun project to hack on when I get bored? And I was like, maybe I should just create a fake influencer, let the community up vote for it. Maybe some other website would pick it up because it's something novel, it's something new.

And I basically just put a few banners every 15 photos to my main product and use it as a marketing channel. And it brings quite some revenue, which is quite fun. And the plan eventually was to like maybe. monetize the technology behind this. I might do it still one day because it's quite exciting. But for now, it's like it's on the back partner. You know, it's five years from now.

You're scrolling on Instagram. What percentage of your feed is virtual influencers like this? How fast this is going now? I'm pretty sure. Like, I think social media, how we know it is going to be dead in five years, because everything is going to be fake. So either people are going to love it or no one's going to use it because it's going to be a black forest effect.

Like, everything is AI. So why would you interact with a post if you can be 95% sure that this is a fake person that's hacking the algorithm? I wonder if, if you look at like Lil Michaela's Instagram, for example, Lil Michaela, I think is probably the most popular virtual influencer in the world. She still gets tens of thousands, if not hundreds of thousands of likes per post. Yeah, but I reckon it's because it's a novelty, right? It's one of a kind.

It's a one thing. What if, yeah, but it can go both ways. What if your whole feat is going to be AI generated? What if you're talking with AI, the comments are AI. Like, you don't even know if you're interacting with a real person or not. It can go both waves. Another way to think about it is in a world where you go into coffee shops and it's all robots, like it's just a robot arm serving you and making you a latte or an espresso.

if they make you an incredible latte, exactly how you like it, would you tip that robot? I mean, I'm from Holland. We don't tip, so I think that's the easy answer here. It depends, right? If you go to Starbucks,

you don't care if it's served by a robot or whatever, because it's like you want to go there, you know what a taste is, so you come for it. But if you're going for a local experience, you know, of course you want a barista, you can talk with. So, yeah, it depends.

I guess. My take on the virtual influencers and engagement with virtual influencers, my thesis is that people will still have that same level of connection with these people because they're invested in the story. Look at what's happening with. I'm sure you've seen what's happening with Twitch with that Seinfeld AI show. Rest in peace. Rest in peace.

Can you actually talk a little bit about how it works? So I don't know how it works So it would be like full speculations But I guess they just use Open AI GPT3 to just generate a lot of content And then have a narrator To turn its text to voice

I guess that's how it works It's basically for those of you haven't seen it It's basically an AI Generated endless Seinfeld parody So you go on Twitch It's actually banned right now It'll come back in a few weeks But that's another story

it's banned because well let me just get into it the the AI they had switched from chat GPT to another AI model and when they did the switch there wasn't the same level of content moderation so it ended up doing some horrible transphobic and homophobic jokes but basically people were going on this Twitch and they were seeing these AI generated version of Seinfeld and there were thousands of people commenting, thousands of people tweeting about it at any given time. And it was one of the most viewed Twitch channels. And it just goes to show you that there's this demand for AI generated content. And there's connection that exists between viewer and content.

Well, but don't you think that in five years, hypothetically, like, why do you need content creators? If TikTok, TikTok already knows so well what you like, what is your interest? like why would you need content creators if the AI models are so fast and GPUs are so fast. Like why do you need anyone to create content for you if you can just create it on the fly personalized towards you? I think that's where the future is going to go to. Yeah, I think if AI could create better content and I define better content as content that connects more with the audience, then there's no place for me and you to post on Instagram.

Yeah. Well, it's already happening, right? It's like these Minecraft video backgrounds with Reddit posts. That's basically being narrated and there's millions and millions and millions of views and people have no clue. It's AI generated. But basically there's one bot that just uploads thousands and thousands of videos and sees what gets sticky. And that's what gets shows to people. Yeah, exactly. So I think that's a trend that's just begun.

And I think with AIS, your product. this model does not exist. I could see this taking off. What does success look like for you with this product? So it's still my mind to turn this into a tool

to let other people create their own influencers, their own characters, whatever, with it. I think that's pretty interesting. People have already reached out on DMs to me, like how can I create my own?

So that's an interesting point, like an interesting way to go to. But for now, yeah, it was just like a fun project, man. I probably have ADD. I like to ship a lot of new products. I think it's just pretty cool that it's possible to do. It's like exploring new ways, learning new things,

seeing what is like the limit of AI right now. And it's mental. It's mental what you can do with it right now. If I were you, this is what I would do. Tell me. So I would build a product studio around your virtual influencers. So one way to bootstrap. any social account is if you do cross posting.

So I would partner with like a CAA, a UTA, a major talent agency, or so, you know, company like doing things media, which owns some of the most popular meme accounts, like middle class fancy and stuff like that on Instagram. And I would say our goal is to launch one new character influencer per month. that you're going to help cross-promote and then you partner with them and do a rev-share with them. I would do that because it just helps you get

to where you want to get too faster because your goal is, the more that this spreads, the more people see your underlying technology, the more revenue ultimately create. Yeah, this would be the easiest, organic, well, easiest would be like the way to get,

like a great way of marketing for this technique for this tool. The thing is, like, I have no clue, I know how Twitter works. I have no clue how Instagram works. I need to dive into it, see if it's interesting to me. Maybe I've got a partner with someone who is, like, interested in this whole technology,

in social media. Is that a downside of being an indie, an indie hacker where you're just one person? Is that, like, is there a cap to how big an indie hacker project could get to? I think it's more limitation of my personality. I get bored after three months. I used to be horrible in school. I couldn't focus.

I didn't want to do my homework. It didn't stick because I can only work on things that really interest me. But then I can go super, super hyper focus on it. And I just lose sense of reality for like a month or whatever. And I can just ship it out. And then one day it's like interest is gone. So I kind of force myself like you saying like it could be such a good idea. So if you're like an entrepreneur who who's not like me who can like focus on the things he doesn't like to do then.

But I know my weakness and my weakness is that I need to work on things that really interest me. You follow your curiosity, put another way. You're just following your curiosity. And I think the beauty about that is you're able to ship a product like this in a morning. Like when you're following your curiosity as an entrepreneur or a maker, it gives you superman-like abilities. It allows you to do something that is impossible. It's like when a parent sees their baby, like going into the ocean and they could swim like never before. That's exactly what it's like when an entrepreneur is following their curiosity.

My opinion, the downside of just following your curiosity is sometimes you do need some sort of lanes or constraints to scale products. And that's what I worry about when you just follow your curiosity. So that's a funny thing, because if you want to dive into that topic, I've realized that I'm good at starting products. I really like to obsess about it. But there's a limit to me where I actually need someone to run it afterwards, build it out afterwards. So I've hired two people, actually, one to do the operational parts of the company,

doing the contact creation, doing the live chat, so basically the operational side of the company to keep it going. And I've hired a developer two weeks ago, a good friend of mine. And he's basically, I build out the products, I'm exploring it,

I'm launching new ideas, and he is hired to build out once I, let's say bluntly, once I get bored of it, and he builds out all the more boring features, continues iterating and developing on it. So I can move on to the next, idea and see how that works.

I like that. I think that's going to be a big unlock for you. And I think once you start seeing success and you're seeing success with a bunch of your projects, which I want to get into, but once you start seeing traction, it makes sense to either hire agencies, hire contractors, build out your little team, as long as you can maintain that like, you know, follow your curiosity. Like if you can just spend all your time following your curiosity, like that's probably ideal for you, right?

Well, but it's a deliberate decision, right? Three years ago, I built headline two years ago. I had talks with Andries and Horowitz, Sequoia, stuff like that. Could you tell people what headline is? So people probably heard about Jasper, copy.competaI. They are, how do they brand themselves? I think they're copyright. They call themselves copywriting tools so they can write blocks for you, emails,

Facebook ads, Twitter ads and stuff like that. I was one of the first to. build a product like that and go live with it. I got a screenshot from OpenAI saying I'm the first developer to actually go live with a product. This was back in December 2020, 2020. And it massively blew up, man. Like, people wanted this. This was when Open AI just went live with GPT3, basically what JetGPT is built on top of. And it just scaled. It went like, it doubled in revenue every two, three weeks, hits a shit ton of revenue in three months. And then it was the decision.

There was so much competition. Am I going to take startup funding, build out a team, hire 50 developers and basically become a CEO. I had an acquisition offer so I could just get a life-changing amount of money, sell the company and do something else. Or I could try to bootstrap it. And I sat down for a week to weeks, talk with my friends, my parents, my wife, wife now. I just knew like it wouldn't be my life to work 80 hours, manage people. It's not going to make me happy. So how am I going to do that?

So I decided to sell it at the time. Yeah, I mean, the C and CEO doesn't stand for curiosity. I think that in general, like, CEO positions are romanticized. You know, you're quote unquote, the boss. But I actually think that like if you're a maker, you're a builder. I hate to be the boss man. Yeah, it's not fun. So I love that you did that.

And you were mindful, right? The great thing about big decisions, crossroads in business life and in life is it allows you to stop and think and ask yourself, what do you want? You need to figure out why are you an entrepreneur? What are you doing it for? Why are you building this company? Yeah. So this model does not exist is based on a product called ProfilePicture.AI, which I love you to talk about. It's massively successful.

It's got thousands of customers. So provo picture. That AI is you basically upload 20 photos of yourself A specific AI of yourself gets trained on your photos It knows how you look like

It can basically spit out anything That looks like you So you can make a prompt That says, okay Danny as a male warrior In Game of Thrones And it will just make a picture for you That looks exactly like you

In the Game of Thrones style Could be an anime style. Anyone already, like everyone knows Lenza. So Lenza launched one month after I launched this product. Coincidence. They did it really well. I think they earned like $30 million on the App Store right now. Should have made an app.

So yeah, but they do it pretty simple and I'm actually trying to build it out into a bigger product. So we're doing LinkedIn headshots basically trying to turn it into a photo studio right now because it actually outputs photo-like quality. Someone shared it on Twitter when he was asking

followers like where do you think I made this photo and no one guess it was AI it's it's pretty mind-blowing that it's possible to just create photos of you so yeah it's been pretty successful didn't expect it i thought it would be a hype but it's pasta hype now three months i think a lot of people ask me Greg how do you build products that foster community well i've got good news that's exactly what late checkout does my kind of company. We partner with the largest brands in the world and fast-paced startups to design products that resonate with your community. We add a couple interesting clients every single year. So if you're interested, and that sounds like you, email front desk at laycheckout.

Dot Studio with what you're working on, what you need help with. And don't forget to mention the Where It Happens Pod. Thank you. When you're building any startup and you see competition, especially well-funded composition, like Lenza is highly funded, Silicon Valley-funded, Silicon Valley-based, I think. You get that bad feeling in your stomach. At least I do when I see a competitor come out with a similar technology or product. How did you feel when you saw it? And how do you think about differentiating your product going forward?

So with Headline, I freak the fuck out every time a big competitor got VC funding, got exposure, got angry that they stole my ideas, blah, blah, blah. Basically, after I sold it, a few months later, I just saw the competition I was afraid of, didn't go that far. But also what I realized is, is like, if you want to stay small, your bootstrap or your indie hacker, it's only good if a competitor built something like you because they have the money to. to create awareness. Like for Lenza, for example, like I got 10 times more sales after they launched a product because suddenly people,

like most of the people, they were suddenly searching for like AI avatar, right? But some people, they didn't use the word AI avatar. They used AI ProVoPicture. So where do they end up? Because my domain name is ProVoPicture. They end up my website.

So now these days I think like, okay, yeah, it's great if some big shots, marketing, money, spender, build something like me because they do the marketing for me. I'll eat the scraps. I'm happy with it.

I don't need to do 30 million revenue because I don't have any employees to pay for them. So I only see it as something positive right now, to be honest. I also think that you, again, you're at a crossroads and you can decide where you want to take the product. So where do you think Lenza is taking it and where do you want to take it? I think what Lenza is eventually going to do,

but I'm pretty sure Instagram or TikTok is going to go there. It's like you will have your own editor of your own. face where you can just type in what you want to see right now and it just outputs the photos for you. I think that's where they might have to. So I've decided not to go that route because there's going to be a lot of competition. I really like to work on new ideas and build it out, but I realize there's so much potential in this, like the whole provol picture. It's massive for some reason. I shouldn't share too much about it, but there's like so much volume about it's a good way to just niche down and just focus on that part.

Just get really good in it. What I realized is I like to build companies that are 100% automated. So I want to bring this to a position where it's 100% automated. Basically only has to do live chat, which my wife does. And then I can move on to another product. SEO traffic comes in. It earns five, six figures a month. And I'm happy with that.

And then build the next product. Add it to the portfolio. And maybe in 10 years, have like 100 AI products. doing the same amount of revenue. It's a really good realization that you had that I think the, you know, the Instagrams, the TikToks are obviously looking at the success of products like yours and Lenza and are like, how do we get involved here? So it might get commoditized.

And the beauty about being an indie hacker, not a bootstrapper, not raising money is you don't need it to be super big for it to be life changing. Which is that that's also like a decision with the indie hacker, right? like you could build it into a big company. You could go to millions and millions of visitors. But that would also mean like if I get million of customers, like how am I going to handle the whole support load and everything? It means I have to hire a whole team.

So it's all deliberately trying to, like, of course I would love to have 30 million sales, whatever tonight. But that would mean I have to build a bigger team. I don't want to manage the whole fucking team. I've got two amazing people around me now, which is my friend and my wife. like I trust them. They're super smart people

and it goes easy because I know they're they know what to do themselves without me having to manage them. So it's perfect. I'm still doing the same thing. I just get some help with it and that's like a deliberative choice.

I love that. I think what's your take on it? What would you do in such a situation? Would you be the one that grows it out bigger? Is there a middle way to go? In a space like this, which is there's a ton of eyeballs on it,

I think that most likely either Instagram wins, like the platforms win, or Lenza figures out how to become the platform, or the platform integrates with Lenza or your product to power it. So those are the scenarios I see. It's very difficult to become a platform. I'm like, for you guys to become like an Instagram competitor or Lenza to become an Instagram competitor, it's remarkably difficult because of how difficult it is to build social products. You know, a social product, you have two sides, right? You have the content creation and then you have the consumption. Although I would argue that building AI based social networks is a lot easier than building

regular social networks. Because if you're building Instagram in 2009, you have to figure out how can I get really interesting content creators on one side. How do I get the right people on and to post the right content so that they interact with it? In an AI world, I mean, you still have to make sure that the content is high quality. People are getting engaged with it, but it's easier to bootstrap the network, so to speak, with AI. But it's still- You can make 10,000. You can make 10,000 A-lis who are posting fake posts on Instagram, right, and bootstrap that way. That's right. However, I do believe that people are going to assume creating virtual celebrities, virtual influencers are, oh, I just put a prompt in and that's it. It's actually going to be a lot harder to develop great stories, great concepts that resonate with people. But yeah, I don't think building a platform is too difficult. If I were you, I would focus on just how do you niche down into something that realistically,

the Instagrams of the world or the lenses of the world are not going to do, but to you could be literally tens of millions of dollars a year of revenue, which would be significant. I don't know the answer to what that exact product is, but I would do the product design sprint around what is something that, like what is something Lenza wouldn't do? Well, I guess they always want to grow bigger, right? So they're going to go broad, they're going to use broad terms.

Yeah. If you think about SEO. So if you focus on a niche keyword, I think that's a really good thing to do. Any startup is going to bloat, and it's the same with any other tool. Like if you can just take a specific little part of any big tool that people love and you're not going to add a shit ton of features to it because you want to grow bigger, then you have a reason to exist. Yeah.

I also think what you're doing is really smart with, you know, building your own virtual influencers with this model. does not exist. That's an example of like Instagram isn't going to create virtual influencers. Like that's too quote unquote small for them. creepy. And creepy too probably. And they can't do it. Yeah. And Lenta too is probably not thinking about that. But I think that is it crazy that a virtual influencer brings in 500,000 to a million dollars a year of ad and affiliate revenue? It's not crazy. And then could you put together 10, 15, 20, 30 of those. It's not crazy. And then could you generate $5, 10, $15, $20 million of people learning

about profile picture.a.I because they see your technology and they want to do it for himself. Not crazy. Maybe I should put some more love into this model that doesn't exist. Well, I mean, that's why I DMD you on Twitter. I was like, this is a guy who understands where the world of AI is going and is building products to speak to it. that are very bespoke and interesting. I'm curious how his brain works. What other products through your product studio have you built? So I build a typical, build a stock photo website that's fully ran on on AI.

I don't know if you saw it, stock AI.com. Actually, just shut it down. Basically, the whole team around all these products I'm trying to build right now is you're trying to replace something where you need a lot of people for and you just automated with some robots. just love building robots that work for me at night doing things that wasn't able to do before. So with stock AI, the thing was, hey, all these stock photo websites, they need photographers to make content for them, right? So it's like, what happens if you just build a robot,

train a model that's really good in stock photo, style photos, and just have it run 24-7 and spit out new photos. So generated like hundreds of thousands of photos. The problem is for now, it's like the quality is too low to be able to use. use as stock photography. It gets a novelty. It works for the profile pictures. It worked for the phases because it doesn't have to be perfect. But with stock photography, if there's

one thing off, no one wants you used it on the website. So what I did is I trained a whole model on it, tried to do it, but it didn't get any engagement. They got a lot of social press, a lot of visitors and everything. Though the conversion rate to a paid product was like 0.004%. And it's not cheap to run AI models, man.

So if you want to keep this up, I had to kill it. But I think someone is probably shutterstock is going to do this really well because they have the whole database of photos, right? They could literally, I think, a two years, put all their photographers out of a job and just having AI generate the photos for them. Just diving into the reason why this failed, is it because you didn't have a big enough data set versus eye stock photo who has like a massive?

I think the quality is not there yet. That's the whole issue. So you can generate all these photos right now and have a database, but if the quality of the models and the AI is not good enough yet, it's like such a waste of time, money and resources. Because no one, like people were barely downloading them either. So I think it's too early yet. And I think it's like you have unsplash, which has millions and millions and millions of free photos, which are really good high quality. That's what you're competing with with a stock photo website. So if your quality is not on par, then why would people use your.

tool in that sense. So even if I stock photo were to launch something today, what you're saying is it probably wouldn't be able to compete. Especially for realistic ones. I think like mid-jorni is getting really close to being really good

in artistic quality, like paintings, that it could replace that but for realistic photos, for real photography, it's not there yet. So if ProVopicture.com. for example, I need to make sure that the hands are hidden under the table in the pockets of the jacket because it sucks in making hands.

All these little limitations that it's like, you could use it for a profile picture because it's fine. You can crop it out. But the AI is not there yet to replace the eye stock photo, get the images or whatever. So this is one of those ideas that you did stock AI that's basically a timing thing. So is this something that you would potentially revisit in six. 12, 18 months. Yes, with two caveats.

I want to see who wins Getty versus stability AI. Because Getty is in a lawsuit now with the creators of this open source model. I don't want to get sued as a one person. Basically, I was going through my head too. I was like, if Getty drops a lawsuit on my doorstep,

which they would do if I grow hard enough, how I'm going to fight it? So I have to kill a product. So it will be a waste of six months working for it. Why does Getty feel like they could run a lawsuit here? Like, what grounds do they have? So this whole stable diffusion AI is trained on a 5 billion image data set called Layen, I believe.

I don't know if you know the name. And I think they found a few of their photos in the dataset. So they were like, all right, you trained on our copyrighted. photos, time to sue, because they're probably fucking scared that they're going to lose their money, right? Plus, it's trained on their copyrighted images. So I guess, yeah, they have a point if that's true. You cannot prove it because the black box of AI. And Getty is a really big business. They're about a two and a half billion dollar business. I believe they're probably traded. And I noticed that I stock photo also has been teasing out some of their AI that they're doing.

My stock photo is owned by Getty, right? Oh, is it? Okay. I think it is. So basically sounds like the stock photography, besides from unsplashed space, is basically monopolized by Getty. With shutterstock having probably a little dip in the toe, but for the rest, yeah. You know, would you ever consider, you know, reaching out to Getty or Shutterstock and being like,

hey, I built this and maybe there's a way we can partner? They're going to do it themselves. They could probably acquire a company that's going to do it. do it, but it's not, like we said, it's the focus. It's not my interest. I think my interest is right now in this whole official one, training on the face has seen, getting to see where you can take these fake Instagram models and whatever in the stock photo. It doesn't, it doesn't interest me that much. It does. It also like, I build in public, it doesn't get that much engagement on Twitter either. It doesn't get shared that much because it's pretty boring, like who the fuck likes stock photography,

right? It's more interesting to do like some interesting businesses. Could you talk more about tattoos.ai? It was the first one I made actually. It's actually getting some sales. So what I did is I've made a little editor. You can basically type in, okay, what kind of tattoo do you want? Let's say a bunny with a black hat. My AI adds some prompt magic around it.

You can select a tattoo artist style and it comes up with a flash sheet of tattoos. there was actually a guy who tattooed one of my tattoos that were generated on his leg, which was pretty dope, to be honest. And how does the pricing model work? It's $9. You get access for 14 days. I think I get one or two sales a day. It's starting to rank on Google, which is quite fun, because every tattoo that gets generated,

I take the prompt, turn it into like a search engine, and then it generates automatically the pages for it. So you could have like bunny tattoos, shark tattoos, wave tattoos. So yeah, Google will pick it up if people search for it. I think that's quite interesting with AI. You can turn out so much content, ranking Google with it and in that way gets search volume to your websites. It's interesting that you talk about SEO because at least the people I follow on Twitter who are in the AI space, they're constantly tweeting about how, you know, SEO is dead, SEO is dead. And I couldn't disagree with them more.

Could you talk more about the role of AI and SEO? So if you know Chet's GBT, right? Probably everyone in this podcast knows it by now, right? Yes. You're not going to go to Chad GBT and be like, hey, I want to make provost pictures. It cannot. Like, there is search intent. search and there is information search.

Information search is going to be completely disrupted by or chat to GPT. Why would you go to Google and ask for a recipe? You can ask the AI that it will come back with some content. But if you're looking for a solution for like making a profile picture, getting a tattoo, whatever, like you're still going to go to Google to find a website

that can do that for you. And there's no way that AI can solve something like that. Does the current way of AI tools make it easier, harder to rank in Google? I don't write my copy that much with those tools because I'm afraid Google penalizes it. I just think it's easier to make content in that sense. Because for example, I have a website landingfolio.com. I sell templates for websites there. I had to hire a developer, a designer,

wait for six months for them to make content. And only then you can start ranking on Google because you made the product. But with AI, you can literally on the fly, produce a content, a user goes to your website, generate something for themselves, and you use it to publish on your website. So your content gets created for you automatically.

So I think it's super easy to grow on social media in that sense. Like, for example, the tattoo, like that content rings. Google likes content where people stay. So if they don't bounce after like one second and say, okay, yeah, this is garbage quality. I don't want to be here. Like, you're going to get penalized, but apparently it's good quality. So Google sees that. You stay on the website for one minute.

You don't go back to the search. So you're going to stay there. Yeah, I think like, you know, the secrets to SEO is exactly what you said is getting people to stay on page. Well, you're going to make something they want and then keep them on the page. You just have to be, you, yeah, there's no fucking heck with it. I think it's such a bullshit. Sorry for the words, but you just need to make things for people that they want.

Use the keyword what they're looking for, which you put on the banner of a supermarket. Like you're searching for a supermarket, you see a supermarket sign, right? And people come to your product. And have other reputable websites point to your website that says, okay, you know, I'm the New York Times and like I'm linking to it.

So I think that having, press, having social links is really, really impactful as well. So, and then a little, if I may a little side note with that, I think that's, that's one of the reasons why you build in public, because it generates press for your new products, which means you're going to have instant backlinks, like on the day you launch. And Google's just going to rocket it to the top of the search engine instead of that you have to wait for like six months before they pick it up. Like people say, oh, you do it to sell your product.

I know it's literally, I just want backlinks. I like to share what I'm building. I like to interact with the community with people I meet. This is how I meet you, right? Which I think is really cool about Twitter. But it's really backlinks. That's the magic use for it. What's really cool about SEO in general is, you know,

it's a very boring way to get passive income, right? Like you invest in SEO. You don't invest in Facebook ads. You invest in SEO because in 12 months, 18 months, 24 months, you're seeing the dividends of all that building public stuff that you've done, all that content you're generating. So I'm excited about how AI is going to help builders, especially indie builders, create more content than they've ever been able to create.

I think if you're making blog post now to rank to the number top of Google, like content marketing, you're done, you're fucked. But if you make tools and products and something like this and that's how you rank in Google, I think you have future success. So tattoos AI, you know, there's already someone with a literal tattoo. Your models have created. You know, do you envision a world where you walk into a tattoo shop and you prompt a robot and the robot, just like we were talking about in the coffee shop, it's like a robot arm. Instead of serving it serving you an espresso or a cappuccino or a latte, it's just starting to. draw on you in permanent ink? Like, are tattoo artists going to be out of work? I genuinely, I've had a lot of discussions with this.

Like, you're not going to get a machine to touch you. That's, I would be so scared, man. Like, there's going to be two types of tattoo artists. Like, some of them obviously are going to be scared because there's suddenly an AI that can do the design. Then on the other side, I think there's going to be a lot of artists who are going to be. use it as an inspiration source. They're going to generate, like, there's going to be a customer.

They can just suddenly generate like 100 options instead of just wanted to, show it to the customer, and then they fine tune it based on what the customer really wants. So I think it's going to be like an inspiration tool, a way for them to communicate faster with the customers and then get to the tattoo they want. Why do you say that you can't imagine a machine tattooing you? Because like, To me, I can't imagine a self-driving car because the logic I make is airplanes are basically self-driving. Like, the pilots are there just to supervise, basically. There's probably more room for human error than there is machine error, especially as we, you know, in two, three, four, five years. So why do you say that you can't imagine the majority of tattoos being machine, machine led?

Might be because I'm scared of needles. So I'd be like even more scared if a rogue robot is going to tattoo me and I'm going to move my arm and it doesn't register that I'm moving my arm so it's going to start to step me or something. Such a wrong thing to think about because there's probably a way less error prone for a robot to do those things. But I want them, man. Would you let a robot tattoo you? I wouldn't let anyone tattoo me because I don't want tattoos. But if I was getting a tattoo, you know, if you're, uh, Richard Branson and you're like, hey, come with me to the moon.

I wouldn't come with you to the moon. I don't want to be the first person or one of the first people to do any of that stuff. I'm one of those people that kind of like, let me wait and see to see like how the first 20,000 tattoos have turned out and like do my due diligence. And then from there make a decision. I do think, though, that the hour of progress is towards machines. driven tattoos. I think that's where it's going to go. What you're probably going to see is it's going to be crazy at first.

Someone's going to do it. I can picture the press release right now or the New York Times article, whatever. Then it's going to be less crazy because more people are going to do it. And you're going to have humans supervising the machine. And then eventually there won't be human supervising the machine. Most tattoos will be done by robots and a small percentage of them will be done by humans. really curious kadoe because it's it's like it's an artistic process in that sense that the artist matters like my wife has a lot of tattoos and she searches for the artist with the style the way of doing it

so i think the tattoo artist it's not just hey i want a picture on me i find a tattoo artist to do it for me but it might also be because that's how it works right now and in the future like how about this like right now i'm i have this like cottage that i'm in in a mountain town and in Quebec in Canada. And there's, you know, probably about 10,000 people who live in this town. And let's say I go on the internet one day and I find this incredible artist in Amsterdam. And for whatever reason, I can't get to Amsterdam. I'm just in my town. Wouldn't it be amazing for me to be able to walk into my main street in my little town, walk into this like robot store and basically just like how you can reuse prompts, just go in and be like, hey, I love this artist.

What would this artist do in the addition of designing my dog or my dead dog or, you know, passed away relative or something like that on my on my shoulder? But wouldn't that be the same as going to a shady, shady store, buy some fake Pokemon cards and play with your friends with your fake Pokemon cards? Like, yeah, it looks the same. it is the same but it's not the same in that sense like okay yeah you got the same style as the artist but the artist didn't do it does it still give you the same fulfillment you could go to a local tattoo artist be like hey i want this exact tattoo from that artist like yeah but it might be

might be difficult to do i mean i i can't answer the question like as as someone who doesn't do tattoos like or doesn't like it's quite hard for both of us it's hard both for both of us but i i do think that if it's done right and if it's blessed by the artist, I could see a world where someone is excited to get that tattoo. That's an interesting future, man. It's an interesting future. Your studio is called Postcrafts. People could check it out at postcrafts.com.

What's one thing you want to leave our listeners with? Let's say you just want to start out, build a new product. You don't know where to start. I'd say go to hdrafts.com. Keep on typing in keywords of things that you might be interested in until you find a collection of keywords that has a keyword difficulty under 10, which means you should be able to go to the first page of Google, build a product around that that could be automized, where you could make some content with, that can rank on the website, build it with no code, without with code

or whatever, and you should be able to start a site business for yourself. that's the thing. I think a lot of people, they know how to build things, they just don't know how to launch it, how to market it or whatever. And I think SEO is the easiest way to start with it if you find some niche to operate inside of.

I think that will be my tip. I think everyone should launch something. Yes. Just putting it out there. Like, man, just like build something to scratch your own it. Like my first tools, I build a headline for myself because I sucked in making headlines.

So I build it for myself. And like, if you're lucky, someone else needs it. If you're lucky, really lucky, thousands of others need it. And if not, you build something for yourself that makes your life easier.

Yeah, absolutely. I love that. People can go follow Danny on Twitter. He's a must follow, especially if you're interested in the world of AI and your account is what? At Danny, P-O-S-T-M-A-A.

Yes, Danny Postma. with the XIA. And I build in public all my startups, tweet about AI mostly these days, tweet about what I'm building, testing and everything. So give me a follow. Come back anytime, Danny.

Thank you so much. It was a great chat.