Debug Information

Processing Details

• VTT File: 127ec3a0.vtt
• Processing Time: September 11, 2025 at 01:02 PM
• Total Chunks: 2
• Transcript Length: 89,725 characters
• Caption Count: 889 captions

Prompts Used

Prompt 1: Context Setup

You are an expert data extractor tasked with analyzing a podcast transcript. 
I will provide you with part 1 of 2 from a podcast transcript. 
I will then ask you to extract different types of information from this content in subsequent messages. Please confirm you have received and understood the transcript content.

Transcript section:
[00:00:06.640 --> 00:00:11.840] I feel like there's something in the water of being an indie hacker where you feel like you need to be really friendly.
[00:00:11.840 --> 00:00:12.960] Like everyone's a community.
[00:00:12.960 --> 00:00:14.400] Everyone helps out.
[00:00:14.400 --> 00:00:19.520] So people don't think about having rivalries when that really is.
[00:00:19.520 --> 00:00:23.680] I mean, if you take Danny and Peter's situation, they aren't enemies.
[00:00:24.320 --> 00:00:25.600] There's no bad blood.
[00:00:26.240 --> 00:00:27.600] You can be a rival and be friendly.
[00:00:28.160 --> 00:00:29.840] Yeah, you can be a rival and be friendly.
[00:00:30.080 --> 00:00:41.520] And then on top of all that, there's also like, Courtland, you probably have seen, like, I mean, I think I've counted no fewer than 10 individual posts on indie hackers complaining about copycats.
[00:00:41.520 --> 00:00:43.920] Like, people just have this idea that you can't copy.
[00:00:44.480 --> 00:00:46.560] If someone else is doing something, like, that's their lane.
[00:00:46.560 --> 00:00:47.760] You got to stay in your lane.
[00:00:47.760 --> 00:00:53.760] And there's no bigger bullshit in the world of business than the idea that you can't have the same thing.
[00:00:53.920 --> 00:00:57.280] It'd be like playing basketball and somebody steals the ball from you.
[00:00:57.280 --> 00:00:58.000] And then you complain.
[00:00:58.160 --> 00:00:59.680] Like, yeah, that's not, that shouldn't be allowed.
[00:00:59.680 --> 00:01:02.480] It's like, actually, that's like part of the rules of the game.
[00:01:02.480 --> 00:01:07.200] Like, one of the rules of business is that if somebody innovates, like, you can do a similar innovation.
[00:01:07.200 --> 00:01:09.600] Someone does a slam dunk and you're like, well, I can't do a slam dunk.
[00:01:09.600 --> 00:01:10.640] Like, that's Michael's move.
[00:01:10.640 --> 00:01:12.160] You know, like, I got to do the crossover.
[00:01:12.160 --> 00:01:13.600] Like, I can't do that same thing.
[00:01:13.600 --> 00:01:14.080] Yeah.
[00:01:14.080 --> 00:01:16.080] Speaking of Danny, Danny, what's up?
[00:01:16.080 --> 00:01:16.880] What's up, guys?
[00:01:16.880 --> 00:01:17.760] Thanks for having me.
[00:01:17.760 --> 00:01:18.000] Hello.
[00:01:18.160 --> 00:01:19.520] Thanks for joining.
[00:01:19.520 --> 00:01:23.600] We're just talking about you and your sort of AI rival, Peter Levels.
[00:01:23.600 --> 00:01:25.040] Do you think it's healthy to have a rivalry?
[00:01:25.600 --> 00:01:27.360] Would you call it a rivalry?
[00:01:27.680 --> 00:01:36.240] I wouldn't call it a rivalry because we've been in Telegram chats for 24-7, just literally sharing all the information we had together when we launched the Provo Picture apps.
[00:01:36.240 --> 00:01:40.240] And then on Twitter, we had like this little rivalry, I guess.
[00:01:40.240 --> 00:01:40.960] There it is.
[00:01:41.920 --> 00:01:43.280] Personally, for me, I gained...
[00:01:43.360 --> 00:01:49.440] I went from 50k followers 1.5 to 60 because Peter was tweeting about me all the time, right?
[00:01:49.440 --> 00:01:58.160] So it feels like we had this little YouTube collaboration where some people put like another YouTuber on the video and then have like cross-followings and stuff.
[00:01:58.160 --> 00:02:00.120] So it really helped me out.
[00:01:59.680 --> 00:02:03.640] Rivalry on the outside and it's warm and fuzzy teddy bears on the inside.
[00:01:59.760 --> 00:02:04.600] Exactly, exactly.
[00:02:04.920 --> 00:02:09.000] Yeah, there's 37 Signals and their book, Rework.
[00:02:09.000 --> 00:02:10.760] They have a chapter called Pick a Fight.
[00:02:10.760 --> 00:02:14.920] And one of the things that they talk about is intentionally going out of your way to have a rival.
[00:02:14.920 --> 00:02:16.920] And they have like one constraint for that.
[00:02:16.920 --> 00:02:18.600] They say, always punch up.
[00:02:18.600 --> 00:02:20.120] Everybody loves an underdog.
[00:02:20.120 --> 00:02:27.160] You never want to be the most popular person picking like somebody who is less, you know, far along to be your rival.
[00:02:27.160 --> 00:02:32.200] And in a way, like, I think like, like you said, like, I think Peter Levels has like over 100,000 followers on Twitter.
[00:02:32.200 --> 00:02:33.720] And you were at, what, like, 15?
[00:02:34.040 --> 00:02:37.240] And so, like, because the two of you are rivals, like, oh, he's at 270.
[00:02:37.240 --> 00:02:38.200] Geez, he's like blown it.
[00:02:38.280 --> 00:02:39.240] He's way up there.
[00:02:39.240 --> 00:02:41.160] He just kind of like brought you along with him, right?
[00:02:41.160 --> 00:02:43.640] Like, you are, like, in many people's eyes, probably the underdog.
[00:02:43.640 --> 00:02:47.320] And so, like, at whatever point you can, like, Channy, who's our rival?
[00:02:47.320 --> 00:02:48.040] We don't really have.
[00:02:48.520 --> 00:02:51.080] You could rival with Sam and his.
[00:02:51.400 --> 00:02:53.000] What is his company called again?
[00:02:53.000 --> 00:02:53.480] Who is this?
[00:02:53.480 --> 00:02:54.760] So he has like this offense.
[00:02:54.840 --> 00:02:56.840] Oh, he beat to pay a shit ton of money.
[00:02:56.840 --> 00:03:01.000] You guys can do it for like the underdogs that can eventually move into New Hampshire.
[00:03:01.880 --> 00:03:02.280] Yeah.
[00:03:02.280 --> 00:03:03.560] Sam Parr could be our rival.
[00:03:03.560 --> 00:03:07.720] Like, in a way, like, Y Combinator is like an aspirational rival.
[00:03:07.720 --> 00:03:12.600] They used to just do funding, but now they're doing all this inspirational founder content.
[00:03:12.920 --> 00:03:13.880] They're doing content, man.
[00:03:13.880 --> 00:03:14.920] They moved into our lane.
[00:03:14.920 --> 00:03:15.640] They're on YouTube.
[00:03:15.640 --> 00:03:16.440] They got their podcast.
[00:03:16.440 --> 00:03:17.080] They got their blog.
[00:03:17.080 --> 00:03:18.360] They're doing founder interviews.
[00:03:18.360 --> 00:03:20.520] And so it's like, that's punching up for us.
[00:03:20.520 --> 00:03:25.000] You know, if we're their rival, and if they tweet about us or say anything back to us, like, that helps us.
[00:03:25.000 --> 00:03:31.480] So, Danny, the reason you're rivals with Peter Level is because you're both doing these AI tools.
[00:03:31.480 --> 00:03:33.400] I don't know how to describe you as an eddy hacker.
[00:03:33.400 --> 00:03:35.720] You've been working on a lot of stuff for a long time.
[00:03:35.720 --> 00:03:38.200] You've had, I think, a seven-figure exit.
[00:03:38.520 --> 00:03:43.720] You have been working with AI in particular for three years now, I think.
[00:03:43.720 --> 00:03:45.000] I saw you tweeting about stuff, like back in 2010.
[00:03:46.560 --> 00:03:54.160] And now you've got like the suite of AI photo editing tools, for a lack of a better word.
[00:03:54.160 --> 00:03:59.520] Like the way that I would describe your portfolio of apps is that they all pick some vertical, right?
[00:03:59.520 --> 00:04:03.360] Photography is a probably $10, $100 billion industry.
[00:04:03.360 --> 00:04:07.840] And it looks like you've just gone down the list of what makes a lot of money and photography.
[00:04:07.840 --> 00:04:10.560] Headshots, okay, I'm doing an AI headshot thing.
[00:04:10.560 --> 00:04:14.320] Like hiring models, okay, I'm doing AI modeling agency.
[00:04:14.320 --> 00:04:15.200] You don't need to hire models.
[00:04:15.200 --> 00:04:18.480] We'll create a model for you in AI and put that person's face all over your website.
[00:04:18.480 --> 00:04:20.480] Team headshots got that unlocked.
[00:04:20.480 --> 00:04:24.800] Like you're just going down the list of what makes money and doing it with AI.
[00:04:24.800 --> 00:04:26.560] I think Peter is doing it all in one tool.
[00:04:26.560 --> 00:04:41.680] He's trying to make like the SaaS one agency model, and I'm trying to do all these different verticals, trying to do all the SEO research and like what, yeah, what you said is like headshots, team headshots, profile pictures, modeling, trying to figure out what is the next step on it.
[00:04:41.680 --> 00:04:49.280] Because I can just use the same tech and I just have to make a new landing page, do some SEO keywords for that, and then just drop into that photocol and just keep repeating it.
[00:04:49.280 --> 00:04:52.560] I'm actually planning on making one backend system so I don't have to rebuild that anymore.
[00:04:52.560 --> 00:04:56.160] So you can just like snipe into one lane two weeks.
[00:04:56.160 --> 00:04:59.760] And I think the great thing is I'm competing with the photographers.
[00:04:59.760 --> 00:05:01.920] And I mean, they don't know how to develop, right?
[00:05:02.240 --> 00:05:05.040] So you're also competing with other people like you and Peter.
[00:05:05.040 --> 00:05:10.800] Like I did kind of a Google search and it's not just like you and Peter, right?
[00:05:11.120 --> 00:05:15.360] If I look for like AI headshot apps, there's Headshot Pro, which is you.
[00:05:15.360 --> 00:05:16.640] There's Headshot AI.
[00:05:16.640 --> 00:05:18.720] There's HeadshotsBuyAI.com.
[00:05:18.720 --> 00:05:20.960] There's like 15 others.
[00:05:20.960 --> 00:05:27.120] So anyone who's actually searching for like, hey, I want headshots by AI, which is probably like 1% of the population.
[00:05:27.120 --> 00:05:29.760] Probably most people don't even think that AI could do this right now.
[00:05:30.120 --> 00:05:32.200] It's you versus a lot of people.
[00:05:32.200 --> 00:05:34.600] Yeah, there's a lot of competitors.
[00:05:35.480 --> 00:05:37.480] I'm doing quite well for some reason, I think.
[00:05:37.480 --> 00:05:39.240] Probably because I have a high conversion rate.
[00:05:39.400 --> 00:05:42.040] So I can outcompete them on Google Ads and stuff like that.
[00:05:42.360 --> 00:05:42.920] Yeah.
[00:05:42.920 --> 00:05:46.360] But I think I'm lucky here because I built my own tech, my own models.
[00:05:46.360 --> 00:05:54.680] So this is harder to copy because I've been spending like three, four weeks really coming up with my own AI, writing Python, training my own models.
[00:05:54.680 --> 00:05:59.640] So this is not just like I'm going to hook up to an API and then get the results out.
[00:05:59.640 --> 00:06:00.200] Yeah.
[00:06:00.200 --> 00:06:01.400] How well are you doing?
[00:06:01.400 --> 00:06:08.280] I know on Headshot Pro, I looked and you said that you've got like the numbers literally on your website.
[00:06:08.280 --> 00:06:10.680] And so you say like, okay, you've got 10,000.
[00:06:11.000 --> 00:06:12.680] You can reverse engineer it.
[00:06:13.400 --> 00:06:17.880] 10,423 happy customers paying $29 a person.
[00:06:17.880 --> 00:06:18.840] And that's where your rates start.
[00:06:18.840 --> 00:06:21.000] So some people are paying more than that.
[00:06:21.000 --> 00:06:29.480] For example, if you do headshots for your team, it's $40 per person for your whole team, which is a pretty good deal compared to like a traditional photographer.
[00:06:29.480 --> 00:06:32.840] Do you share your total revenue numbers or should I just do some math here?
[00:06:32.840 --> 00:06:42.040] Yeah, I quit doing it after I launched Headshot Pro because I felt, okay, A, it's going to get so many competitors, and B, I feel bad about sharing it right now.
[00:06:42.520 --> 00:06:45.640] These numbers are not that share-friendly anymore.
[00:06:45.640 --> 00:06:45.880] Right.
[00:06:45.960 --> 00:06:47.320] But you can do the math.
[00:06:47.320 --> 00:06:47.720] Yeah, sure.
[00:06:47.720 --> 00:06:48.360] It's public.
[00:06:48.360 --> 00:06:51.240] I feel like we're going to just do some arithmetic real quick here.
[00:06:51.240 --> 00:06:52.200] It's over $300,000.
[00:06:52.360 --> 00:06:53.800] No, no public math, though.
[00:06:53.800 --> 00:06:54.280] Yeah.
[00:06:54.520 --> 00:06:55.640] I got a calculator.
[00:06:55.880 --> 00:06:57.560] You've got other products too.
[00:06:57.560 --> 00:06:59.480] It's not just Headshot Pro.
[00:06:59.480 --> 00:07:03.000] So you're making hundreds of thousands of dollars in a very short period of time.
[00:07:03.000 --> 00:07:05.480] Like, these aren't, like, when did you launch Headshot Pro?
[00:07:05.480 --> 00:07:07.240] I think that's now five weeks ago.
[00:07:07.240 --> 00:07:07.920] Yeah, five weeks.
[00:07:07.800 --> 00:07:08.560] Provo Picture.
[00:07:08.680 --> 00:07:12.680] ProfoPicture was November, so that's three, five months.
[00:07:12.680 --> 00:07:15.840] I mean, AI, official AI didn't exist six months ago.
[00:07:15.840 --> 00:07:16.800] Yeah, it's crazy.
[00:07:14.840 --> 00:07:21.600] And you have this tweet that I think encapsulates kind of the opportunity here because sometimes people get like upset.
[00:07:21.840 --> 00:07:27.360] Like, one of the most popular tweets that I see on Twitter nowadays is, God, I've muted everything to do with AI.
[00:07:27.360 --> 00:07:28.400] Everything is just a clone.
[00:07:28.400 --> 00:07:29.280] Blah, blah, blah, blah, blah.
[00:07:29.440 --> 00:07:31.520] People are so tired of seeing AI stuff.
[00:07:31.520 --> 00:07:36.000] And then I look at you and you're like, Kill it, made $300,000 in the last five weeks.
[00:07:37.600 --> 00:07:40.560] I think I got lucky though, because I sold my company back in 2020, right?
[00:07:40.560 --> 00:07:43.600] And I got a non-compete on ChatGPT, OpenAI.
[00:07:43.600 --> 00:07:45.200] I cannot do that for three years.
[00:07:45.200 --> 00:07:51.520] So I'm like, wow, forced to not do any of the fancy, hyped stuff on Twitter that everyone's doing.
[00:07:51.520 --> 00:07:53.040] So I need to find something else to do.
[00:07:53.280 --> 00:07:53.760] Wait, wait, wait.
[00:07:53.920 --> 00:07:56.880] So what exactly is your non-compete?
[00:07:56.880 --> 00:07:58.080] Like, what are you not allowed to do?
[00:07:58.080 --> 00:08:02.800] I'm not allowed to do any text generation for three years, which ends next year.
[00:08:03.840 --> 00:08:04.560] Wow.
[00:08:05.120 --> 00:08:06.880] And this is because of your acquisition?
[00:08:06.880 --> 00:08:08.320] Yeah, by the company that you sold.
[00:08:08.320 --> 00:08:11.600] So I sold my copyright generator to Jasper back in the days.
[00:08:11.600 --> 00:08:16.800] And obviously, they don't want me to compete anymore because the founders said you ship too fast.
[00:08:16.800 --> 00:08:18.800] I don't want anything to do with you anymore.
[00:08:19.760 --> 00:08:20.400] Wow.
[00:08:20.400 --> 00:08:20.720] Okay.
[00:08:20.720 --> 00:08:24.640] So that severely handicaps you, but like you said, constraints, creativity.
[00:08:24.640 --> 00:08:24.960] Now you're going to be able to do it.
[00:08:25.120 --> 00:08:26.000] Dude, I feel like that doesn't.
[00:08:26.000 --> 00:08:32.480] Yeah, I feel like that doesn't even handicap him because what that does, if you have constraints, then you end up doing things that are new.
[00:08:32.480 --> 00:08:40.240] And in this particular case, we're in a situation where all you see today is basically chat GPT stuff.
[00:08:40.240 --> 00:08:41.600] You see a bunch of text generation.
[00:08:41.600 --> 00:08:45.760] All you saw five months ago was like image generation, right?
[00:08:46.080 --> 00:08:58.720] So because you're handcuffed from doing the thing that is like the hot thing that everyone is like constantly bored by, you have this like built-in forcing function to like, okay, what else can I do, right?
[00:08:58.720 --> 00:08:59.280] Yeah.
[00:08:59.000 --> 00:09:01.720] Yeah, yeah, and the other thing is like you move really fast, right?
[00:08:59.440 --> 00:09:04.920] That's why this like company that acquired your last product doesn't want to compete with you.
[00:08:59.600 --> 00:09:06.600] And guess what's also moving fast?
[00:08:59.680 --> 00:09:14.200] Like the state of AI, things that were cutting edge six months ago, nine months ago now look like crap and they're yesterday's news.
[00:09:14.200 --> 00:09:16.200] And if you're still doing that stuff, like you're behind.
[00:09:16.200 --> 00:09:17.880] So you have to constantly innovate.
[00:09:17.880 --> 00:09:27.080] And I think this is one of the reasons why people, especially sort of like the more stuffy, tried and true business theoretical crowd are like, oh, you know, you need to build a business that can last.
[00:09:27.080 --> 00:09:28.760] But like, do you?
[00:09:28.760 --> 00:09:34.920] Is there anything wrong with building a business that makes a bunch of money in a short period of time and then becomes obsolete?
[00:09:34.920 --> 00:09:36.600] But by that time, you're on to the next thing?
[00:09:36.600 --> 00:09:38.440] Like, I see no problem doing that.
[00:09:38.440 --> 00:09:45.000] And that might be the way the world is going considering the pace of technological advancement right now.
[00:09:45.000 --> 00:09:46.600] Like, look at Linktree, for example.
[00:09:46.600 --> 00:09:49.560] Like, they are trying to build a sustainable product, right?
[00:09:49.560 --> 00:09:55.560] And then last week, I think Instagram suddenly decided, hey, guys, you can add five links and suddenly your sustainable feature business is gone.
[00:09:55.560 --> 00:09:55.880] Yeah.
[00:09:55.880 --> 00:09:57.480] It's the same with TikTok.
[00:09:57.480 --> 00:09:59.960] TikTok just added the avatars that they were ProvoPicture.
[00:09:59.960 --> 00:10:03.640] So if I was building ProvoPicture as a sustainable business, it's still sustainable, right?
[00:10:03.640 --> 00:10:09.400] But if you build on a hype and if you build on a feature, like you want to get the cash out as soon as possible, right?
[00:10:09.400 --> 00:10:16.120] Because someone that ships slower than you, six months down a drain that has like a mass audience, they're just gonna, they're just gonna take your audience.
[00:10:16.360 --> 00:10:17.880] Incumbent sales are already down.
[00:10:17.880 --> 00:10:19.160] Like you can already notice it.
[00:10:19.160 --> 00:10:27.720] So all these huge companies employ tens of thousands of really smart programmers and they are getting better and better at building things quickly.
[00:10:27.720 --> 00:10:33.080] And like you said, if you don't release, you know, capture as much money as you can now, like, okay, maybe they move slower than you.
[00:10:33.080 --> 00:10:36.760] Maybe it takes them six months to get to where you are, but like, then they have the distribution advantage.
[00:10:36.760 --> 00:10:41.480] And it turns out that having a distribution advantage is worth more than six months.
[00:10:41.480 --> 00:10:43.160] You know, like they could way more.
[00:10:43.160 --> 00:10:44.360] It's worth so much.
[00:10:44.360 --> 00:10:46.000] And so you've got to build like super fast.
[00:10:46.000 --> 00:10:47.200] So I like what you're doing.
[00:10:47.200 --> 00:10:48.000] I mean, there's haters.
[00:10:44.760 --> 00:10:50.560] There's people who say, like, okay, these businesses aren't going to last.
[00:10:44.840 --> 00:10:51.840] They're just clones of each other.
[00:10:52.160 --> 00:10:54.080] I guess it depends on your personality type, right?
[00:10:54.080 --> 00:10:59.120] Like, I have this little ADD personality type where after three months, I'm absolutely bored with a product.
[00:10:59.200 --> 00:11:00.960] I don't want to work on it anymore.
[00:11:00.960 --> 00:11:02.720] Like, I don't even want to touch it.
[00:11:02.720 --> 00:11:04.240] So for me, it's perfect.
[00:11:04.480 --> 00:11:17.200] But for other people, if you like more slowly building, getting out, getting like a stable life, you have children, whatever, then probably you absolutely hate this because you can build any stability on top of it.
[00:11:17.200 --> 00:11:17.920] Right.
[00:11:17.920 --> 00:11:22.080] If you didn't bring up the ADHD thing, I was looking at your list of products here.
[00:11:22.080 --> 00:11:23.920] I'm like, I'm just scrolling through.
[00:11:23.920 --> 00:11:25.920] You've got like parody bar discontinued.
[00:11:25.920 --> 00:11:28.480] Headline acquired like rare blocks discontinued.
[00:11:28.480 --> 00:11:30.480] This one, 2021, this one, 2023.
[00:11:30.480 --> 00:11:37.360] And I'm like, this is a timeline of your products, but it's also like a timeline of your mental focus and like the shiny objects.
[00:11:37.360 --> 00:11:40.080] Yeah, postcrafts.com, 19 different products.
[00:11:40.080 --> 00:11:41.760] 19 startups by Danny Postcraft.
[00:11:41.840 --> 00:11:42.560] Oh, it's 19.
[00:11:44.400 --> 00:11:45.440] You're not just shipping products.
[00:11:45.440 --> 00:11:47.120] You're shipping products and you're tweeting about it.
[00:11:47.120 --> 00:11:48.240] You're talking about your revenue.
[00:11:48.240 --> 00:11:49.440] You're talking about your decisions.
[00:11:49.440 --> 00:11:51.920] You're talking about the disputes you're dealing with with customers.
[00:11:51.920 --> 00:12:01.680] Like you're doing all the juicy behind-the-scenes stuff that makes people want to follow along because it's just far more interesting to follow somebody like you than somebody who doesn't ever tweet about it.
[00:12:01.840 --> 00:12:13.840] For me, for me, it's also a diary because I read back tweets I did a year ago if I needed to solve something because I remember I had the same issue at that time and I would be reading it back and then like I'm kind of also doing it for myself just so it has like cross-reference thing.
[00:12:13.840 --> 00:12:17.200] I like this sort of like 12 startups in 12 months approach.
[00:12:17.200 --> 00:12:21.920] Also, something that like Peter came up with a long time ago, where you basically say, Screw trying to work on one thing.
[00:12:21.920 --> 00:12:24.640] I'm going to do a ton of things in the course of a year.
[00:12:24.640 --> 00:12:27.440] I'm going to learn a lot and I'm going to see what sticks.
[00:12:27.440 --> 00:12:31.880] You said you built like you know a dozen products or tons of products in 2020.
[00:12:31.880 --> 00:12:34.200] What were they and what stuck?
[00:12:34.520 --> 00:12:38.600] So the first one was Landic Folio, which I did for four years, which is the inspiration gallery.
[00:12:38.600 --> 00:12:48.600] After that, I took a little feature from that website and turned it into a drag and drop inspiration builder so you could make a landing page based on examples of other websites.
[00:12:48.600 --> 00:12:49.400] So that was another one.
[00:12:49.400 --> 00:12:54.360] So I like to keep, like, take a little feature of another product and iterate on top of it.
[00:12:54.360 --> 00:12:56.760] How much money did these products make?
[00:12:57.080 --> 00:12:59.080] 2K tops, I think, per month.
[00:12:59.080 --> 00:13:02.760] In SpireFrame, maybe 400 at the time?
[00:13:02.760 --> 00:13:03.560] Nothing like that.
[00:13:03.720 --> 00:13:05.960] Enough to survive if you're living in Bali.
[00:13:05.960 --> 00:13:07.960] But not what you're looking for.
[00:13:07.960 --> 00:13:09.720] Yeah, for many, many years.
[00:13:09.720 --> 00:13:10.360] Right.
[00:13:10.360 --> 00:13:11.400] For a long time.
[00:13:11.400 --> 00:13:11.720] Okay.
[00:13:11.720 --> 00:13:11.880] Yeah.
[00:13:11.880 --> 00:13:14.280] And then I got lucky with the headline.
[00:13:14.680 --> 00:13:14.840] Yeah.
[00:13:15.160 --> 00:13:18.280] That's like your fifth product on your list.
[00:13:18.280 --> 00:13:19.560] This is the one that got acquired.
[00:13:19.560 --> 00:13:24.440] You started in 2020, and it was an AI-based copywriting thing.
[00:13:24.440 --> 00:13:28.120] The headline on the website now says, Writing copy has never been easier.
[00:13:28.120 --> 00:13:33.720] The automated copywriter for busy designers, anti-hackers, marketing agencies, business owners.
[00:13:33.720 --> 00:13:36.360] How did you come up with an idea that you were able to sell?
[00:13:36.680 --> 00:13:40.120] Also, let me point out something that I noticed before you even say that.
[00:13:40.120 --> 00:13:42.520] Like, there were four products that happened before that.
[00:13:42.520 --> 00:13:47.640] You know, you had the landing page design, you have this pop-up library for JavaScript, you have mock-ups.
[00:13:47.640 --> 00:13:52.280] So you have all of these basically like these products that have to do with just basically sort of helping people code.
[00:13:52.280 --> 00:13:54.600] And then, boom, you just have AI.
[00:13:54.600 --> 00:13:56.200] And then that's the one that got acquired.
[00:13:56.200 --> 00:14:01.800] So, and there is a really nice bridge going from that programming to the actual AI content.
[00:14:01.800 --> 00:14:04.920] So on Linux Folio, I have marketers coming there.
[00:14:04.920 --> 00:14:09.880] So we wrote an e-book with headline templates because that's the hardest way to come up for the Linux page.
[00:14:09.880 --> 00:14:19.120] So we sold an e-book with copies, like get blank in blank hours, and then people could fill in the variables in the e-book.
[00:14:14.680 --> 00:14:21.600] And then I was thinking, okay, I just learned the program.
[00:14:21.760 --> 00:14:24.400] I was like, you can sell an e-book for $19, $19, right?
[00:14:24.480 --> 00:14:25.200] Like, not much.
[00:14:25.200 --> 00:14:28.560] But if you sell a SaaS tool, you can suddenly increase the price over that.
[00:14:28.560 --> 00:14:34.960] So I was like, maybe if I could tell, like, if I put those templates into a SaaS tool and you can sell it for $89.
[00:14:34.960 --> 00:14:37.520] So that one exploded when I sold it.
[00:14:37.520 --> 00:14:39.920] I got like $60K revenue, I think, in two weeks.
[00:14:39.920 --> 00:14:40.800] That was my first experience.
[00:14:40.960 --> 00:14:41.200] Wow.
[00:14:41.600 --> 00:14:43.440] And that was Pushfolio, right?
[00:14:43.440 --> 00:14:47.440] No, that was a headline actually before they went.
[00:14:48.400 --> 00:14:49.280] Wow.
[00:14:49.280 --> 00:14:50.720] Headline AI.
[00:14:51.120 --> 00:14:52.400] Free AI.
[00:14:52.400 --> 00:14:57.040] So I was working on that product, and then a friend of mine showed me GPT-3.
[00:14:57.040 --> 00:15:01.840] And I was like, all right, I'm going to TM, I'm going to email Greg, the CTO of OpenAI.
[00:15:01.840 --> 00:15:02.480] Can I get access?
[00:15:02.480 --> 00:15:03.200] So I got access.
[00:15:03.200 --> 00:15:05.200] It was like one of the first batch.
[00:15:05.200 --> 00:15:10.480] And I let it write the templates for my SaaS for headline with AI.
[00:15:10.560 --> 00:15:11.200] I was like, wait.
[00:15:11.200 --> 00:15:12.240] Oh, that's awesome.
[00:15:12.240 --> 00:15:14.960] Wait, this fucking AI can write copy for me.
[00:15:14.960 --> 00:15:19.760] What if I turn this tool into a copywriting tool instead of just using templates?
[00:15:19.760 --> 00:15:20.320] Right.
[00:15:20.320 --> 00:15:27.680] So I was lucky to be working on a copywriting tool by accident when a copywriting AI came out.
[00:15:27.680 --> 00:15:28.160] Wow.
[00:15:28.160 --> 00:15:36.320] I think I have a screenshot that I'm one of the first people to go live with GPT-3 actually in production together with copy.ai and three tools at the same time.
[00:15:36.400 --> 00:15:39.840] Copy.ai is valued at whatever hundreds of millions of dollars now.
[00:15:39.840 --> 00:15:49.200] I interviewed somebody back in the day who was doing AI copywriting, and I've never gotten so many negative responses to an interview that I've done.
[00:15:49.200 --> 00:15:50.640] People hated that guy.
[00:15:50.640 --> 00:15:52.160] They were so upset.
[00:15:52.160 --> 00:15:53.920] People don't like low-quality content.
[00:15:53.920 --> 00:15:55.600] They don't like spam content.
[00:15:55.600 --> 00:16:11.320] And I think the idea, and this was back in 2020, the idea that we might be heading toward a world where the internet is just gunked up with nothing but low-quality AI spam content and people are creating this at the click of a button just like enraged a lot of people who are listening.
[00:16:11.320 --> 00:16:24.680] The full circle on that is that the first AI feature that we've built for indie hackers was your like AI spam, you know, block or spam moderator that is like going in and zapping members of its own space.
[00:16:24.920 --> 00:16:27.880] People who are getting like AI spam on the indie hackers forum.
[00:16:29.560 --> 00:16:33.080] So this guy, like, so like the fears weren't exactly completely unfounded.
[00:16:33.080 --> 00:16:34.360] I just want to say that.
[00:16:34.360 --> 00:16:34.920] Yeah.
[00:16:34.920 --> 00:16:37.560] Well, back in 2020, this was like, it was a very different landscape.
[00:16:37.560 --> 00:16:38.760] Nobody knew this was possible.
[00:16:38.760 --> 00:16:44.840] And Danny, you said, you just happened to be doing a copywriting tool at the same time that AI came out that could write copy.
[00:16:44.840 --> 00:16:46.360] How did you get it into the hands of people?
[00:16:46.360 --> 00:16:50.840] Because like, I mean, I imagine at this time you had like five followers on Twitter.
[00:16:50.840 --> 00:16:52.280] Nobody knew who you were.
[00:16:52.280 --> 00:16:54.040] You know, you didn't have a giant email list.
[00:16:54.040 --> 00:16:56.680] Like, how did anybody even find out about Headline?
[00:16:56.680 --> 00:17:01.480] I think I got lucky that same with Hetshop Pro and ProvoPicture.
[00:17:01.480 --> 00:17:03.800] Like, it is so novelty.
[00:17:03.800 --> 00:17:06.600] Like, it's so new that it's magic to people.
[00:17:06.600 --> 00:17:10.440] So I was just tweeting out tweets, like, using the tool, and it was new to people.
[00:17:10.440 --> 00:17:12.840] I had 200 followers back in the day, I think.
[00:17:12.840 --> 00:17:14.440] And people just started retweeting it.
[00:17:14.440 --> 00:17:15.480] People went mentor over there.
[00:17:15.480 --> 00:17:16.920] We're like, what the fuck's going on?
[00:17:17.320 --> 00:17:18.840] This tool is writing copy.
[00:17:18.840 --> 00:17:19.800] Like, what?
[00:17:20.120 --> 00:17:21.720] Because we live in a bubble, right?
[00:17:21.720 --> 00:17:23.560] There's like 0.1% bubble.
[00:17:23.560 --> 00:17:26.120] But once you go outside a little bit of it, it's magic to people.
[00:17:26.120 --> 00:17:27.400] It was magic to me.
[00:17:27.400 --> 00:17:28.840] I was like so amazed by it.
[00:17:29.640 --> 00:17:38.280] I grew from, I think, 200 followers in that say to 10K before I sold Headline, I guess.
[00:17:38.280 --> 00:17:41.160] And that was the only marketing I did, only on Twitter.
[00:17:41.160 --> 00:17:42.440] Just word of mouth.
[00:17:42.680 --> 00:17:47.600] That's how I grew to 25K MRR before I sold it.
[00:17:44.840 --> 00:17:49.600] Yeah, it was only word of mouth.
[00:17:50.000 --> 00:17:53.040] No SEO, no marketing, just Twitter.
[00:17:53.040 --> 00:17:53.760] Just Twitter.
[00:17:53.760 --> 00:18:01.760] But I feel like Twitter is like, if you are an indie hacker, if you're sort of a small startup founder, that's where we hang out.
[00:18:02.160 --> 00:18:03.920] That's like your people, our people.
[00:18:03.920 --> 00:18:05.840] But I want to say one thing about you.
[00:18:05.920 --> 00:18:08.240] So you say indie hackers are the right way to hang out, Jenny.
[00:18:10.400 --> 00:18:17.280] But what I was going to say is like, dude, it's so easy to sort of see the narrative as you were doing the right kind of product at the right time.
[00:18:17.280 --> 00:18:19.360] And the wind was in the air.
[00:18:19.360 --> 00:18:24.000] And it just so happened to be that the AI stuff became an option.
[00:18:24.000 --> 00:18:34.160] But dude, you did something in that story that 999 out of 1,000 people would not do, which is you heard about it and you were like, oh, this thing is sick.
[00:18:34.160 --> 00:18:39.200] I'm going to just reach out to the founder and ask for favors.
[00:18:41.760 --> 00:18:43.600] That's not a minor play that you made, right?
[00:18:43.600 --> 00:18:46.080] Number one, you noticed that this thing was awesome.
[00:18:46.400 --> 00:18:50.320] And then the second thing is like, you had the balls or whatever, right?
[00:18:50.640 --> 00:18:53.200] I don't like, you know, most people don't even think of that as an option.
[00:18:53.200 --> 00:18:56.320] Like, I'm just going to reach out to the guy who made this at the very top.
[00:18:56.320 --> 00:19:00.000] And then, but for that decision, none of this would have happened.
[00:19:00.000 --> 00:19:00.400] Yep.
[00:19:00.400 --> 00:19:00.800] Yeah.
[00:19:00.800 --> 00:19:07.040] It's crazy how many little small life decisions escalate into like something big in your life, to be honest.
[00:19:07.520 --> 00:19:14.640] There's also like this piece of advice that's super common, which is like, don't build a solution in search of a problem.
[00:19:14.640 --> 00:19:19.520] This is like foundational startup wisdom that has been around for like, it's been said for like 30 years.
[00:19:19.520 --> 00:19:33.480] Like, it's the most stereotypical thing in the world to be just a tech nerd or someone who loves, like, somebody with ADHD who just loves spritting about, and a new opportunity comes along, and you're like, oh, how do I take this, you know, crypto, AI, whatever thing, and just like slap it onto something that exists.
[00:19:33.480 --> 00:19:35.080] And most people say, don't do that.
[00:19:35.080 --> 00:19:38.760] But there are times when that is like absolutely the right thing to do.
[00:19:38.760 --> 00:19:49.560] When the new technology, I think, is so revolutionary that like when it comes out, if you just apply it to existing problems, it makes them 10 times better than the previous solutions.
[00:19:49.560 --> 00:19:51.080] And like, that's exactly what you did.
[00:19:51.080 --> 00:19:56.120] So it makes me kind of skeptical of some of this advice that's like tried and true advice when there's such clear exceptions.
[00:19:56.520 --> 00:19:58.600] Same thing with the birth of the internet, right?
[00:19:58.600 --> 00:20:03.240] Anyone who didn't slap the internet onto their newspaper in the 90s, their newspaper is dead.
[00:20:03.240 --> 00:20:04.280] They did not make it.
[00:20:04.360 --> 00:20:15.560] So sometimes you just like, I kind of feel like we're in that moment right now with AI where it's like, hey, this is like so foundationally different and better than hiring a bunch of people to do things manually.
[00:20:15.560 --> 00:20:18.040] If like with this keystroke, you could do it automatically.
[00:20:18.040 --> 00:20:20.760] Like it's easy to be a curmudgeon and be like, no, it's just a trend.
[00:20:20.760 --> 00:20:21.320] It's just a trend.
[00:20:21.320 --> 00:20:21.960] It's just a fad.
[00:20:22.600 --> 00:20:23.720] It's not just fad.
[00:20:23.720 --> 00:20:26.440] Like, it's seriously good.
[00:20:26.440 --> 00:20:42.440] The crazy thing I realize, and I haven't been able to put it in words, but AI can, like, you can go into industries that normally take so much human labor, and you can basically just, as a one-person business, and this is why I think this is a golden time for indie hackers.
[00:20:42.440 --> 00:20:44.280] Like the Photo Studio, for example.
[00:20:44.280 --> 00:20:47.800] I'm competing with people who probably have, and it's also a negative thing, right?
[00:20:47.800 --> 00:20:52.840] But it's positive for the consumer because no one wants to take headshots or make $500 worth of.
[00:20:53.160 --> 00:20:55.880] So I see it as like, I do it for the customers.
[00:20:55.880 --> 00:21:08.200] I can basically by myself scale unlimited times if I have this GPU power and just do photography by myself because I have an AI that's just doing the work for me.
[00:21:08.200 --> 00:21:11.640] I have all these job works, all these other AIs filtering, blah, blah, blah, blah, blah.
[00:21:11.800 --> 00:21:16.720] The only thing I need to manually do is, yeah, do the programming and the customer support.
[00:21:14.920 --> 00:21:21.280] And thank God for my wife who helps me out with that because that's a lot of work.
[00:21:21.600 --> 00:21:30.560] And if you apply AI to solving real problems that are done by real humans, I think there's a lot of markets you can get into in that way.
[00:21:30.560 --> 00:21:31.040] Yeah.
[00:21:31.040 --> 00:21:32.720] Shanning hinted at this a second ago.
[00:21:32.720 --> 00:21:36.800] So we've had like a spam problem on indie actors forever.
[00:21:36.800 --> 00:21:41.360] And a lot of the people who are spammers, like it's not just robots, it's actual people sometimes.
[00:21:41.360 --> 00:21:44.560] You come on and I'll implement spam fighting stuff.
[00:21:44.560 --> 00:21:51.520] You know, add a captcha or I'll change like all the classes in my CSS so like the robots can't find it and people just figure it out and they just do it differently.
[00:21:51.520 --> 00:21:52.880] I'll ban certain words.
[00:21:52.880 --> 00:21:54.080] They'll just start using different words.
[00:21:54.080 --> 00:21:56.800] If I ban links, they start putting a space in front of the dot-com.
[00:21:56.800 --> 00:21:57.680] They try everything, right?
[00:21:57.680 --> 00:21:59.840] It's like real people fighting.
[00:21:59.840 --> 00:22:03.840] And two or three weeks ago, I was like, okay, should we hire another community moderator?
[00:22:03.840 --> 00:22:04.880] It's expensive.
[00:22:04.880 --> 00:22:06.960] It's hard to manage the community moderator.
[00:22:06.960 --> 00:22:09.840] You don't really necessarily want to outsource all the community work to somebody else.
[00:22:09.840 --> 00:22:12.080] So we should really be doing a lot of this ourselves.
[00:22:12.560 --> 00:22:14.720] But, you know, mostly it's just expensive.
[00:22:14.720 --> 00:22:15.600] We're independent now.
[00:22:15.600 --> 00:22:16.720] We can't afford that kind of thing.
[00:22:17.120 --> 00:22:18.240] Congratulations, by the way.
[00:22:18.240 --> 00:22:18.640] Awesome.
[00:22:18.720 --> 00:22:18.960] Thanks.
[00:22:18.960 --> 00:22:20.480] It's nice to be nice to be indie.
[00:22:20.480 --> 00:22:22.000] So I'm like, okay, what can I do?
[00:22:22.000 --> 00:22:30.000] Well, I could just take a solution that exists, AI, and just like slap it onto this problem and say, okay, well, what can like, what can GPT-4 do?
[00:22:30.000 --> 00:22:34.480] And so in the span of like a couple days, I just played around GPT-4 and it wasn't really even a couple days.
[00:22:34.480 --> 00:22:35.760] It was like a few hours.
[00:22:35.760 --> 00:22:40.080] And now I have a spam fighting bot who is just like dispensing justice on the forum.
[00:22:40.080 --> 00:22:41.360] I call him botman.
[00:22:41.360 --> 00:22:43.360] And he will message me and Telegram.
[00:22:43.360 --> 00:22:47.040] Every time he sees a spammer, he will give them a spam score from zero to one.
[00:22:47.040 --> 00:22:48.400] He'll give me a confidence score.
[00:22:48.400 --> 00:22:50.560] How confident is he from zero to one?
[00:22:50.560 --> 00:22:53.600] He will give me a reason and actual pros.
[00:22:53.600 --> 00:22:58.560] They'll tell me, oh, this person, their account is really new and it seems like this, blah, blah, blah.
[00:22:58.560 --> 00:23:00.920] And I think this is definitely a spammer.
[00:23:00.920 --> 00:23:03.560] And I just set him today to auto-ban people.
[00:22:59.840 --> 00:23:06.440] And dude, I think it's all, I think it's like all hits and no misses.
[00:23:06.680 --> 00:23:08.280] No, I've just been checking over two weeks.
[00:23:08.680 --> 00:23:09.960] He has never, it is never wrong.
[00:23:09.960 --> 00:23:11.160] It is right every single time.
[00:23:11.160 --> 00:23:18.840] Every time it finds its teacher, you just train like a really long, long text in ChatGPT where you put like examples inside and then it learns from it.
[00:23:18.840 --> 00:23:21.880] Or are you using like a vector database and it finds like the distance?
[00:23:22.040 --> 00:23:25.320] I'm doing the simplest, dumbest few shot learning in GPT-4.
[00:23:25.640 --> 00:23:27.560] I've given it like 10 examples.
[00:23:27.560 --> 00:23:30.360] And GPT-4's context is like, how many tokens is it?
[00:23:30.360 --> 00:23:35.560] It's like 12,000 or some crazy amount of tokens, which channeling a token is basically like three or four letters.
[00:23:35.560 --> 00:23:39.000] And so it can handle a gigantic wall of text of examples.
[00:23:39.000 --> 00:23:41.400] Plus, it's already been trained on basically the entire internet.
[00:23:41.400 --> 00:23:42.760] So it knows what spam is.
[00:23:42.760 --> 00:23:45.160] It knows like what moderating a forum is.
[00:23:45.160 --> 00:23:49.560] And it's like, okay, well, that just like wiped out the job of like policing spam on the forum.
[00:23:49.800 --> 00:23:54.680] I not only don't have to hire anyone, but I also don't need to go like pay some other business that's built a cool.
[00:23:54.680 --> 00:23:56.920] Like people have pitched me tools for this for years.
[00:23:56.920 --> 00:23:59.160] Hey, we've got some spam planning algorithms we train.
[00:23:59.160 --> 00:24:02.840] Sorry, GPT-4 is better than everything you've been working on for years.
[00:24:02.840 --> 00:24:04.760] And I just rigged it up in like an afternoon.
[00:24:04.760 --> 00:24:05.720] Like, I'm going to do this.
[00:24:05.720 --> 00:24:06.680] It's basically free.
[00:24:06.680 --> 00:24:07.320] Goodbye.
[00:24:07.320 --> 00:24:07.640] Right.
[00:24:07.640 --> 00:24:08.520] And it's like entertaining.
[00:24:08.520 --> 00:24:09.400] I can tell it to be funny.
[00:24:09.400 --> 00:24:10.520] I could tell it to be humorous.
[00:24:10.520 --> 00:24:13.560] It'll send me interesting, funny messages and then ban the spammers.
[00:24:13.560 --> 00:24:20.520] And so it's like, I don't see how people can look at this technology and say, like, this is a passing fad, just like crypto or something.
[00:24:21.480 --> 00:24:23.080] It's not like that's like head in the sand.
[00:24:23.240 --> 00:24:25.320] There's so much more use case to it.
[00:24:25.320 --> 00:24:26.760] Like, this is not a fad.
[00:24:26.760 --> 00:24:31.960] This is completely industry changing, I think, even for the better and for the worse, right?
[00:24:31.960 --> 00:24:32.920] Like, yeah.
[00:24:33.240 --> 00:24:43.640] To me, I think that the sign of it is like, I said this a few weeks ago, where it's like the difference between, you know, you feeling like there might be an opportunity, like there's new technology, and you're like, oh, there's an opportunity.
[00:24:43.640 --> 00:24:47.200] Maybe we should do some like RD, and like maybe there's something there.
[00:24:44.520 --> 00:24:49.120] But this is not, it doesn't feel like an opportunity.
[00:24:49.360 --> 00:25:04.640] We got that spam fighting bot going, and now I'm almost like, dude, it's a liability that like there's almost certainly other parts of our stack, whether it's our technology stack or like our process stack, like how we do things, where we're currently wasting tons of time.
[00:25:04.640 --> 00:25:11.280] And like, I'm almost like, I want to like, you know, sort of do this like an audit, like an internal audit.
[00:25:11.280 --> 00:25:12.240] Like, what are we doing?
[00:25:12.240 --> 00:25:14.960] What are we doing manually that we should be using AI for?
[00:25:14.960 --> 00:25:18.800] Not just like as a distraction to mess around, but it's like a real time-saving technique.
[00:25:18.800 --> 00:25:20.800] Like, we can't afford not to do that.
[00:25:20.800 --> 00:25:23.920] And the amazing thing is, like, you don't have the code anymore.
[00:25:23.920 --> 00:25:30.800] Like, what OpenAI did basically is they turned the written English word or whatever word into a programming language.
[00:25:30.800 --> 00:25:31.840] So anyone can use it.
[00:25:32.000 --> 00:25:35.920] Like, some people listening to this podcast are probably like, yeah, you know, these guys are all programmers.
[00:25:35.920 --> 00:25:36.560] Only they can do it.
[00:25:36.560 --> 00:25:41.920] Like, no, man, you can just send text to it, tell it what to do, and it sends you text back.
[00:25:41.920 --> 00:25:47.920] And if you tell it to send it back as like a one or a zero or a true or a false, like this is spam, yes or no?
[00:25:47.920 --> 00:25:48.880] And you can just read it out.
[00:25:48.880 --> 00:25:50.160] Like, you don't need to know program.
[00:25:50.320 --> 00:25:52.800] So I think people don't like fully comprehend what you're saying.
[00:25:52.800 --> 00:25:56.080] Because this is something that I was telling my friend about when I was coding my spam bot.
[00:25:56.160 --> 00:25:57.120] Like, oh, you're coding this thing.
[00:25:57.120 --> 00:25:59.680] I'm like, I'm literally writing text.
[00:25:59.680 --> 00:26:00.400] That's what I'm saying.
[00:26:00.400 --> 00:26:02.560] Like, there's some code involved, right?
[00:26:02.560 --> 00:26:07.600] Like, if I tried to fight spam a year ago or a month ago, I would be doing nothing but coding.
[00:26:07.600 --> 00:26:11.360] Now I wrote a tiny amount of code that basically just says, hey, ban this person.
[00:26:11.360 --> 00:26:18.080] But the vast majority of what I'm doing is tweaking a single sentence that says, You are a forum moderator for the anti-hackers community.
[00:26:18.080 --> 00:26:22.800] You're reviewing comments made by users on the forum and assigning a spam score ranging from zero to one.
[00:26:22.800 --> 00:26:24.400] Zero means it's unlikely to be spam.
[00:26:24.400 --> 00:26:25.920] One means it's very likely to be spam.
[00:26:25.920 --> 00:26:26.480] Blah, blah, blah.
[00:26:26.640 --> 00:26:30.920] I just keep writing all this stuff, and I end up with like a couple paragraphs of instructions and then it does it.
[00:26:29.120 --> 00:26:35.880] So like coding is literally me just giving instructions as if I was talking to a human.
[00:26:36.200 --> 00:26:39.080] It's like the highest level of abstraction of no code.
[00:26:39.080 --> 00:26:45.240] It's like no code with just like, okay, the interface for coding here, the syntax is, do you know English?
[00:26:45.240 --> 00:26:47.560] Okay, so just like use English and then talk to it.
[00:26:47.560 --> 00:26:48.520] Like that's the syntax for this.
[00:26:48.760 --> 00:26:50.680] So does your computer, it turns out.
[00:26:50.920 --> 00:26:53.880] If you think about it, like programming is like, it's such a bug.
[00:26:53.880 --> 00:26:59.000] Like why would we have to talk in a way that a computer understands if we made a computer?
[00:26:59.000 --> 00:27:02.520] Like we should just be able to write in English to a computer.
[00:27:02.520 --> 00:27:03.080] Yeah.
[00:27:03.080 --> 00:27:04.520] And that's getting possible now.
[00:27:04.520 --> 00:27:06.120] So everyone is becoming a programmer.
[00:27:06.120 --> 00:27:07.320] I guess people hate that.
[00:27:08.440 --> 00:27:10.360] And by the way, I've seen this in practice.
[00:27:10.360 --> 00:27:16.440] So Cortland, I told Dave, a buddy of ours that works at Lyft, I share all this stuff with him.
[00:27:16.440 --> 00:27:18.760] I share all the new tech stuff.
[00:27:19.000 --> 00:27:23.560] Maybe he's so interested because he works at Lyft and jobs are getting sliced left and right at these kind of companies.
[00:27:23.560 --> 00:27:25.800] So he's looking for his escape hatch.
[00:27:25.800 --> 00:27:31.320] But the first thing that he said when I showed him our bot was he was like, he said two things.
[00:27:31.320 --> 00:27:33.640] He goes, number one, that's awesome.
[00:27:33.640 --> 00:27:37.160] And then number two, bummer, like it's pointless for me to know about this.
[00:27:37.160 --> 00:27:37.960] I'm not a developer.
[00:27:37.960 --> 00:27:41.720] So I guess I'll just wait three years until somehow I can do something like this.
[00:27:41.960 --> 00:27:45.880] And I'm like, dude, no, like, you know, there are some limitations.
[00:27:45.880 --> 00:27:56.040] You have to be a little bit technical for some aspects of this, but like, you really should look into tools and ways that you can implement these kinds of things with your own workflows.
[00:27:56.040 --> 00:28:01.240] And I don't know if you listen to the all-in podcast, but it's got a few investors on it.
[00:28:01.240 --> 00:28:02.920] And one of them is Jason Calicanis.
[00:28:02.920 --> 00:28:19.200] And like the next day, there was an episode where Jason Calicanis, who's not really a developer, he was a podcast host, was like, oh, what I do these days is I have my, like, whenever I open a new tab in my browser, it is set to go to Auto GPT.
[00:28:14.760 --> 00:28:25.200] Like, I'm sort of creating this forcing function where I automatically just have to sort of go through Auto GPT.
[00:28:25.200 --> 00:28:32.720] So, no matter what I'm doing, I'm constantly like reminding myself, like, oh, this thing, could that be done better with Auto GPT?
[00:28:32.720 --> 00:28:34.560] Could this be done better with Auto GPT?
[00:28:34.560 --> 00:28:48.240] And it just sort of, he's like, I'm just training myself because no matter what, the thing I just said, like, that idea that like 100%, there are liabilities in the way that you're doing whatever it is that you're doing, like, some AI tool or feature could do it better.
[00:28:48.240 --> 00:28:50.560] There's all sorts of apps that are basically like glue.
[00:28:50.560 --> 00:28:54.880] So, the no-code community has been really into apps like Zapier for years and years and years and years.
[00:28:54.880 --> 00:29:01.840] And it's like, you can't code, but you need, you know, your Gmail to talk to your Telegram, to talk to Slack, to talk to, you know, Google Docs or something.
[00:29:01.840 --> 00:29:06.080] Okay, whenever I get a new email that is this subject, you know, send me this message and Telegram.
[00:29:06.080 --> 00:29:09.280] You could hook that up without code in like five seconds with Zapier, right?
[00:29:09.280 --> 00:29:12.400] And now all these apps are adding chat GPT and AI.
[00:29:12.400 --> 00:29:15.360] So, you know, if I type this query in, have it, you know, go here.
[00:29:15.360 --> 00:29:18.320] So you really, really don't need to learn how to code.
[00:29:18.320 --> 00:29:23.200] Like, it's, if you know the right apps, you could just sign up and start like hooking stuff up.
[00:29:23.200 --> 00:29:30.640] Everyone should spend like, probably take a few days off and just try to automate some of their most repetitive things in their life.
[00:29:30.640 --> 00:29:32.640] And they will just save so much more time.
[00:29:33.360 --> 00:29:35.680] I want to talk about, like, I wanted so much I want to talk to you about.
[00:29:35.680 --> 00:29:37.440] Like, what are you automating in your life?
[00:29:37.440 --> 00:29:42.800] But, like, first, I want to finish a story about Headline because you started it, it blew up, it's making money.
[00:29:43.120 --> 00:29:44.160] Just like, let's get a story out of it.
[00:29:44.800 --> 00:29:46.400] We did a really big detour there.
[00:29:46.480 --> 00:29:47.760] Huge detour.
[00:29:48.400 --> 00:29:49.280] How did you sell it?
[00:29:49.280 --> 00:29:50.880] And how much money did you make from selling it?
[00:29:50.880 --> 00:29:51.280] Because it's not really.
[00:29:51.520 --> 00:29:52.160] Twitter, dude.
[00:29:52.240 --> 00:29:53.680] This is all Twitter again.
[00:29:54.000 --> 00:29:59.360] I get reached out by sort of guys of jasper.ai and unbounce.
[00:29:59.360 --> 00:30:02.280] Unbounce.com, the landing page builder.
[00:30:02.520 --> 00:30:04.120] They both reached out via Twitter.
[00:29:59.920 --> 00:30:07.000] Dave from Jasper was just like super upfront.
[00:30:07.240 --> 00:30:10.200] Like, you just said, like, yeah, I want to buy you guys.
[00:30:10.200 --> 00:30:11.880] And they were a team of three people.
[00:30:11.880 --> 00:30:14.120] So I was hopping on calls with them to talk about it.
[00:30:14.120 --> 00:30:15.000] They just dropped the number.
[00:30:15.000 --> 00:30:16.040] I was like, double it.
[00:30:16.040 --> 00:30:17.080] And they were like, okay.
[00:30:17.080 --> 00:30:18.360] So we just started negotiating.
[00:30:18.360 --> 00:30:20.040] This was like super fast.
[00:30:20.040 --> 00:30:24.520] And I was getting burned out because I was doing customer support, development marketing, everything myself.
[00:30:24.520 --> 00:30:25.960] Like, I was done.
[00:30:25.960 --> 00:30:27.080] I was talking with my wife about it.
[00:30:27.160 --> 00:30:28.920] She was like, just sell it, dude.
[00:30:29.480 --> 00:30:34.600] So I was also entertaining offers from Unbounce, but they were like this obviously massive corporation.
[00:30:34.600 --> 00:30:36.120] So you have to go through procurement.
[00:30:36.120 --> 00:30:39.640] They have to wait two weeks to set up a meeting with the whole board.
[00:30:40.600 --> 00:30:43.560] And I was like, yeah, I have like an offer here.
[00:30:43.560 --> 00:30:44.760] I should just go through it.
[00:30:45.240 --> 00:30:46.680] I told them I don't want to work for them.
[00:30:46.680 --> 00:30:51.160] They were like, after two weeks, if you do the program, like the switch over, you're done with us.
[00:30:51.480 --> 00:30:51.880] Cool.
[00:30:51.880 --> 00:30:54.120] They asked me for Echo T, I was like, nah, I don't need it.
[00:30:54.120 --> 00:30:55.880] It's fucking stupid.
[00:30:57.160 --> 00:30:57.880] Oh, man.
[00:30:57.880 --> 00:31:02.440] Yeah, and then, and, yeah, I think it went through in four weeks.
[00:31:02.440 --> 00:31:07.880] I got 33% of the money in the bank, 66% over 12 months.
[00:31:08.280 --> 00:31:10.040] Two weeks in, I was done.
[00:31:10.040 --> 00:31:14.200] I was getting depressed afterwards because I went from working 16 hours a day to nothing.
[00:31:14.200 --> 00:31:15.480] I had nothing anymore.
[00:31:15.480 --> 00:31:15.880] Right.
[00:31:15.880 --> 00:31:16.280] Yeah.
[00:31:16.760 --> 00:31:18.520] It was odd.
[00:31:18.520 --> 00:31:21.800] A lot of people think they're going to quit and retire and go live on a beach.
[00:31:22.120 --> 00:31:24.120] But it's just like this grass is greener thing.
[00:31:24.360 --> 00:31:29.400] When you go from being somebody who does stuff constantly to being somebody who's like when you got money, but you're doing nothing.
[00:31:29.640 --> 00:31:33.480] The number one thing you want to do is start doing stuff again.
[00:31:33.480 --> 00:31:44.760] Well, but if your personality is built around you learning and doing things and suddenly you don't have anything like that, like your whole purpose in life, you're like, you're doing like, what is my purpose in life?
[00:31:44.760 --> 00:31:47.360] Yeah, was it to sell a company and make money and then do nothing?
[00:31:47.360 --> 00:31:48.800] Like, clearly not.
[00:31:48.800 --> 00:31:49.760] So, what's going to happen?
[00:31:44.840 --> 00:31:50.240] Let's talk about it.
[00:31:50.320 --> 00:31:57.680] Let's talk about this AI stuff because, like, I'm thinking, like, it just feels like there's a huge elephant in the room where half the stuff I build or consider building now.
[00:31:57.680 --> 00:32:01.760] I'm like, but isn't AI just going to make this obsolete in six months?
[00:32:01.760 --> 00:32:06.960] Like, we're building landing pages for indie hackers, like, basically a little profile.
[00:32:06.960 --> 00:32:16.400] And it's like, why would anybody ever use my website builder if six months from now or literally last week or two months ago, they could find some AI tool that will build them a landing page?
[00:32:16.720 --> 00:32:18.320] You got to build AI into it.
[00:32:18.320 --> 00:32:18.640] Right?
[00:32:18.640 --> 00:32:21.040] It's like, we have to sort of upgrade our stuff.
[00:32:21.040 --> 00:32:23.680] Like, we have to, or we're just going to be toast.
[00:32:24.000 --> 00:32:26.880] And it's like, it's so hard to wrap my mind around that.
[00:32:26.880 --> 00:32:28.800] Like, to some degree, I feel like a dinosaur.
[00:32:28.960 --> 00:32:32.320] I'm like, no, the old way of doing things is still here.
[00:32:32.320 --> 00:32:36.560] And the tried and true business fundamentals still apply.
[00:32:36.560 --> 00:32:39.760] But the technologist in me is like, no, it's all different.
[00:32:40.240 --> 00:32:41.920] Even just things in my life.
[00:32:42.080 --> 00:32:45.760] I want to spend a weekend just building an AI assistant for my own life.
[00:32:45.760 --> 00:32:46.960] I want it to text me.
[00:32:46.960 --> 00:32:48.000] I want it to email me.
[00:32:48.000 --> 00:32:49.120] I want it to call me.
[00:32:49.120 --> 00:32:50.160] I want it to remind me.
[00:32:50.160 --> 00:32:54.400] I want it to ask me questions about what I'm doing and then think and then tell me what I should do.
[00:32:54.400 --> 00:32:58.320] A year ago, this would have been extremely difficult to conceive of how it could even be.
[00:32:58.480 --> 00:33:01.280] It would be a huge, require a huge team to do something, right?
[00:33:01.280 --> 00:33:03.600] Well, now it's like, I can hook this up this weekend myself.
[00:33:03.600 --> 00:33:09.520] And like, I don't need a friend or a colleague or an expensive coach to motivate me or ask me what's going on.
[00:33:09.520 --> 00:33:12.800] Like, AI can do it extremely reliably.
[00:33:12.800 --> 00:33:14.480] It's the age of the entrepreneur, right?
[00:33:14.560 --> 00:33:19.360] Like, every repetitive job, everything that can be automated will be automated, right?
[00:33:19.360 --> 00:33:20.880] So, what stays?
[00:33:20.880 --> 00:33:21.200] Right.
[00:33:21.200 --> 00:33:25.520] And I think it's like the creativity in launching businesses.
[00:33:25.520 --> 00:33:27.840] Like, good luck trying to get that away with AI.
[00:33:27.840 --> 00:33:29.200] I don't think you can automate it.
[00:33:29.360 --> 00:33:30.360] And this is why I love AI.
[00:33:30.360 --> 00:33:33.160] Like, I don't like to do all these manual, boring stuff.
[00:33:33.160 --> 00:33:34.600] I don't want to make my own landing pages.
[00:33:29.840 --> 00:33:36.120] I don't want to write my own copy.
[00:33:36.280 --> 00:33:38.040] I don't want to do any of that.
[00:33:38.040 --> 00:33:41.880] I just want to build businesses, automate it, and move on to the next one.
[00:33:41.880 --> 00:33:46.520] So, I think that's why, like, as an entrepreneur, as an indie hacker, it's going to be an amazing time.
[00:33:46.520 --> 00:33:46.920] Agreed.
[00:33:46.920 --> 00:33:54.200] But, like, if you do a repetitive job, like financials, I think EBM just announced that they're going to fire 10% of their stuff.
[00:33:54.200 --> 00:33:55.080] And they said it's because of AI.
[00:33:56.280 --> 00:33:56.920] Yeah.
[00:33:56.920 --> 00:33:57.400] Because of AI.
[00:33:57.480 --> 00:33:58.360] Which I don't believe.
[00:33:58.360 --> 00:33:59.720] I don't think it's really because of AI.
[00:33:59.720 --> 00:34:04.280] I think that they needed to do layoffs and they're just what's going to happen to the SP.
[00:34:04.280 --> 00:34:10.040] Like, we're going into a recession, but all these companies just probably gonna save so much money in this.
[00:34:10.040 --> 00:34:15.320] Like, what is it going to do to the economy in like a stock market versus normal people way?
[00:34:15.320 --> 00:34:18.520] And I think, like, we, we are enthusiastic about it, right?
[00:34:18.600 --> 00:34:23.800] Like, but yeah, if you have an old job, yeah, I understand why you hate it.
[00:34:23.800 --> 00:34:26.680] So, there's this kind of hilarious book by this.
[00:34:26.680 --> 00:34:27.960] I think he was like an anthropologist.
[00:34:27.960 --> 00:34:32.360] He just died a couple of years ago, whatever, cognitive scientist, and it's called Bullshit Jobs.
[00:34:32.360 --> 00:34:39.720] And he's very like left-wing, so he's not coming at this from a perspective of like, you know, caring a lot about certain economic developments.
[00:34:39.720 --> 00:34:47.320] He really is just looking at our society and he's like, hey, capitalism creates a bunch of jobs that are horseshit, that aren't very meaningful, et cetera.
[00:34:47.720 --> 00:34:57.560] And his list of the jobs, I'll name a couple of them in a second, are the exact kinds of jobs that I think are in a lot of danger from AI.
[00:34:57.560 --> 00:35:07.240] And it makes me feel like there's going to be a lot of short-term pain and you know sort of a lot of job security is going to be gone there's you know this is going to be a very unstable place.
[00:35:07.240 --> 00:36:50.120] But in the long term, these types of jobs, like repetitive jobs, aren't the ones that I really imagine being that much loved lost for his book says over half of societal work is pointless and becomes psychologically destructive when paired with the work ethic that associates work with self-worth and so like that's like the like we're just looking this up now i haven't read this book but like this thesis of his book is that half the jobs we already have we shouldn't even have and then he has a list of like okay why what are these crappy jobs and so yeah and for context this is like 2018 so this wasn't you know he didn't see like he might have seen ai on the horizon but like you know he hadn't heard the term gpt before so flunkies right people that serve to make their superiors feel important goons who act a harm or to see others on behalf of their employer for example lobbyists corporate lawyers telemarketers duct tapers who temporarily fix problems that could be fixed permanently programmers repairing shoddy code airline desk staff who calm passengers whose bags do not arrive et cetera etc um and when i look at this list i'm just like oh that that is just the like the you know the first on the the victims that are going to get lined up against the wall from ai i disagree and it's like looking at his list like i don't like i think the things that he's saying are useless jobs are actually the jobs that are the safest the reason why is because a lot of the jobs that he's saying are useless are jobs that have to do with people's emotional state right like you lose your your bags at the airline you feel stressed you feel worried you're not sure if you're going to get it back people will pay money to feel better right or like flunkies he said of being a receptionist or being a door attendant or store greeter that just serve to make other people feel important people will pay money to feel important and feel good like i mean like there's some stuff in the ai landscape, like replica, like have an ai friend that you can talk to that is making some people feel good.
[00:36:50.120 --> 00:36:55.640] But I think that, generally speaking, for a while, people are mostly going to rely on other people to feel good.
[00:36:55.640 --> 00:36:59.400] And that's something ai is not going to be able to do, even when it can do everything else.
[00:36:59.400 --> 00:37:09.320] Which it seems like I'm pretty sure if an AI is gonna be nice to you and like tell you, oh, your bag will be back, like people would fucking hate it to be calmed by an AI in such a position.
[00:37:09.320 --> 00:37:19.720] I guess anyone who doesn't have a startup idea, they should get this book and just browse through it and be like, hmm, which one of these can I just like get a little piece of the pie on and automate out of it?
[00:37:20.120 --> 00:37:21.080] What are you doing for yourself?
[00:37:21.080 --> 00:37:28.920] Do you have any like behind the scenes Danny Postma AI tools that are making you more efficient and allowing you not to do repetitive tasks?
[00:37:28.920 --> 00:37:37.720] Mostly just rewriting my tweets, rewriting my content, because I'm not native English and I'm horrible in writing cohesion sentences for some reason.
[00:37:37.720 --> 00:37:44.920] So I will just write garbage in, like just do a brain fart of like 500 words for my tweets, put it in typefully.
[00:37:44.920 --> 00:37:46.840] Shout out to typefully.com.
[00:37:46.840 --> 00:37:50.600] Tell the AI to rewrite it in a cohesive way.
[00:37:50.600 --> 00:37:53.720] And then if the AI doesn't understand what I say, I know, okay.
[00:37:53.720 --> 00:37:57.400] Like if the AI doesn't understand what I say, no one is going to understand what I say.
[00:37:57.400 --> 00:38:03.480] Then I tell it to summarize it and then I tell it to make it a little bit more poppy and then I edit it for like 10 minutes.
[00:38:03.480 --> 00:38:05.240] So mostly those kind of things.
[00:38:05.240 --> 00:38:10.040] We should be doing this on the hacker scanning because we have a site where essentially people have to come in and write a lot.
[00:38:10.040 --> 00:38:13.320] They have to write posts, they have to write updates, they have to write comments.
[00:38:13.320 --> 00:38:16.120] We should be helping them write better content with.
[00:38:16.360 --> 00:38:19.000] I bet MicroAcquire already does this, for example.
[00:38:19.160 --> 00:38:24.120] If you go to microacquire.com, I think all their posts are written in the same way.
[00:38:24.120 --> 00:38:25.960] I wouldn't be surprised if they have.
[00:38:25.960 --> 00:38:26.360] Yeah.
[00:38:26.600 --> 00:38:36.680] You just put all these content in and then have an AI to generate it for them and then probably like one reviewer that makes some manual changes towards it.
[00:38:36.680 --> 00:38:37.080] Yeah.
[00:38:37.640 --> 00:38:39.000] I did this for Headline back in the day.
[00:38:39.000 --> 00:38:42.240] Like the first customers, they told me the output was garbage.
[00:38:42.240 --> 00:38:45.360] And you just look at the input and it's like, yeah, you input garbage.
[00:38:44.200 --> 00:38:47.200] Like input garbage in whoever's tied.
[00:38:47.200 --> 00:38:48.960] It's the same with Headshot Pro right now.
[00:38:44.600 --> 00:38:51.200] Like we refund 20% of the people because they upload garbage.
[00:38:51.840 --> 00:38:57.920] So for Headline, I was actually rewriting their inputs in the back end without them seeing it just to make it better.
[00:38:57.920 --> 00:38:59.280] So stuff like that is.
[00:38:59.280 --> 00:38:59.840] Yeah.
[00:38:59.840 --> 00:39:02.400] Well, this is one of the things that's already getting better.
[00:39:02.400 --> 00:39:04.960] Like the difference between GPT 3.5.
[00:39:04.960 --> 00:39:10.160] So GPT is like, you know, OpenAI's family of text-based AIs.
[00:39:10.160 --> 00:39:14.400] And you can use chat GPT to interface with it, or you can use the API.
[00:39:14.400 --> 00:39:19.360] But as technology gets better, you don't have to be as good of a prompt whiz kid.
[00:39:19.360 --> 00:39:29.520] Like right now, the people who are the best at using AI are just really good at understanding how the AI thinks, what kinds of things they should put in their prompt and their questions and their conversations with the AI to get it to do what they want.
[00:39:29.520 --> 00:39:32.560] But as it gets smarter, it gets more and more human-esque.
[00:39:32.720 --> 00:39:38.720] You could imagine five or ten years from now, maybe even a year from now, you could write absolute garbage.
[00:39:38.720 --> 00:39:40.080] Mid-Journey does it right now.
[00:39:40.080 --> 00:39:46.080] You can just type in and Midjourney actually optimizes your prompt and adds all these specifications to it so the AI better understands what you mean.
[00:39:46.080 --> 00:39:46.640] Boom.
[00:39:46.640 --> 00:39:47.200] There it is.
[00:39:47.520 --> 00:39:48.240] So yeah.
[00:39:48.240 --> 00:39:48.640] Yeah.
[00:39:48.960 --> 00:39:50.880] Are you guys using any tools I should know of?
[00:39:50.880 --> 00:39:52.720] Because it's actually quite stupid.
[00:39:52.720 --> 00:39:53.920] I only use one AI tool.
[00:39:54.880 --> 00:40:07.600] It's hard because I think being a builder and being an eddy hacker, I would always get stuck in these creative black holes where I would just be spending so much of my time creating that I would just get really behind on like using, right?
[00:40:07.920 --> 00:40:10.240] And I think I see like Peter Levels is like this too.
[00:40:10.240 --> 00:40:12.240] He's like, I'm using PHP, blah, blah, blah, blah.
[00:40:12.320 --> 00:40:14.320] He's like, he doesn't care how his code looks, right?
[00:40:14.320 --> 00:40:15.680] He's just trying to create stuff.
[00:40:15.680 --> 00:40:20.640] He's not trying to be like up to date with the latest technology and build systems and whatever as an engineer.
[00:40:20.640 --> 00:40:22.080] And I feel like that too as an Eddie hacker.
[00:40:22.080 --> 00:40:24.560] I'm like, I'm trying to like create value for other people.
[00:40:24.560 --> 00:40:30.840] So I'm not like sitting around like, you know, licking my finger and sticking it in the air and like trying to figure out what toys I should be playing with.
[00:40:30.840 --> 00:40:31.640] Like I'm building.
[00:40:29.760 --> 00:40:32.840] And that's kind of how you are too.
[00:40:33.160 --> 00:40:40.840] But like in a way, it's like we kind of need to strike this balance between like becoming more efficient builders and continuing to build.
[00:40:40.840 --> 00:40:49.720] I would say that that is something that like basically sort of, you know, what balance do you strike between like building in the ways that you know and exploring, right?
[00:40:49.720 --> 00:40:56.040] Like basically sort of having almost like a skunk works built into your own like way of working.
[00:40:56.040 --> 00:40:58.280] Basically what you should do is you should look at the times.
[00:40:58.280 --> 00:41:03.480] I think three or four years ago, hey, you know, this is the big, you know, the crypto wave is coming.
[00:41:03.480 --> 00:41:07.320] I think that, you know, some of the other waves that have come, no code, et cetera.
[00:41:07.320 --> 00:41:11.720] I think you can kind of poke your head out of the hole and go, ah, that's not really for me, right?
[00:41:11.720 --> 00:41:13.160] That's not revolutionary.
[00:41:13.160 --> 00:41:14.120] It's not a huge change.
[00:41:14.120 --> 00:41:18.600] I don't necessarily need to divide my time away from the things that I know are working and that I know.
[00:41:18.680 --> 00:41:20.600] None of those like made you more productive.
[00:41:20.600 --> 00:41:26.440] Maybe, maybe the creator economy thing where it's like, oh, hey, you can make money without writing any code.
[00:41:26.440 --> 00:41:34.760] You could just basically show your work, build in public, have an additional revenue channel that's lighter weight than building a full fledged SaaS app.
[00:41:34.760 --> 00:41:37.560] Maybe like that was like, it's not really making you more productive.
[00:41:37.560 --> 00:41:39.640] It's just like, oh, there's easier ways to make money.
[00:41:39.640 --> 00:41:43.960] But now it's like, oh, like the new sort of technological wave is actually making you more productive.
[00:41:43.960 --> 00:41:46.040] To me, it's AI and then the gig economy.
[00:41:46.040 --> 00:41:54.680] The gig economy, it becomes way easier, for example, to quickly hire a contractor or quickly get someone to whip you up a quick design, like a quick logo.
[00:41:55.160 --> 00:41:55.640] That kind of thing.
[00:41:55.640 --> 00:41:57.720] It's like, hey, you need to jump on that bus, right?
[00:41:57.720 --> 00:42:04.920] Like take a long weekend and just figure this out because it's going to, I don't know, that was, you know, 0.3x your productivity.
[00:42:04.920 --> 00:42:11.080] Now AI is here and it's like, dude, you might 0.5, you know, 1x, 2x your productivity, right?
[00:42:11.080 --> 00:42:11.800] Easily.
[00:42:11.800 --> 00:42:30.960] I wonder when someone is going to launch Fiverr with an extra A in the site, where five, like AI Fiverr, and it's just going to get all the jobs that are listed on their copywriting, checking, translating, is basically just going to have the same UI on it, have the gig works done by an AI, so it doesn't feel like an AI is doing it anymore.
[00:42:30.960 --> 00:42:36.560] I've seen at least five or six AI landing page generator tools out there right now, right?
[00:42:36.560 --> 00:42:38.240] It's like people are already doing this.
[00:42:38.640 --> 00:42:47.440] 100%, by the way, when this episode ends, I'll give it two days, and then there's going to be someone who launches and it's going to be called one or because Fiverr is about $5.
[00:42:47.440 --> 00:42:52.800] Now there's going to be an AI tool, it's going to be five times less the operating cost, and they're going to call it Wanner.
[00:42:53.120 --> 00:42:54.800] But I think that's what people are doing wrong.
[00:42:54.800 --> 00:42:59.760] For example, a friend of mine, Mark, they have this landing page generated, but they priced it $19.
[00:42:59.760 --> 00:43:06.880] But you need to understand, like, you are removing the value of a landing page designer that's $5,000.
[00:43:06.880 --> 00:43:11.040] Like, you don't have to price it cheap just because it is cheap.
[00:43:11.040 --> 00:43:19.840] Like, for Hetshot Pro, for example, like, it cost me a few dollars to do it, but I'm not going to price it at $10 because you're competing with a photo shoot.
[00:43:19.840 --> 00:43:20.400] Right.
[00:43:20.400 --> 00:43:30.560] And as long as they're not competitors and you're not going, like, as long as you're like a fancy tool and you know what content you're doing and you're good, like you can charge a high price now while it's still new.
[00:43:30.560 --> 00:43:35.200] Like probably in five years, this doesn't work anymore, but for now, be expensive, man.
[00:43:35.200 --> 00:43:36.880] Dude, this is the exact point.
[00:43:36.880 --> 00:43:40.560] This is the call back to the point about the benefit of being fast.
[00:43:40.560 --> 00:43:53.000] I mean, look, if the ultimate cost of like, you know, sort of whatever it is, right, a headshot builder for AI or whatever is, you know, say it's like, you know, a 0.5, you know, or it's five times less expensive to create it.
[00:43:53.000 --> 00:43:59.280] Look, in the fullness of time, a bunch of competitors are going to like, you know, sort of enter into the space and drive down the cost.
[00:43:59.280 --> 00:44:05.800] But if you're fast, then the pricing theory just becomes, right, what's the alternative to this, right?
[00:44:05.800 --> 00:44:07.560] What is the incumbent or whatever it is?
[00:44:07.880 --> 00:44:14.680] This is what Peter, this, what Peter did when he launched the Avatar AI, he charged $50, and he was the only one doing it.
[00:44:15.000 --> 00:44:18.840] So he could just charge whatever the fuck he wants because everyone wants to use it.
[00:44:18.840 --> 00:44:23.080] And then gradually I came in, we had to like lower a little bit of a price.
[00:44:23.080 --> 00:44:27.080] And then, of course, Lanza came out and we had to slice our price to 80% down.
[00:44:27.080 --> 00:44:31.240] I think we were at $40 and we had to go down to $10, $7.
[00:44:31.240 --> 00:44:40.280] But most of our revenue, I think Peter got like 90% of his revenue in the first weeks because he could charge so much money and then everyone had to go down.
[00:44:40.280 --> 00:44:43.320] So yeah, this is coming back again to your point and what I said.
[00:44:43.800 --> 00:44:48.920] If it's a hype, get the most money out immediately because you can.
[00:44:48.920 --> 00:44:50.680] You're not going to last.
[00:44:50.680 --> 00:44:56.520] There's so many reasons right now the best strategy is to have small products that you move fast on.
[00:44:56.520 --> 00:44:59.480] Because number one, you have no idea when you're going to be disrupted.
[00:44:59.480 --> 00:45:05.800] These incumbents might come by and crush you and whatever money you were able to make for six weeks or six months, like that's now gone.
[00:45:05.800 --> 00:45:08.600] You might be disrupted not only by incumbents, but by technology itself.
[00:45:08.600 --> 00:45:11.160] What you're building might become completely obsolete.
[00:45:11.160 --> 00:45:22.040] Like one of the things I see right now that's happening is everybody's kind of taking existing applications and slapping AI on top of it and getting like this 5x boost of like, hey, this is like five times better than what I had in the past, which is awesome.
[00:45:22.040 --> 00:45:29.160] But like, okay, we're still waiting for the wave of companies to come that are building products from the ground up with this AI in mind.
[00:45:29.160 --> 00:45:30.520] And that's going to be much better.
[00:45:30.520 --> 00:45:44.280] Like a community forum, for example, that helps you occasionally, you know, use AI, or like a social product like Twitter, that helps you occasionally use AI to like write better tweets, is not going to be as good as a social network that was built from the ground up with AI in mind.
[00:45:44.280 --> 00:45:45.600] And that stuff is coming, right?
[00:45:44.440 --> 00:45:47.040] Like that stuff's going to stand the test of time.
[00:45:47.280 --> 00:45:56.240] And that's going to lead to another second-order effect where a world full of all these AI products is going to create a whole bunch of new problems that we can't even imagine right now that are hard to imagine.
[00:45:56.240 --> 00:45:58.400] And there's going to be startups that solve those problems.
[00:45:58.400 --> 00:46:05.440] And so if you're building something that's going to take 10 years or five years to slowly come to fruition, I think you're dead.
[00:46:05.440 --> 00:46:10.080] Yeah, this is why as indie hacker, I think don't compete with the big companies.
[00:46:10.080 --> 00:46:12.480] Like I see so many people, I made the same mistake.
[00:46:12.960 --> 00:46:17.040] The first AI thing I made was like a stock photo, AI stock photo website.
[00:46:17.040 --> 00:46:20.720] But it's fucking dumb because Adobe has all the photos and libraries.
[00:46:21.040 --> 00:46:22.560] Why would I compete with it?
[00:46:22.560 --> 00:46:25.600] You have three months of time, and then someone else will come out.
[00:46:25.600 --> 00:46:30.160] And I see other people, like, they make this AI generator where you can generate pictures.
[00:46:30.160 --> 00:46:31.040] It's like an editor.
[00:46:31.040 --> 00:46:34.560] But Adobe just launched the Firefly thing and you're done for.
[00:46:34.720 --> 00:46:36.320] There's no way you can compete with it.
[00:46:36.560 --> 00:46:40.160] You should find a niche, ship really fast in that niche.
[00:46:40.160 --> 00:46:43.920] Because a big company is not going to outcompete a niche in that sense.
[00:46:43.920 --> 00:46:44.160] Nope.
[00:46:44.160 --> 00:46:45.360] It's not worth it to them.
[00:46:45.360 --> 00:46:59.920] And I think this raises an interesting question, which is like every time there's a technological shift, some percentage of it goes to the incumbents, the big players, some percentage of it goes to the startups, and some percentage of it goes like even a level below, like the anti-hackers.
[00:46:59.920 --> 00:47:02.240] Like I'm trying to make some money online right now, right?
[00:47:02.240 --> 00:47:06.960] Like with something like crypto, for example, the incumbents didn't make very much money from crypto.
[00:47:06.960 --> 00:47:15.120] All the biggest crypto companies were kind of startups because they moved really fast and it was like kind of niche enough that big companies were like, I don't know about this.
[00:47:15.120 --> 00:47:18.960] Whereas something like mobile, for example, all the incumbents took all 

Prompt 2: Key Takeaways

Now please extract the key takeaways from the transcript content I provided.

Extract the most important key takeaways from this part of the conversation. Use a single sentence statement (the key takeaway) rather than milquetoast descriptions like "the hosts discuss...". 

Limit the key takeaways to a maximum of 3. The key takeaways should be insightful and knowledge-additive. 

IMPORTANT: Return ONLY valid JSON, no explanations or markdown. Ensure:
- All strings are properly quoted and escaped
- No trailing commas
- All braces and brackets are balanced
Format: {"key_takeaways": ["takeaway 1", "takeaway 2"]}

Prompt 3: Segments

Now identify 2-4 distinct topical segments from this part of the conversation.

For each segment, identify:
- Descriptive title (3-6 words)  
- START timestamp when this topic begins (HH:MM:SS format)
- Double check that the timestamp is accurate - a timestamp will NEVER be greater than the total length of the audio
- Most important Key takeaway from that segment. Key takeaway must be specific and knowledge-additive.
- Brief summary of the discussion

IMPORTANT: The timestamp should mark when the topic/segment STARTS, not a range. Look for topic transitions and conversation shifts.

Return ONLY valid JSON. Ensure all strings are properly quoted, no trailing commas:
{
  "segments": [
    {
      "segment_title": "Topic Discussion", 
      "timestamp": "01:15:30",
      "key_takeaway": "main point from this segment",
      "segment_summary": "brief description of what was discussed"
    }
  ]
}

Timestamp format: HH:MM:SS (e.g., 00:05:30, 01:22:45) marking the START of each segment.

Prompt 4: Media Mentions

Now scan the transcript content I provided for ACTUAL mentions of specific media titles:

Find explicit mentions of:
- Books (with specific titles)
- Movies (with specific titles)  
- TV Shows (with specific titles)
- Music/Songs (with specific titles)

DO NOT include:
- Websites, URLs, or web services
- Other podcasts or podcast names

IMPORTANT: 
- Only include items explicitly mentioned by name. Do not invent titles.
- Valid categories are: "Book", "Movie", "TV Show", "Music"
- Include the exact phrase where each item was mentioned
- Find the nearest proximate timestamp where it appears in the conversation
- THE TIMESTAMP OF THE MEDIA MENTION IS IMPORTANT - DO NOT INVENT TIMESTAMPS AND DO NOT MISATTRIBUTE TIMESTAMPS
- Double check that the timestamp is accurate - a timestamp will NEVER be greater than the total length of the audio
- Timestamps are given as ranges, e.g. 01:13:42.520 --> 01:13:46.720. Use the EARLIER of the 2 timestamps in the range.

Return ONLY valid JSON. Ensure all strings are properly quoted and escaped, no trailing commas:
{
  "media_mentions": [
    {
      "title": "Exact Title as Mentioned",
      "category": "Book",
      "author_artist": "N/A",
      "context": "Brief context of why it was mentioned",
      "context_phrase": "The exact sentence or phrase where it was mentioned",
      "timestamp": "estimated time like 01:15:30"
    }
  ]
}

If no media is mentioned, return: {"media_mentions": []}

Prompt 5: Context Setup

You are an expert data extractor tasked with analyzing a podcast transcript. 
I will provide you with part 2 of 2 from a podcast transcript. 
I will then ask you to extract different types of information from this content in subsequent messages. Please confirm you have received and understood the transcript content.

Transcript section:
ker, I think don't compete with the big companies.
[00:46:10.080 --> 00:46:12.480] Like I see so many people, I made the same mistake.
[00:46:12.960 --> 00:46:17.040] The first AI thing I made was like a stock photo, AI stock photo website.
[00:46:17.040 --> 00:46:20.720] But it's fucking dumb because Adobe has all the photos and libraries.
[00:46:21.040 --> 00:46:22.560] Why would I compete with it?
[00:46:22.560 --> 00:46:25.600] You have three months of time, and then someone else will come out.
[00:46:25.600 --> 00:46:30.160] And I see other people, like, they make this AI generator where you can generate pictures.
[00:46:30.160 --> 00:46:31.040] It's like an editor.
[00:46:31.040 --> 00:46:34.560] But Adobe just launched the Firefly thing and you're done for.
[00:46:34.720 --> 00:46:36.320] There's no way you can compete with it.
[00:46:36.560 --> 00:46:40.160] You should find a niche, ship really fast in that niche.
[00:46:40.160 --> 00:46:43.920] Because a big company is not going to outcompete a niche in that sense.
[00:46:43.920 --> 00:46:44.160] Nope.
[00:46:44.160 --> 00:46:45.360] It's not worth it to them.
[00:46:45.360 --> 00:46:59.920] And I think this raises an interesting question, which is like every time there's a technological shift, some percentage of it goes to the incumbents, the big players, some percentage of it goes to the startups, and some percentage of it goes like even a level below, like the anti-hackers.
[00:46:59.920 --> 00:47:02.240] Like I'm trying to make some money online right now, right?
[00:47:02.240 --> 00:47:06.960] Like with something like crypto, for example, the incumbents didn't make very much money from crypto.
[00:47:06.960 --> 00:47:15.120] All the biggest crypto companies were kind of startups because they moved really fast and it was like kind of niche enough that big companies were like, I don't know about this.
[00:47:15.120 --> 00:47:18.960] Whereas something like mobile, for example, all the incumbents took all the money, right?
[00:47:18.960 --> 00:47:21.360] Like Apple and Google created the phones.
[00:47:21.360 --> 00:47:23.600] Every big website made its own mobile app.
[00:47:23.600 --> 00:47:26.400] And like very few startups got in there and disrupted things.
[00:47:26.400 --> 00:47:28.480] I'm curious what it's going to be like with AI.
[00:47:28.480 --> 00:47:37.800] Like you were saying, like Adobe, like they can just slap AI into Photoshop and have a giant gallery of images and AI effects, and they're just going to kill everything because everybody already has Photoshop.
[00:47:37.800 --> 00:47:45.160] But from the bottom up, indie hackers can do what you're saying and target these super tiny niches that no incumbent's going to target.
[00:47:45.160 --> 00:47:51.480] And I think what's really cool is the fact that it's so fast and easy to build these AI apps now because you're basically writing prompts.
[00:47:51.480 --> 00:47:53.560] You're not writing a ton of code.
[00:47:53.560 --> 00:48:04.520] That there's no reason not to have super specialized apps for tons of little niches, which I think is going to make it harder for anyone to get a bigger foothold with a bigger, broader company that's targeting more people.
[00:48:04.520 --> 00:48:13.640] Like to your idea of like a Fiverr powered by AI, I can imagine 100 different profiles of AI copywriters, but they all have their own little different flavor because they all have a different prompt.
[00:48:13.640 --> 00:48:15.320] And essentially, you could pick a different one.
[00:48:15.560 --> 00:48:18.680] There isn't just going to be one AI copywriter to rule them all.
[00:48:18.680 --> 00:48:23.960] Why would there be if you could make a specialized AI that could be super duper good at each individual task?
[00:48:23.960 --> 00:48:30.040] Yeah, I think what I was shocked with on my Twitter is like, it feels like almost no indie hackers is doing SEO.
[00:48:30.040 --> 00:48:32.760] Well, SEO, like we indie hackers, we hate marketing, right?
[00:48:32.760 --> 00:48:43.400] And this is the easiest way to A, find a product idea, B, launch a niche product, and then C, don't have to do any marketing for it because it just stands there.
[00:48:43.400 --> 00:48:46.680] So that was like the lucky part I had with the ProvoPictor website.
[00:48:46.680 --> 00:48:48.280] Like there was no competition on it.
[00:48:48.280 --> 00:48:52.040] So I would just launch my website, get a backlink, and then just rank instantly.
[00:48:52.040 --> 00:48:54.520] So you can find like hyper-specific niches.
[00:48:54.520 --> 00:48:55.320] You don't have to compete.
[00:48:55.320 --> 00:48:56.280] You don't have to do marketing.
[00:48:56.280 --> 00:49:02.440] And you would just keep it because why would any other big company go down on that keyword for you?
[00:49:02.760 --> 00:49:04.200] By the way, I was going to ask you that.
[00:49:04.680 --> 00:49:15.600] You mentioned at the beginning of this call, actually, that you spend a lot of time on Google looking at keywords in relation to a lot of these, like the latest of your products on Postgrafts.
[00:49:14.840 --> 00:49:21.120] And then you also, you don't, you know, sort of, you're not building cool AI to help you with your own processes.
[00:49:21.280 --> 00:49:23.680] Like, you're not necessarily scratching your own itch.
[00:49:23.680 --> 00:49:30.240] Like, is Google is like, are you know, keyword search and SEO the thing that you use to find ideas?
[00:49:30.880 --> 00:49:33.280] Yeah, well, to find ideas, kind of.
[00:49:33.280 --> 00:49:38.800] So last year I started trying to do this with the program with programmatic SEO, like Fig components, tailbits.
[00:49:39.360 --> 00:49:42.080] These were just completely based on SEO research.
[00:49:42.080 --> 00:49:49.840] I was just searching all day, all week for ideas what I could build next without AI, because AI generation didn't exist yet and I had an On Copy.
[00:49:49.920 --> 00:49:52.000] So what am I going to do?
[00:49:52.400 --> 00:49:54.400] So yeah, I built based on that.
[00:49:54.400 --> 00:49:59.840] And this time, so with the Avatar race, everyone chose the avatar work, right?
[00:49:59.840 --> 00:50:01.120] But there's no searches on it.
[00:50:01.120 --> 00:50:04.320] So I chose Profile Picture specifically.
[00:50:04.320 --> 00:50:08.320] So I got the idea before as something I want to explore.
[00:50:08.320 --> 00:50:18.720] But then when I want to launch something, I also check, hey, which keyword is underrepresented, which is on the keyboard difficulty of 10, and then use that in your domain name, in your titles.
[00:50:18.720 --> 00:50:24.640] And this is like if you put my website in href, like I have a shit ton of traffic on Alders Profile Picture.
[00:50:24.640 --> 00:50:29.760] Well, I know, and this is why levels, YPETA went down, for example, because there are no searches on Avatar.
[00:50:29.760 --> 00:50:32.960] So once the hype is gone, there is nothing sustainable in it.
[00:50:32.960 --> 00:50:34.480] And the same with Headshot right now, right?
[00:50:35.040 --> 00:50:38.960] I get hype on TikTok, I get backlinks, but there is a lot of search on Headshot.
[00:50:38.960 --> 00:50:47.280] So you take advantage of the hype, and then you build for the longevity of it based on the keywords.
[00:50:47.280 --> 00:50:49.680] What do you think about the future of being like a founder, right?
[00:50:49.680 --> 00:50:52.960] Like it's cool because it's kind of like a golden era for indie hackers.
[00:50:52.960 --> 00:50:54.720] We're going to be much more efficient.
[00:50:54.720 --> 00:50:56.080] We can build much more.
[00:50:56.080 --> 00:50:57.920] But also, the competition is ramped up.
[00:50:57.920 --> 00:51:02.440] Way more people now can build apps that do very impressive things.
[00:50:59.760 --> 00:51:07.000] Any idea that's basic, there's going to be 10 million versions of it in like two days.
[00:51:07.880 --> 00:51:09.320] And so you're going to have to be super creative.
[00:51:09.320 --> 00:51:12.200] But even then, people can copy you super duper quickly.
[00:51:12.200 --> 00:51:14.120] So where is the defensibility?
[00:51:14.120 --> 00:51:14.840] Where are the moats?
[00:51:14.840 --> 00:51:17.320] Is it all just going to be these fast cash grabs?
[00:51:17.320 --> 00:51:21.320] Are there businesses that are going to be immune to this and that can stand the test of time?
[00:51:21.320 --> 00:51:24.680] Well, the thing is, I have a lot of competition with Hedgehog Pro and ProfoPicts, right?
[00:51:24.680 --> 00:51:25.560] But I think I have a mode.
[00:51:25.560 --> 00:51:34.680] Like all my competitors with Headshot, they have a lower quality because I figured out how to make my own custom model that looks more real life using some other techniques.
[00:51:34.680 --> 00:51:36.040] Like it's not just one AI.
[00:51:36.040 --> 00:51:38.840] It's like 10, 20 AI stacked on top of each other.
[00:51:38.840 --> 00:51:41.560] So you have like that is a moat.
[00:51:41.560 --> 00:51:45.720] On the other side, if you don't want competition, don't tweet too much about it.
[00:51:45.720 --> 00:51:47.000] Just keep your revenue numbers quiet.
[00:51:47.560 --> 00:51:52.600] A niche that's really boring that no one wants to jump into and just like be caching.
[00:51:52.600 --> 00:51:54.600] Yeah, don't come on the Andy Hackers podcast.
[00:51:55.000 --> 00:51:56.120] I mean, you've been through a lot, Danny.
[00:51:56.120 --> 00:51:57.320] You sold a company.
[00:51:57.320 --> 00:52:00.280] You have launching products, it seems like every other week.
[00:52:00.600 --> 00:52:05.320] You're at the forefront of like these new AI image generation photo editing tools.
[00:52:05.320 --> 00:52:11.720] What's something you've learned that you think other indie hackers who are just getting started, I mean, people listening to the show, some of them don't know how to code.
[00:52:11.720 --> 00:52:13.160] Many have never built anything.
[00:52:13.160 --> 00:52:16.600] Some people are two years into a product that's not making any money.
[00:52:16.600 --> 00:52:19.560] What do you think they can take away from your story?
[00:52:19.880 --> 00:52:29.320] Ship fast, build a lot of different things to find what works, and do SEO for marketing if you don't like marketing.
[00:52:29.320 --> 00:52:34.360] I think those are the three things to get you out of the slum.
[00:52:34.360 --> 00:52:34.840] Love it.
[00:52:34.840 --> 00:52:36.520] Danny, thanks again for coming on the show.
[00:52:36.520 --> 00:52:39.080] Can you tell people where they can find your stuff?
[00:52:39.080 --> 00:52:41.160] Is it just postcrafts.com?
[00:52:41.160 --> 00:52:42.680] Yeah, you can, mostly on my Twitter.
[00:52:42.680 --> 00:52:52.000] So if you go to twitter.com/slash Danny Postma with two A's in the end because the other one wasn't available, I post there with everything that I learned.
[00:52:52.160 --> 00:52:54.880] Yeah, I actually own the other, I own the other handle.
[00:52:54.880 --> 00:52:56.240] I just cannot get access to it.
[00:52:56.240 --> 00:52:56.720] It sucks.
[00:52:56.720 --> 00:52:57.200] Nice.
[00:52:57.520 --> 00:52:58.320] Yeah, just follow me there.
[00:52:58.320 --> 00:53:00.960] And there's like a bunch of links to all my latest projects, what I'm working on.
[00:53:00.960 --> 00:53:03.040] I'm sharing all my struggles, my learnings.
[00:53:03.440 --> 00:53:04.000] Yeah.
[00:53:04.000 --> 00:53:04.240] All right.
[00:53:04.240 --> 00:53:05.360] We'll put you in the show notes, too.
[00:53:05.440 --> 00:53:06.000] Thanks again, Danny.
[00:53:06.480 --> 00:53:07.520] Cheers, guys.

Prompt 6: Key Takeaways

Now please extract the key takeaways from the transcript content I provided.

Extract the most important key takeaways from this part of the conversation. Use a single sentence statement (the key takeaway) rather than milquetoast descriptions like "the hosts discuss...". 

Limit the key takeaways to a maximum of 3. The key takeaways should be insightful and knowledge-additive. 

IMPORTANT: Return ONLY valid JSON, no explanations or markdown. Ensure:
- All strings are properly quoted and escaped
- No trailing commas
- All braces and brackets are balanced
Format: {"key_takeaways": ["takeaway 1", "takeaway 2"]}

Prompt 7: Segments

Now identify 2-4 distinct topical segments from this part of the conversation.

For each segment, identify:
- Descriptive title (3-6 words)  
- START timestamp when this topic begins (HH:MM:SS format)
- Double check that the timestamp is accurate - a timestamp will NEVER be greater than the total length of the audio
- Most important Key takeaway from that segment. Key takeaway must be specific and knowledge-additive.
- Brief summary of the discussion

IMPORTANT: The timestamp should mark when the topic/segment STARTS, not a range. Look for topic transitions and conversation shifts.

Return ONLY valid JSON. Ensure all strings are properly quoted, no trailing commas:
{
  "segments": [
    {
      "segment_title": "Topic Discussion", 
      "timestamp": "01:15:30",
      "key_takeaway": "main point from this segment",
      "segment_summary": "brief description of what was discussed"
    }
  ]
}

Timestamp format: HH:MM:SS (e.g., 00:05:30, 01:22:45) marking the START of each segment.

Prompt 8: Media Mentions

Now scan the transcript content I provided for ACTUAL mentions of specific media titles:

Find explicit mentions of:
- Books (with specific titles)
- Movies (with specific titles)  
- TV Shows (with specific titles)
- Music/Songs (with specific titles)

DO NOT include:
- Websites, URLs, or web services
- Other podcasts or podcast names

IMPORTANT: 
- Only include items explicitly mentioned by name. Do not invent titles.
- Valid categories are: "Book", "Movie", "TV Show", "Music"
- Include the exact phrase where each item was mentioned
- Find the nearest proximate timestamp where it appears in the conversation
- THE TIMESTAMP OF THE MEDIA MENTION IS IMPORTANT - DO NOT INVENT TIMESTAMPS AND DO NOT MISATTRIBUTE TIMESTAMPS
- Double check that the timestamp is accurate - a timestamp will NEVER be greater than the total length of the audio
- Timestamps are given as ranges, e.g. 01:13:42.520 --> 01:13:46.720. Use the EARLIER of the 2 timestamps in the range.

Return ONLY valid JSON. Ensure all strings are properly quoted and escaped, no trailing commas:
{
  "media_mentions": [
    {
      "title": "Exact Title as Mentioned",
      "category": "Book",
      "author_artist": "N/A",
      "context": "Brief context of why it was mentioned",
      "context_phrase": "The exact sentence or phrase where it was mentioned",
      "timestamp": "estimated time like 01:15:30"
    }
  ]
}

If no media is mentioned, return: {"media_mentions": []}

Full Transcript

[00:00:06.640 --> 00:00:11.840] I feel like there's something in the water of being an indie hacker where you feel like you need to be really friendly.
[00:00:11.840 --> 00:00:12.960] Like everyone's a community.
[00:00:12.960 --> 00:00:14.400] Everyone helps out.
[00:00:14.400 --> 00:00:19.520] So people don't think about having rivalries when that really is.
[00:00:19.520 --> 00:00:23.680] I mean, if you take Danny and Peter's situation, they aren't enemies.
[00:00:24.320 --> 00:00:25.600] There's no bad blood.
[00:00:26.240 --> 00:00:27.600] You can be a rival and be friendly.
[00:00:28.160 --> 00:00:29.840] Yeah, you can be a rival and be friendly.
[00:00:30.080 --> 00:00:41.520] And then on top of all that, there's also like, Courtland, you probably have seen, like, I mean, I think I've counted no fewer than 10 individual posts on indie hackers complaining about copycats.
[00:00:41.520 --> 00:00:43.920] Like, people just have this idea that you can't copy.
[00:00:44.480 --> 00:00:46.560] If someone else is doing something, like, that's their lane.
[00:00:46.560 --> 00:00:47.760] You got to stay in your lane.
[00:00:47.760 --> 00:00:53.760] And there's no bigger bullshit in the world of business than the idea that you can't have the same thing.
[00:00:53.920 --> 00:00:57.280] It'd be like playing basketball and somebody steals the ball from you.
[00:00:57.280 --> 00:00:58.000] And then you complain.
[00:00:58.160 --> 00:00:59.680] Like, yeah, that's not, that shouldn't be allowed.
[00:00:59.680 --> 00:01:02.480] It's like, actually, that's like part of the rules of the game.
[00:01:02.480 --> 00:01:07.200] Like, one of the rules of business is that if somebody innovates, like, you can do a similar innovation.
[00:01:07.200 --> 00:01:09.600] Someone does a slam dunk and you're like, well, I can't do a slam dunk.
[00:01:09.600 --> 00:01:10.640] Like, that's Michael's move.
[00:01:10.640 --> 00:01:12.160] You know, like, I got to do the crossover.
[00:01:12.160 --> 00:01:13.600] Like, I can't do that same thing.
[00:01:13.600 --> 00:01:14.080] Yeah.
[00:01:14.080 --> 00:01:16.080] Speaking of Danny, Danny, what's up?
[00:01:16.080 --> 00:01:16.880] What's up, guys?
[00:01:16.880 --> 00:01:17.760] Thanks for having me.
[00:01:17.760 --> 00:01:18.000] Hello.
[00:01:18.160 --> 00:01:19.520] Thanks for joining.
[00:01:19.520 --> 00:01:23.600] We're just talking about you and your sort of AI rival, Peter Levels.
[00:01:23.600 --> 00:01:25.040] Do you think it's healthy to have a rivalry?
[00:01:25.600 --> 00:01:27.360] Would you call it a rivalry?
[00:01:27.680 --> 00:01:36.240] I wouldn't call it a rivalry because we've been in Telegram chats for 24-7, just literally sharing all the information we had together when we launched the Provo Picture apps.
[00:01:36.240 --> 00:01:40.240] And then on Twitter, we had like this little rivalry, I guess.
[00:01:40.240 --> 00:01:40.960] There it is.
[00:01:41.920 --> 00:01:43.280] Personally, for me, I gained...
[00:01:43.360 --> 00:01:49.440] I went from 50k followers 1.5 to 60 because Peter was tweeting about me all the time, right?
[00:01:49.440 --> 00:01:58.160] So it feels like we had this little YouTube collaboration where some people put like another YouTuber on the video and then have like cross-followings and stuff.
[00:01:58.160 --> 00:02:00.120] So it really helped me out.
[00:01:59.680 --> 00:02:03.640] Rivalry on the outside and it's warm and fuzzy teddy bears on the inside.
[00:01:59.760 --> 00:02:04.600] Exactly, exactly.
[00:02:04.920 --> 00:02:09.000] Yeah, there's 37 Signals and their book, Rework.
[00:02:09.000 --> 00:02:10.760] They have a chapter called Pick a Fight.
[00:02:10.760 --> 00:02:14.920] And one of the things that they talk about is intentionally going out of your way to have a rival.
[00:02:14.920 --> 00:02:16.920] And they have like one constraint for that.
[00:02:16.920 --> 00:02:18.600] They say, always punch up.
[00:02:18.600 --> 00:02:20.120] Everybody loves an underdog.
[00:02:20.120 --> 00:02:27.160] You never want to be the most popular person picking like somebody who is less, you know, far along to be your rival.
[00:02:27.160 --> 00:02:32.200] And in a way, like, I think like, like you said, like, I think Peter Levels has like over 100,000 followers on Twitter.
[00:02:32.200 --> 00:02:33.720] And you were at, what, like, 15?
[00:02:34.040 --> 00:02:37.240] And so, like, because the two of you are rivals, like, oh, he's at 270.
[00:02:37.240 --> 00:02:38.200] Geez, he's like blown it.
[00:02:38.280 --> 00:02:39.240] He's way up there.
[00:02:39.240 --> 00:02:41.160] He just kind of like brought you along with him, right?
[00:02:41.160 --> 00:02:43.640] Like, you are, like, in many people's eyes, probably the underdog.
[00:02:43.640 --> 00:02:47.320] And so, like, at whatever point you can, like, Channy, who's our rival?
[00:02:47.320 --> 00:02:48.040] We don't really have.
[00:02:48.520 --> 00:02:51.080] You could rival with Sam and his.
[00:02:51.400 --> 00:02:53.000] What is his company called again?
[00:02:53.000 --> 00:02:53.480] Who is this?
[00:02:53.480 --> 00:02:54.760] So he has like this offense.
[00:02:54.840 --> 00:02:56.840] Oh, he beat to pay a shit ton of money.
[00:02:56.840 --> 00:03:01.000] You guys can do it for like the underdogs that can eventually move into New Hampshire.
[00:03:01.880 --> 00:03:02.280] Yeah.
[00:03:02.280 --> 00:03:03.560] Sam Parr could be our rival.
[00:03:03.560 --> 00:03:07.720] Like, in a way, like, Y Combinator is like an aspirational rival.
[00:03:07.720 --> 00:03:12.600] They used to just do funding, but now they're doing all this inspirational founder content.
[00:03:12.920 --> 00:03:13.880] They're doing content, man.
[00:03:13.880 --> 00:03:14.920] They moved into our lane.
[00:03:14.920 --> 00:03:15.640] They're on YouTube.
[00:03:15.640 --> 00:03:16.440] They got their podcast.
[00:03:16.440 --> 00:03:17.080] They got their blog.
[00:03:17.080 --> 00:03:18.360] They're doing founder interviews.
[00:03:18.360 --> 00:03:20.520] And so it's like, that's punching up for us.
[00:03:20.520 --> 00:03:25.000] You know, if we're their rival, and if they tweet about us or say anything back to us, like, that helps us.
[00:03:25.000 --> 00:03:31.480] So, Danny, the reason you're rivals with Peter Level is because you're both doing these AI tools.
[00:03:31.480 --> 00:03:33.400] I don't know how to describe you as an eddy hacker.
[00:03:33.400 --> 00:03:35.720] You've been working on a lot of stuff for a long time.
[00:03:35.720 --> 00:03:38.200] You've had, I think, a seven-figure exit.
[00:03:38.520 --> 00:03:43.720] You have been working with AI in particular for three years now, I think.
[00:03:43.720 --> 00:03:45.000] I saw you tweeting about stuff, like back in 2010.
[00:03:46.560 --> 00:03:54.160] And now you've got like the suite of AI photo editing tools, for a lack of a better word.
[00:03:54.160 --> 00:03:59.520] Like the way that I would describe your portfolio of apps is that they all pick some vertical, right?
[00:03:59.520 --> 00:04:03.360] Photography is a probably $10, $100 billion industry.
[00:04:03.360 --> 00:04:07.840] And it looks like you've just gone down the list of what makes a lot of money and photography.
[00:04:07.840 --> 00:04:10.560] Headshots, okay, I'm doing an AI headshot thing.
[00:04:10.560 --> 00:04:14.320] Like hiring models, okay, I'm doing AI modeling agency.
[00:04:14.320 --> 00:04:15.200] You don't need to hire models.
[00:04:15.200 --> 00:04:18.480] We'll create a model for you in AI and put that person's face all over your website.
[00:04:18.480 --> 00:04:20.480] Team headshots got that unlocked.
[00:04:20.480 --> 00:04:24.800] Like you're just going down the list of what makes money and doing it with AI.
[00:04:24.800 --> 00:04:26.560] I think Peter is doing it all in one tool.
[00:04:26.560 --> 00:04:41.680] He's trying to make like the SaaS one agency model, and I'm trying to do all these different verticals, trying to do all the SEO research and like what, yeah, what you said is like headshots, team headshots, profile pictures, modeling, trying to figure out what is the next step on it.
[00:04:41.680 --> 00:04:49.280] Because I can just use the same tech and I just have to make a new landing page, do some SEO keywords for that, and then just drop into that photocol and just keep repeating it.
[00:04:49.280 --> 00:04:52.560] I'm actually planning on making one backend system so I don't have to rebuild that anymore.
[00:04:52.560 --> 00:04:56.160] So you can just like snipe into one lane two weeks.
[00:04:56.160 --> 00:04:59.760] And I think the great thing is I'm competing with the photographers.
[00:04:59.760 --> 00:05:01.920] And I mean, they don't know how to develop, right?
[00:05:02.240 --> 00:05:05.040] So you're also competing with other people like you and Peter.
[00:05:05.040 --> 00:05:10.800] Like I did kind of a Google search and it's not just like you and Peter, right?
[00:05:11.120 --> 00:05:15.360] If I look for like AI headshot apps, there's Headshot Pro, which is you.
[00:05:15.360 --> 00:05:16.640] There's Headshot AI.
[00:05:16.640 --> 00:05:18.720] There's HeadshotsBuyAI.com.
[00:05:18.720 --> 00:05:20.960] There's like 15 others.
[00:05:20.960 --> 00:05:27.120] So anyone who's actually searching for like, hey, I want headshots by AI, which is probably like 1% of the population.
[00:05:27.120 --> 00:05:29.760] Probably most people don't even think that AI could do this right now.
[00:05:30.120 --> 00:05:32.200] It's you versus a lot of people.
[00:05:32.200 --> 00:05:34.600] Yeah, there's a lot of competitors.
[00:05:35.480 --> 00:05:37.480] I'm doing quite well for some reason, I think.
[00:05:37.480 --> 00:05:39.240] Probably because I have a high conversion rate.
[00:05:39.400 --> 00:05:42.040] So I can outcompete them on Google Ads and stuff like that.
[00:05:42.360 --> 00:05:42.920] Yeah.
[00:05:42.920 --> 00:05:46.360] But I think I'm lucky here because I built my own tech, my own models.
[00:05:46.360 --> 00:05:54.680] So this is harder to copy because I've been spending like three, four weeks really coming up with my own AI, writing Python, training my own models.
[00:05:54.680 --> 00:05:59.640] So this is not just like I'm going to hook up to an API and then get the results out.
[00:05:59.640 --> 00:06:00.200] Yeah.
[00:06:00.200 --> 00:06:01.400] How well are you doing?
[00:06:01.400 --> 00:06:08.280] I know on Headshot Pro, I looked and you said that you've got like the numbers literally on your website.
[00:06:08.280 --> 00:06:10.680] And so you say like, okay, you've got 10,000.
[00:06:11.000 --> 00:06:12.680] You can reverse engineer it.
[00:06:13.400 --> 00:06:17.880] 10,423 happy customers paying $29 a person.
[00:06:17.880 --> 00:06:18.840] And that's where your rates start.
[00:06:18.840 --> 00:06:21.000] So some people are paying more than that.
[00:06:21.000 --> 00:06:29.480] For example, if you do headshots for your team, it's $40 per person for your whole team, which is a pretty good deal compared to like a traditional photographer.
[00:06:29.480 --> 00:06:32.840] Do you share your total revenue numbers or should I just do some math here?
[00:06:32.840 --> 00:06:42.040] Yeah, I quit doing it after I launched Headshot Pro because I felt, okay, A, it's going to get so many competitors, and B, I feel bad about sharing it right now.
[00:06:42.520 --> 00:06:45.640] These numbers are not that share-friendly anymore.
[00:06:45.640 --> 00:06:45.880] Right.
[00:06:45.960 --> 00:06:47.320] But you can do the math.
[00:06:47.320 --> 00:06:47.720] Yeah, sure.
[00:06:47.720 --> 00:06:48.360] It's public.
[00:06:48.360 --> 00:06:51.240] I feel like we're going to just do some arithmetic real quick here.
[00:06:51.240 --> 00:06:52.200] It's over $300,000.
[00:06:52.360 --> 00:06:53.800] No, no public math, though.
[00:06:53.800 --> 00:06:54.280] Yeah.
[00:06:54.520 --> 00:06:55.640] I got a calculator.
[00:06:55.880 --> 00:06:57.560] You've got other products too.
[00:06:57.560 --> 00:06:59.480] It's not just Headshot Pro.
[00:06:59.480 --> 00:07:03.000] So you're making hundreds of thousands of dollars in a very short period of time.
[00:07:03.000 --> 00:07:05.480] Like, these aren't, like, when did you launch Headshot Pro?
[00:07:05.480 --> 00:07:07.240] I think that's now five weeks ago.
[00:07:07.240 --> 00:07:07.920] Yeah, five weeks.
[00:07:07.800 --> 00:07:08.560] Provo Picture.
[00:07:08.680 --> 00:07:12.680] ProfoPicture was November, so that's three, five months.
[00:07:12.680 --> 00:07:15.840] I mean, AI, official AI didn't exist six months ago.
[00:07:15.840 --> 00:07:16.800] Yeah, it's crazy.
[00:07:14.840 --> 00:07:21.600] And you have this tweet that I think encapsulates kind of the opportunity here because sometimes people get like upset.
[00:07:21.840 --> 00:07:27.360] Like, one of the most popular tweets that I see on Twitter nowadays is, God, I've muted everything to do with AI.
[00:07:27.360 --> 00:07:28.400] Everything is just a clone.
[00:07:28.400 --> 00:07:29.280] Blah, blah, blah, blah, blah.
[00:07:29.440 --> 00:07:31.520] People are so tired of seeing AI stuff.
[00:07:31.520 --> 00:07:36.000] And then I look at you and you're like, Kill it, made $300,000 in the last five weeks.
[00:07:37.600 --> 00:07:40.560] I think I got lucky though, because I sold my company back in 2020, right?
[00:07:40.560 --> 00:07:43.600] And I got a non-compete on ChatGPT, OpenAI.
[00:07:43.600 --> 00:07:45.200] I cannot do that for three years.
[00:07:45.200 --> 00:07:51.520] So I'm like, wow, forced to not do any of the fancy, hyped stuff on Twitter that everyone's doing.
[00:07:51.520 --> 00:07:53.040] So I need to find something else to do.
[00:07:53.280 --> 00:07:53.760] Wait, wait, wait.
[00:07:53.920 --> 00:07:56.880] So what exactly is your non-compete?
[00:07:56.880 --> 00:07:58.080] Like, what are you not allowed to do?
[00:07:58.080 --> 00:08:02.800] I'm not allowed to do any text generation for three years, which ends next year.
[00:08:03.840 --> 00:08:04.560] Wow.
[00:08:05.120 --> 00:08:06.880] And this is because of your acquisition?
[00:08:06.880 --> 00:08:08.320] Yeah, by the company that you sold.
[00:08:08.320 --> 00:08:11.600] So I sold my copyright generator to Jasper back in the days.
[00:08:11.600 --> 00:08:16.800] And obviously, they don't want me to compete anymore because the founders said you ship too fast.
[00:08:16.800 --> 00:08:18.800] I don't want anything to do with you anymore.
[00:08:19.760 --> 00:08:20.400] Wow.
[00:08:20.400 --> 00:08:20.720] Okay.
[00:08:20.720 --> 00:08:24.640] So that severely handicaps you, but like you said, constraints, creativity.
[00:08:24.640 --> 00:08:24.960] Now you're going to be able to do it.
[00:08:25.120 --> 00:08:26.000] Dude, I feel like that doesn't.
[00:08:26.000 --> 00:08:32.480] Yeah, I feel like that doesn't even handicap him because what that does, if you have constraints, then you end up doing things that are new.
[00:08:32.480 --> 00:08:40.240] And in this particular case, we're in a situation where all you see today is basically chat GPT stuff.
[00:08:40.240 --> 00:08:41.600] You see a bunch of text generation.
[00:08:41.600 --> 00:08:45.760] All you saw five months ago was like image generation, right?
[00:08:46.080 --> 00:08:58.720] So because you're handcuffed from doing the thing that is like the hot thing that everyone is like constantly bored by, you have this like built-in forcing function to like, okay, what else can I do, right?
[00:08:58.720 --> 00:08:59.280] Yeah.
[00:08:59.000 --> 00:09:01.720] Yeah, yeah, and the other thing is like you move really fast, right?
[00:08:59.440 --> 00:09:04.920] That's why this like company that acquired your last product doesn't want to compete with you.
[00:08:59.600 --> 00:09:06.600] And guess what's also moving fast?
[00:08:59.680 --> 00:09:14.200] Like the state of AI, things that were cutting edge six months ago, nine months ago now look like crap and they're yesterday's news.
[00:09:14.200 --> 00:09:16.200] And if you're still doing that stuff, like you're behind.
[00:09:16.200 --> 00:09:17.880] So you have to constantly innovate.
[00:09:17.880 --> 00:09:27.080] And I think this is one of the reasons why people, especially sort of like the more stuffy, tried and true business theoretical crowd are like, oh, you know, you need to build a business that can last.
[00:09:27.080 --> 00:09:28.760] But like, do you?
[00:09:28.760 --> 00:09:34.920] Is there anything wrong with building a business that makes a bunch of money in a short period of time and then becomes obsolete?
[00:09:34.920 --> 00:09:36.600] But by that time, you're on to the next thing?
[00:09:36.600 --> 00:09:38.440] Like, I see no problem doing that.
[00:09:38.440 --> 00:09:45.000] And that might be the way the world is going considering the pace of technological advancement right now.
[00:09:45.000 --> 00:09:46.600] Like, look at Linktree, for example.
[00:09:46.600 --> 00:09:49.560] Like, they are trying to build a sustainable product, right?
[00:09:49.560 --> 00:09:55.560] And then last week, I think Instagram suddenly decided, hey, guys, you can add five links and suddenly your sustainable feature business is gone.
[00:09:55.560 --> 00:09:55.880] Yeah.
[00:09:55.880 --> 00:09:57.480] It's the same with TikTok.
[00:09:57.480 --> 00:09:59.960] TikTok just added the avatars that they were ProvoPicture.
[00:09:59.960 --> 00:10:03.640] So if I was building ProvoPicture as a sustainable business, it's still sustainable, right?
[00:10:03.640 --> 00:10:09.400] But if you build on a hype and if you build on a feature, like you want to get the cash out as soon as possible, right?
[00:10:09.400 --> 00:10:16.120] Because someone that ships slower than you, six months down a drain that has like a mass audience, they're just gonna, they're just gonna take your audience.
[00:10:16.360 --> 00:10:17.880] Incumbent sales are already down.
[00:10:17.880 --> 00:10:19.160] Like you can already notice it.
[00:10:19.160 --> 00:10:27.720] So all these huge companies employ tens of thousands of really smart programmers and they are getting better and better at building things quickly.
[00:10:27.720 --> 00:10:33.080] And like you said, if you don't release, you know, capture as much money as you can now, like, okay, maybe they move slower than you.
[00:10:33.080 --> 00:10:36.760] Maybe it takes them six months to get to where you are, but like, then they have the distribution advantage.
[00:10:36.760 --> 00:10:41.480] And it turns out that having a distribution advantage is worth more than six months.
[00:10:41.480 --> 00:10:43.160] You know, like they could way more.
[00:10:43.160 --> 00:10:44.360] It's worth so much.
[00:10:44.360 --> 00:10:46.000] And so you've got to build like super fast.
[00:10:46.000 --> 00:10:47.200] So I like what you're doing.
[00:10:47.200 --> 00:10:48.000] I mean, there's haters.
[00:10:44.760 --> 00:10:50.560] There's people who say, like, okay, these businesses aren't going to last.
[00:10:44.840 --> 00:10:51.840] They're just clones of each other.
[00:10:52.160 --> 00:10:54.080] I guess it depends on your personality type, right?
[00:10:54.080 --> 00:10:59.120] Like, I have this little ADD personality type where after three months, I'm absolutely bored with a product.
[00:10:59.200 --> 00:11:00.960] I don't want to work on it anymore.
[00:11:00.960 --> 00:11:02.720] Like, I don't even want to touch it.
[00:11:02.720 --> 00:11:04.240] So for me, it's perfect.
[00:11:04.480 --> 00:11:17.200] But for other people, if you like more slowly building, getting out, getting like a stable life, you have children, whatever, then probably you absolutely hate this because you can build any stability on top of it.
[00:11:17.200 --> 00:11:17.920] Right.
[00:11:17.920 --> 00:11:22.080] If you didn't bring up the ADHD thing, I was looking at your list of products here.
[00:11:22.080 --> 00:11:23.920] I'm like, I'm just scrolling through.
[00:11:23.920 --> 00:11:25.920] You've got like parody bar discontinued.
[00:11:25.920 --> 00:11:28.480] Headline acquired like rare blocks discontinued.
[00:11:28.480 --> 00:11:30.480] This one, 2021, this one, 2023.
[00:11:30.480 --> 00:11:37.360] And I'm like, this is a timeline of your products, but it's also like a timeline of your mental focus and like the shiny objects.
[00:11:37.360 --> 00:11:40.080] Yeah, postcrafts.com, 19 different products.
[00:11:40.080 --> 00:11:41.760] 19 startups by Danny Postcraft.
[00:11:41.840 --> 00:11:42.560] Oh, it's 19.
[00:11:44.400 --> 00:11:45.440] You're not just shipping products.
[00:11:45.440 --> 00:11:47.120] You're shipping products and you're tweeting about it.
[00:11:47.120 --> 00:11:48.240] You're talking about your revenue.
[00:11:48.240 --> 00:11:49.440] You're talking about your decisions.
[00:11:49.440 --> 00:11:51.920] You're talking about the disputes you're dealing with with customers.
[00:11:51.920 --> 00:12:01.680] Like you're doing all the juicy behind-the-scenes stuff that makes people want to follow along because it's just far more interesting to follow somebody like you than somebody who doesn't ever tweet about it.
[00:12:01.840 --> 00:12:13.840] For me, for me, it's also a diary because I read back tweets I did a year ago if I needed to solve something because I remember I had the same issue at that time and I would be reading it back and then like I'm kind of also doing it for myself just so it has like cross-reference thing.
[00:12:13.840 --> 00:12:17.200] I like this sort of like 12 startups in 12 months approach.
[00:12:17.200 --> 00:12:21.920] Also, something that like Peter came up with a long time ago, where you basically say, Screw trying to work on one thing.
[00:12:21.920 --> 00:12:24.640] I'm going to do a ton of things in the course of a year.
[00:12:24.640 --> 00:12:27.440] I'm going to learn a lot and I'm going to see what sticks.
[00:12:27.440 --> 00:12:31.880] You said you built like you know a dozen products or tons of products in 2020.
[00:12:31.880 --> 00:12:34.200] What were they and what stuck?
[00:12:34.520 --> 00:12:38.600] So the first one was Landic Folio, which I did for four years, which is the inspiration gallery.
[00:12:38.600 --> 00:12:48.600] After that, I took a little feature from that website and turned it into a drag and drop inspiration builder so you could make a landing page based on examples of other websites.
[00:12:48.600 --> 00:12:49.400] So that was another one.
[00:12:49.400 --> 00:12:54.360] So I like to keep, like, take a little feature of another product and iterate on top of it.
[00:12:54.360 --> 00:12:56.760] How much money did these products make?
[00:12:57.080 --> 00:12:59.080] 2K tops, I think, per month.
[00:12:59.080 --> 00:13:02.760] In SpireFrame, maybe 400 at the time?
[00:13:02.760 --> 00:13:03.560] Nothing like that.
[00:13:03.720 --> 00:13:05.960] Enough to survive if you're living in Bali.
[00:13:05.960 --> 00:13:07.960] But not what you're looking for.
[00:13:07.960 --> 00:13:09.720] Yeah, for many, many years.
[00:13:09.720 --> 00:13:10.360] Right.
[00:13:10.360 --> 00:13:11.400] For a long time.
[00:13:11.400 --> 00:13:11.720] Okay.
[00:13:11.720 --> 00:13:11.880] Yeah.
[00:13:11.880 --> 00:13:14.280] And then I got lucky with the headline.
[00:13:14.680 --> 00:13:14.840] Yeah.
[00:13:15.160 --> 00:13:18.280] That's like your fifth product on your list.
[00:13:18.280 --> 00:13:19.560] This is the one that got acquired.
[00:13:19.560 --> 00:13:24.440] You started in 2020, and it was an AI-based copywriting thing.
[00:13:24.440 --> 00:13:28.120] The headline on the website now says, Writing copy has never been easier.
[00:13:28.120 --> 00:13:33.720] The automated copywriter for busy designers, anti-hackers, marketing agencies, business owners.
[00:13:33.720 --> 00:13:36.360] How did you come up with an idea that you were able to sell?
[00:13:36.680 --> 00:13:40.120] Also, let me point out something that I noticed before you even say that.
[00:13:40.120 --> 00:13:42.520] Like, there were four products that happened before that.
[00:13:42.520 --> 00:13:47.640] You know, you had the landing page design, you have this pop-up library for JavaScript, you have mock-ups.
[00:13:47.640 --> 00:13:52.280] So you have all of these basically like these products that have to do with just basically sort of helping people code.
[00:13:52.280 --> 00:13:54.600] And then, boom, you just have AI.
[00:13:54.600 --> 00:13:56.200] And then that's the one that got acquired.
[00:13:56.200 --> 00:14:01.800] So, and there is a really nice bridge going from that programming to the actual AI content.
[00:14:01.800 --> 00:14:04.920] So on Linux Folio, I have marketers coming there.
[00:14:04.920 --> 00:14:09.880] So we wrote an e-book with headline templates because that's the hardest way to come up for the Linux page.
[00:14:09.880 --> 00:14:19.120] So we sold an e-book with copies, like get blank in blank hours, and then people could fill in the variables in the e-book.
[00:14:14.680 --> 00:14:21.600] And then I was thinking, okay, I just learned the program.
[00:14:21.760 --> 00:14:24.400] I was like, you can sell an e-book for $19, $19, right?
[00:14:24.480 --> 00:14:25.200] Like, not much.
[00:14:25.200 --> 00:14:28.560] But if you sell a SaaS tool, you can suddenly increase the price over that.
[00:14:28.560 --> 00:14:34.960] So I was like, maybe if I could tell, like, if I put those templates into a SaaS tool and you can sell it for $89.
[00:14:34.960 --> 00:14:37.520] So that one exploded when I sold it.
[00:14:37.520 --> 00:14:39.920] I got like $60K revenue, I think, in two weeks.
[00:14:39.920 --> 00:14:40.800] That was my first experience.
[00:14:40.960 --> 00:14:41.200] Wow.
[00:14:41.600 --> 00:14:43.440] And that was Pushfolio, right?
[00:14:43.440 --> 00:14:47.440] No, that was a headline actually before they went.
[00:14:48.400 --> 00:14:49.280] Wow.
[00:14:49.280 --> 00:14:50.720] Headline AI.
[00:14:51.120 --> 00:14:52.400] Free AI.
[00:14:52.400 --> 00:14:57.040] So I was working on that product, and then a friend of mine showed me GPT-3.
[00:14:57.040 --> 00:15:01.840] And I was like, all right, I'm going to TM, I'm going to email Greg, the CTO of OpenAI.
[00:15:01.840 --> 00:15:02.480] Can I get access?
[00:15:02.480 --> 00:15:03.200] So I got access.
[00:15:03.200 --> 00:15:05.200] It was like one of the first batch.
[00:15:05.200 --> 00:15:10.480] And I let it write the templates for my SaaS for headline with AI.
[00:15:10.560 --> 00:15:11.200] I was like, wait.
[00:15:11.200 --> 00:15:12.240] Oh, that's awesome.
[00:15:12.240 --> 00:15:14.960] Wait, this fucking AI can write copy for me.
[00:15:14.960 --> 00:15:19.760] What if I turn this tool into a copywriting tool instead of just using templates?
[00:15:19.760 --> 00:15:20.320] Right.
[00:15:20.320 --> 00:15:27.680] So I was lucky to be working on a copywriting tool by accident when a copywriting AI came out.
[00:15:27.680 --> 00:15:28.160] Wow.
[00:15:28.160 --> 00:15:36.320] I think I have a screenshot that I'm one of the first people to go live with GPT-3 actually in production together with copy.ai and three tools at the same time.
[00:15:36.400 --> 00:15:39.840] Copy.ai is valued at whatever hundreds of millions of dollars now.
[00:15:39.840 --> 00:15:49.200] I interviewed somebody back in the day who was doing AI copywriting, and I've never gotten so many negative responses to an interview that I've done.
[00:15:49.200 --> 00:15:50.640] People hated that guy.
[00:15:50.640 --> 00:15:52.160] They were so upset.
[00:15:52.160 --> 00:15:53.920] People don't like low-quality content.
[00:15:53.920 --> 00:15:55.600] They don't like spam content.
[00:15:55.600 --> 00:16:11.320] And I think the idea, and this was back in 2020, the idea that we might be heading toward a world where the internet is just gunked up with nothing but low-quality AI spam content and people are creating this at the click of a button just like enraged a lot of people who are listening.
[00:16:11.320 --> 00:16:24.680] The full circle on that is that the first AI feature that we've built for indie hackers was your like AI spam, you know, block or spam moderator that is like going in and zapping members of its own space.
[00:16:24.920 --> 00:16:27.880] People who are getting like AI spam on the indie hackers forum.
[00:16:29.560 --> 00:16:33.080] So this guy, like, so like the fears weren't exactly completely unfounded.
[00:16:33.080 --> 00:16:34.360] I just want to say that.
[00:16:34.360 --> 00:16:34.920] Yeah.
[00:16:34.920 --> 00:16:37.560] Well, back in 2020, this was like, it was a very different landscape.
[00:16:37.560 --> 00:16:38.760] Nobody knew this was possible.
[00:16:38.760 --> 00:16:44.840] And Danny, you said, you just happened to be doing a copywriting tool at the same time that AI came out that could write copy.
[00:16:44.840 --> 00:16:46.360] How did you get it into the hands of people?
[00:16:46.360 --> 00:16:50.840] Because like, I mean, I imagine at this time you had like five followers on Twitter.
[00:16:50.840 --> 00:16:52.280] Nobody knew who you were.
[00:16:52.280 --> 00:16:54.040] You know, you didn't have a giant email list.
[00:16:54.040 --> 00:16:56.680] Like, how did anybody even find out about Headline?
[00:16:56.680 --> 00:17:01.480] I think I got lucky that same with Hetshop Pro and ProvoPicture.
[00:17:01.480 --> 00:17:03.800] Like, it is so novelty.
[00:17:03.800 --> 00:17:06.600] Like, it's so new that it's magic to people.
[00:17:06.600 --> 00:17:10.440] So I was just tweeting out tweets, like, using the tool, and it was new to people.
[00:17:10.440 --> 00:17:12.840] I had 200 followers back in the day, I think.
[00:17:12.840 --> 00:17:14.440] And people just started retweeting it.
[00:17:14.440 --> 00:17:15.480] People went mentor over there.
[00:17:15.480 --> 00:17:16.920] We're like, what the fuck's going on?
[00:17:17.320 --> 00:17:18.840] This tool is writing copy.
[00:17:18.840 --> 00:17:19.800] Like, what?
[00:17:20.120 --> 00:17:21.720] Because we live in a bubble, right?
[00:17:21.720 --> 00:17:23.560] There's like 0.1% bubble.
[00:17:23.560 --> 00:17:26.120] But once you go outside a little bit of it, it's magic to people.
[00:17:26.120 --> 00:17:27.400] It was magic to me.
[00:17:27.400 --> 00:17:28.840] I was like so amazed by it.
[00:17:29.640 --> 00:17:38.280] I grew from, I think, 200 followers in that say to 10K before I sold Headline, I guess.
[00:17:38.280 --> 00:17:41.160] And that was the only marketing I did, only on Twitter.
[00:17:41.160 --> 00:17:42.440] Just word of mouth.
[00:17:42.680 --> 00:17:47.600] That's how I grew to 25K MRR before I sold it.
[00:17:44.840 --> 00:17:49.600] Yeah, it was only word of mouth.
[00:17:50.000 --> 00:17:53.040] No SEO, no marketing, just Twitter.
[00:17:53.040 --> 00:17:53.760] Just Twitter.
[00:17:53.760 --> 00:18:01.760] But I feel like Twitter is like, if you are an indie hacker, if you're sort of a small startup founder, that's where we hang out.
[00:18:02.160 --> 00:18:03.920] That's like your people, our people.
[00:18:03.920 --> 00:18:05.840] But I want to say one thing about you.
[00:18:05.920 --> 00:18:08.240] So you say indie hackers are the right way to hang out, Jenny.
[00:18:10.400 --> 00:18:17.280] But what I was going to say is like, dude, it's so easy to sort of see the narrative as you were doing the right kind of product at the right time.
[00:18:17.280 --> 00:18:19.360] And the wind was in the air.
[00:18:19.360 --> 00:18:24.000] And it just so happened to be that the AI stuff became an option.
[00:18:24.000 --> 00:18:34.160] But dude, you did something in that story that 999 out of 1,000 people would not do, which is you heard about it and you were like, oh, this thing is sick.
[00:18:34.160 --> 00:18:39.200] I'm going to just reach out to the founder and ask for favors.
[00:18:41.760 --> 00:18:43.600] That's not a minor play that you made, right?
[00:18:43.600 --> 00:18:46.080] Number one, you noticed that this thing was awesome.
[00:18:46.400 --> 00:18:50.320] And then the second thing is like, you had the balls or whatever, right?
[00:18:50.640 --> 00:18:53.200] I don't like, you know, most people don't even think of that as an option.
[00:18:53.200 --> 00:18:56.320] Like, I'm just going to reach out to the guy who made this at the very top.
[00:18:56.320 --> 00:19:00.000] And then, but for that decision, none of this would have happened.
[00:19:00.000 --> 00:19:00.400] Yep.
[00:19:00.400 --> 00:19:00.800] Yeah.
[00:19:00.800 --> 00:19:07.040] It's crazy how many little small life decisions escalate into like something big in your life, to be honest.
[00:19:07.520 --> 00:19:14.640] There's also like this piece of advice that's super common, which is like, don't build a solution in search of a problem.
[00:19:14.640 --> 00:19:19.520] This is like foundational startup wisdom that has been around for like, it's been said for like 30 years.
[00:19:19.520 --> 00:19:33.480] Like, it's the most stereotypical thing in the world to be just a tech nerd or someone who loves, like, somebody with ADHD who just loves spritting about, and a new opportunity comes along, and you're like, oh, how do I take this, you know, crypto, AI, whatever thing, and just like slap it onto something that exists.
[00:19:33.480 --> 00:19:35.080] And most people say, don't do that.
[00:19:35.080 --> 00:19:38.760] But there are times when that is like absolutely the right thing to do.
[00:19:38.760 --> 00:19:49.560] When the new technology, I think, is so revolutionary that like when it comes out, if you just apply it to existing problems, it makes them 10 times better than the previous solutions.
[00:19:49.560 --> 00:19:51.080] And like, that's exactly what you did.
[00:19:51.080 --> 00:19:56.120] So it makes me kind of skeptical of some of this advice that's like tried and true advice when there's such clear exceptions.
[00:19:56.520 --> 00:19:58.600] Same thing with the birth of the internet, right?
[00:19:58.600 --> 00:20:03.240] Anyone who didn't slap the internet onto their newspaper in the 90s, their newspaper is dead.
[00:20:03.240 --> 00:20:04.280] They did not make it.
[00:20:04.360 --> 00:20:15.560] So sometimes you just like, I kind of feel like we're in that moment right now with AI where it's like, hey, this is like so foundationally different and better than hiring a bunch of people to do things manually.
[00:20:15.560 --> 00:20:18.040] If like with this keystroke, you could do it automatically.
[00:20:18.040 --> 00:20:20.760] Like it's easy to be a curmudgeon and be like, no, it's just a trend.
[00:20:20.760 --> 00:20:21.320] It's just a trend.
[00:20:21.320 --> 00:20:21.960] It's just a fad.
[00:20:22.600 --> 00:20:23.720] It's not just fad.
[00:20:23.720 --> 00:20:26.440] Like, it's seriously good.
[00:20:26.440 --> 00:20:42.440] The crazy thing I realize, and I haven't been able to put it in words, but AI can, like, you can go into industries that normally take so much human labor, and you can basically just, as a one-person business, and this is why I think this is a golden time for indie hackers.
[00:20:42.440 --> 00:20:44.280] Like the Photo Studio, for example.
[00:20:44.280 --> 00:20:47.800] I'm competing with people who probably have, and it's also a negative thing, right?
[00:20:47.800 --> 00:20:52.840] But it's positive for the consumer because no one wants to take headshots or make $500 worth of.
[00:20:53.160 --> 00:20:55.880] So I see it as like, I do it for the customers.
[00:20:55.880 --> 00:21:08.200] I can basically by myself scale unlimited times if I have this GPU power and just do photography by myself because I have an AI that's just doing the work for me.
[00:21:08.200 --> 00:21:11.640] I have all these job works, all these other AIs filtering, blah, blah, blah, blah, blah.
[00:21:11.800 --> 00:21:16.720] The only thing I need to manually do is, yeah, do the programming and the customer support.
[00:21:14.920 --> 00:21:21.280] And thank God for my wife who helps me out with that because that's a lot of work.
[00:21:21.600 --> 00:21:30.560] And if you apply AI to solving real problems that are done by real humans, I think there's a lot of markets you can get into in that way.
[00:21:30.560 --> 00:21:31.040] Yeah.
[00:21:31.040 --> 00:21:32.720] Shanning hinted at this a second ago.
[00:21:32.720 --> 00:21:36.800] So we've had like a spam problem on indie actors forever.
[00:21:36.800 --> 00:21:41.360] And a lot of the people who are spammers, like it's not just robots, it's actual people sometimes.
[00:21:41.360 --> 00:21:44.560] You come on and I'll implement spam fighting stuff.
[00:21:44.560 --> 00:21:51.520] You know, add a captcha or I'll change like all the classes in my CSS so like the robots can't find it and people just figure it out and they just do it differently.
[00:21:51.520 --> 00:21:52.880] I'll ban certain words.
[00:21:52.880 --> 00:21:54.080] They'll just start using different words.
[00:21:54.080 --> 00:21:56.800] If I ban links, they start putting a space in front of the dot-com.
[00:21:56.800 --> 00:21:57.680] They try everything, right?
[00:21:57.680 --> 00:21:59.840] It's like real people fighting.
[00:21:59.840 --> 00:22:03.840] And two or three weeks ago, I was like, okay, should we hire another community moderator?
[00:22:03.840 --> 00:22:04.880] It's expensive.
[00:22:04.880 --> 00:22:06.960] It's hard to manage the community moderator.
[00:22:06.960 --> 00:22:09.840] You don't really necessarily want to outsource all the community work to somebody else.
[00:22:09.840 --> 00:22:12.080] So we should really be doing a lot of this ourselves.
[00:22:12.560 --> 00:22:14.720] But, you know, mostly it's just expensive.
[00:22:14.720 --> 00:22:15.600] We're independent now.
[00:22:15.600 --> 00:22:16.720] We can't afford that kind of thing.
[00:22:17.120 --> 00:22:18.240] Congratulations, by the way.
[00:22:18.240 --> 00:22:18.640] Awesome.
[00:22:18.720 --> 00:22:18.960] Thanks.
[00:22:18.960 --> 00:22:20.480] It's nice to be nice to be indie.
[00:22:20.480 --> 00:22:22.000] So I'm like, okay, what can I do?
[00:22:22.000 --> 00:22:30.000] Well, I could just take a solution that exists, AI, and just like slap it onto this problem and say, okay, well, what can like, what can GPT-4 do?
[00:22:30.000 --> 00:22:34.480] And so in the span of like a couple days, I just played around GPT-4 and it wasn't really even a couple days.
[00:22:34.480 --> 00:22:35.760] It was like a few hours.
[00:22:35.760 --> 00:22:40.080] And now I have a spam fighting bot who is just like dispensing justice on the forum.
[00:22:40.080 --> 00:22:41.360] I call him botman.
[00:22:41.360 --> 00:22:43.360] And he will message me and Telegram.
[00:22:43.360 --> 00:22:47.040] Every time he sees a spammer, he will give them a spam score from zero to one.
[00:22:47.040 --> 00:22:48.400] He'll give me a confidence score.
[00:22:48.400 --> 00:22:50.560] How confident is he from zero to one?
[00:22:50.560 --> 00:22:53.600] He will give me a reason and actual pros.
[00:22:53.600 --> 00:22:58.560] They'll tell me, oh, this person, their account is really new and it seems like this, blah, blah, blah.
[00:22:58.560 --> 00:23:00.920] And I think this is definitely a spammer.
[00:23:00.920 --> 00:23:03.560] And I just set him today to auto-ban people.
[00:22:59.840 --> 00:23:06.440] And dude, I think it's all, I think it's like all hits and no misses.
[00:23:06.680 --> 00:23:08.280] No, I've just been checking over two weeks.
[00:23:08.680 --> 00:23:09.960] He has never, it is never wrong.
[00:23:09.960 --> 00:23:11.160] It is right every single time.
[00:23:11.160 --> 00:23:18.840] Every time it finds its teacher, you just train like a really long, long text in ChatGPT where you put like examples inside and then it learns from it.
[00:23:18.840 --> 00:23:21.880] Or are you using like a vector database and it finds like the distance?
[00:23:22.040 --> 00:23:25.320] I'm doing the simplest, dumbest few shot learning in GPT-4.
[00:23:25.640 --> 00:23:27.560] I've given it like 10 examples.
[00:23:27.560 --> 00:23:30.360] And GPT-4's context is like, how many tokens is it?
[00:23:30.360 --> 00:23:35.560] It's like 12,000 or some crazy amount of tokens, which channeling a token is basically like three or four letters.
[00:23:35.560 --> 00:23:39.000] And so it can handle a gigantic wall of text of examples.
[00:23:39.000 --> 00:23:41.400] Plus, it's already been trained on basically the entire internet.
[00:23:41.400 --> 00:23:42.760] So it knows what spam is.
[00:23:42.760 --> 00:23:45.160] It knows like what moderating a forum is.
[00:23:45.160 --> 00:23:49.560] And it's like, okay, well, that just like wiped out the job of like policing spam on the forum.
[00:23:49.800 --> 00:23:54.680] I not only don't have to hire anyone, but I also don't need to go like pay some other business that's built a cool.
[00:23:54.680 --> 00:23:56.920] Like people have pitched me tools for this for years.
[00:23:56.920 --> 00:23:59.160] Hey, we've got some spam planning algorithms we train.
[00:23:59.160 --> 00:24:02.840] Sorry, GPT-4 is better than everything you've been working on for years.
[00:24:02.840 --> 00:24:04.760] And I just rigged it up in like an afternoon.
[00:24:04.760 --> 00:24:05.720] Like, I'm going to do this.
[00:24:05.720 --> 00:24:06.680] It's basically free.
[00:24:06.680 --> 00:24:07.320] Goodbye.
[00:24:07.320 --> 00:24:07.640] Right.
[00:24:07.640 --> 00:24:08.520] And it's like entertaining.
[00:24:08.520 --> 00:24:09.400] I can tell it to be funny.
[00:24:09.400 --> 00:24:10.520] I could tell it to be humorous.
[00:24:10.520 --> 00:24:13.560] It'll send me interesting, funny messages and then ban the spammers.
[00:24:13.560 --> 00:24:20.520] And so it's like, I don't see how people can look at this technology and say, like, this is a passing fad, just like crypto or something.
[00:24:21.480 --> 00:24:23.080] It's not like that's like head in the sand.
[00:24:23.240 --> 00:24:25.320] There's so much more use case to it.
[00:24:25.320 --> 00:24:26.760] Like, this is not a fad.
[00:24:26.760 --> 00:24:31.960] This is completely industry changing, I think, even for the better and for the worse, right?
[00:24:31.960 --> 00:24:32.920] Like, yeah.
[00:24:33.240 --> 00:24:43.640] To me, I think that the sign of it is like, I said this a few weeks ago, where it's like the difference between, you know, you feeling like there might be an opportunity, like there's new technology, and you're like, oh, there's an opportunity.
[00:24:43.640 --> 00:24:47.200] Maybe we should do some like RD, and like maybe there's something there.
[00:24:44.520 --> 00:24:49.120] But this is not, it doesn't feel like an opportunity.
[00:24:49.360 --> 00:25:04.640] We got that spam fighting bot going, and now I'm almost like, dude, it's a liability that like there's almost certainly other parts of our stack, whether it's our technology stack or like our process stack, like how we do things, where we're currently wasting tons of time.
[00:25:04.640 --> 00:25:11.280] And like, I'm almost like, I want to like, you know, sort of do this like an audit, like an internal audit.
[00:25:11.280 --> 00:25:12.240] Like, what are we doing?
[00:25:12.240 --> 00:25:14.960] What are we doing manually that we should be using AI for?
[00:25:14.960 --> 00:25:18.800] Not just like as a distraction to mess around, but it's like a real time-saving technique.
[00:25:18.800 --> 00:25:20.800] Like, we can't afford not to do that.
[00:25:20.800 --> 00:25:23.920] And the amazing thing is, like, you don't have the code anymore.
[00:25:23.920 --> 00:25:30.800] Like, what OpenAI did basically is they turned the written English word or whatever word into a programming language.
[00:25:30.800 --> 00:25:31.840] So anyone can use it.
[00:25:32.000 --> 00:25:35.920] Like, some people listening to this podcast are probably like, yeah, you know, these guys are all programmers.
[00:25:35.920 --> 00:25:36.560] Only they can do it.
[00:25:36.560 --> 00:25:41.920] Like, no, man, you can just send text to it, tell it what to do, and it sends you text back.
[00:25:41.920 --> 00:25:47.920] And if you tell it to send it back as like a one or a zero or a true or a false, like this is spam, yes or no?
[00:25:47.920 --> 00:25:48.880] And you can just read it out.
[00:25:48.880 --> 00:25:50.160] Like, you don't need to know program.
[00:25:50.320 --> 00:25:52.800] So I think people don't like fully comprehend what you're saying.
[00:25:52.800 --> 00:25:56.080] Because this is something that I was telling my friend about when I was coding my spam bot.
[00:25:56.160 --> 00:25:57.120] Like, oh, you're coding this thing.
[00:25:57.120 --> 00:25:59.680] I'm like, I'm literally writing text.
[00:25:59.680 --> 00:26:00.400] That's what I'm saying.
[00:26:00.400 --> 00:26:02.560] Like, there's some code involved, right?
[00:26:02.560 --> 00:26:07.600] Like, if I tried to fight spam a year ago or a month ago, I would be doing nothing but coding.
[00:26:07.600 --> 00:26:11.360] Now I wrote a tiny amount of code that basically just says, hey, ban this person.
[00:26:11.360 --> 00:26:18.080] But the vast majority of what I'm doing is tweaking a single sentence that says, You are a forum moderator for the anti-hackers community.
[00:26:18.080 --> 00:26:22.800] You're reviewing comments made by users on the forum and assigning a spam score ranging from zero to one.
[00:26:22.800 --> 00:26:24.400] Zero means it's unlikely to be spam.
[00:26:24.400 --> 00:26:25.920] One means it's very likely to be spam.
[00:26:25.920 --> 00:26:26.480] Blah, blah, blah.
[00:26:26.640 --> 00:26:30.920] I just keep writing all this stuff, and I end up with like a couple paragraphs of instructions and then it does it.
[00:26:29.120 --> 00:26:35.880] So like coding is literally me just giving instructions as if I was talking to a human.
[00:26:36.200 --> 00:26:39.080] It's like the highest level of abstraction of no code.
[00:26:39.080 --> 00:26:45.240] It's like no code with just like, okay, the interface for coding here, the syntax is, do you know English?
[00:26:45.240 --> 00:26:47.560] Okay, so just like use English and then talk to it.
[00:26:47.560 --> 00:26:48.520] Like that's the syntax for this.
[00:26:48.760 --> 00:26:50.680] So does your computer, it turns out.
[00:26:50.920 --> 00:26:53.880] If you think about it, like programming is like, it's such a bug.
[00:26:53.880 --> 00:26:59.000] Like why would we have to talk in a way that a computer understands if we made a computer?
[00:26:59.000 --> 00:27:02.520] Like we should just be able to write in English to a computer.
[00:27:02.520 --> 00:27:03.080] Yeah.
[00:27:03.080 --> 00:27:04.520] And that's getting possible now.
[00:27:04.520 --> 00:27:06.120] So everyone is becoming a programmer.
[00:27:06.120 --> 00:27:07.320] I guess people hate that.
[00:27:08.440 --> 00:27:10.360] And by the way, I've seen this in practice.
[00:27:10.360 --> 00:27:16.440] So Cortland, I told Dave, a buddy of ours that works at Lyft, I share all this stuff with him.
[00:27:16.440 --> 00:27:18.760] I share all the new tech stuff.
[00:27:19.000 --> 00:27:23.560] Maybe he's so interested because he works at Lyft and jobs are getting sliced left and right at these kind of companies.
[00:27:23.560 --> 00:27:25.800] So he's looking for his escape hatch.
[00:27:25.800 --> 00:27:31.320] But the first thing that he said when I showed him our bot was he was like, he said two things.
[00:27:31.320 --> 00:27:33.640] He goes, number one, that's awesome.
[00:27:33.640 --> 00:27:37.160] And then number two, bummer, like it's pointless for me to know about this.
[00:27:37.160 --> 00:27:37.960] I'm not a developer.
[00:27:37.960 --> 00:27:41.720] So I guess I'll just wait three years until somehow I can do something like this.
[00:27:41.960 --> 00:27:45.880] And I'm like, dude, no, like, you know, there are some limitations.
[00:27:45.880 --> 00:27:56.040] You have to be a little bit technical for some aspects of this, but like, you really should look into tools and ways that you can implement these kinds of things with your own workflows.
[00:27:56.040 --> 00:28:01.240] And I don't know if you listen to the all-in podcast, but it's got a few investors on it.
[00:28:01.240 --> 00:28:02.920] And one of them is Jason Calicanis.
[00:28:02.920 --> 00:28:19.200] And like the next day, there was an episode where Jason Calicanis, who's not really a developer, he was a podcast host, was like, oh, what I do these days is I have my, like, whenever I open a new tab in my browser, it is set to go to Auto GPT.
[00:28:14.760 --> 00:28:25.200] Like, I'm sort of creating this forcing function where I automatically just have to sort of go through Auto GPT.
[00:28:25.200 --> 00:28:32.720] So, no matter what I'm doing, I'm constantly like reminding myself, like, oh, this thing, could that be done better with Auto GPT?
[00:28:32.720 --> 00:28:34.560] Could this be done better with Auto GPT?
[00:28:34.560 --> 00:28:48.240] And it just sort of, he's like, I'm just training myself because no matter what, the thing I just said, like, that idea that like 100%, there are liabilities in the way that you're doing whatever it is that you're doing, like, some AI tool or feature could do it better.
[00:28:48.240 --> 00:28:50.560] There's all sorts of apps that are basically like glue.
[00:28:50.560 --> 00:28:54.880] So, the no-code community has been really into apps like Zapier for years and years and years and years.
[00:28:54.880 --> 00:29:01.840] And it's like, you can't code, but you need, you know, your Gmail to talk to your Telegram, to talk to Slack, to talk to, you know, Google Docs or something.
[00:29:01.840 --> 00:29:06.080] Okay, whenever I get a new email that is this subject, you know, send me this message and Telegram.
[00:29:06.080 --> 00:29:09.280] You could hook that up without code in like five seconds with Zapier, right?
[00:29:09.280 --> 00:29:12.400] And now all these apps are adding chat GPT and AI.
[00:29:12.400 --> 00:29:15.360] So, you know, if I type this query in, have it, you know, go here.
[00:29:15.360 --> 00:29:18.320] So you really, really don't need to learn how to code.
[00:29:18.320 --> 00:29:23.200] Like, it's, if you know the right apps, you could just sign up and start like hooking stuff up.
[00:29:23.200 --> 00:29:30.640] Everyone should spend like, probably take a few days off and just try to automate some of their most repetitive things in their life.
[00:29:30.640 --> 00:29:32.640] And they will just save so much more time.
[00:29:33.360 --> 00:29:35.680] I want to talk about, like, I wanted so much I want to talk to you about.
[00:29:35.680 --> 00:29:37.440] Like, what are you automating in your life?
[00:29:37.440 --> 00:29:42.800] But, like, first, I want to finish a story about Headline because you started it, it blew up, it's making money.
[00:29:43.120 --> 00:29:44.160] Just like, let's get a story out of it.
[00:29:44.800 --> 00:29:46.400] We did a really big detour there.
[00:29:46.480 --> 00:29:47.760] Huge detour.
[00:29:48.400 --> 00:29:49.280] How did you sell it?
[00:29:49.280 --> 00:29:50.880] And how much money did you make from selling it?
[00:29:50.880 --> 00:29:51.280] Because it's not really.
[00:29:51.520 --> 00:29:52.160] Twitter, dude.
[00:29:52.240 --> 00:29:53.680] This is all Twitter again.
[00:29:54.000 --> 00:29:59.360] I get reached out by sort of guys of jasper.ai and unbounce.
[00:29:59.360 --> 00:30:02.280] Unbounce.com, the landing page builder.
[00:30:02.520 --> 00:30:04.120] They both reached out via Twitter.
[00:29:59.920 --> 00:30:07.000] Dave from Jasper was just like super upfront.
[00:30:07.240 --> 00:30:10.200] Like, you just said, like, yeah, I want to buy you guys.
[00:30:10.200 --> 00:30:11.880] And they were a team of three people.
[00:30:11.880 --> 00:30:14.120] So I was hopping on calls with them to talk about it.
[00:30:14.120 --> 00:30:15.000] They just dropped the number.
[00:30:15.000 --> 00:30:16.040] I was like, double it.
[00:30:16.040 --> 00:30:17.080] And they were like, okay.
[00:30:17.080 --> 00:30:18.360] So we just started negotiating.
[00:30:18.360 --> 00:30:20.040] This was like super fast.
[00:30:20.040 --> 00:30:24.520] And I was getting burned out because I was doing customer support, development marketing, everything myself.
[00:30:24.520 --> 00:30:25.960] Like, I was done.
[00:30:25.960 --> 00:30:27.080] I was talking with my wife about it.
[00:30:27.160 --> 00:30:28.920] She was like, just sell it, dude.
[00:30:29.480 --> 00:30:34.600] So I was also entertaining offers from Unbounce, but they were like this obviously massive corporation.
[00:30:34.600 --> 00:30:36.120] So you have to go through procurement.
[00:30:36.120 --> 00:30:39.640] They have to wait two weeks to set up a meeting with the whole board.
[00:30:40.600 --> 00:30:43.560] And I was like, yeah, I have like an offer here.
[00:30:43.560 --> 00:30:44.760] I should just go through it.
[00:30:45.240 --> 00:30:46.680] I told them I don't want to work for them.
[00:30:46.680 --> 00:30:51.160] They were like, after two weeks, if you do the program, like the switch over, you're done with us.
[00:30:51.480 --> 00:30:51.880] Cool.
[00:30:51.880 --> 00:30:54.120] They asked me for Echo T, I was like, nah, I don't need it.
[00:30:54.120 --> 00:30:55.880] It's fucking stupid.
[00:30:57.160 --> 00:30:57.880] Oh, man.
[00:30:57.880 --> 00:31:02.440] Yeah, and then, and, yeah, I think it went through in four weeks.
[00:31:02.440 --> 00:31:07.880] I got 33% of the money in the bank, 66% over 12 months.
[00:31:08.280 --> 00:31:10.040] Two weeks in, I was done.
[00:31:10.040 --> 00:31:14.200] I was getting depressed afterwards because I went from working 16 hours a day to nothing.
[00:31:14.200 --> 00:31:15.480] I had nothing anymore.
[00:31:15.480 --> 00:31:15.880] Right.
[00:31:15.880 --> 00:31:16.280] Yeah.
[00:31:16.760 --> 00:31:18.520] It was odd.
[00:31:18.520 --> 00:31:21.800] A lot of people think they're going to quit and retire and go live on a beach.
[00:31:22.120 --> 00:31:24.120] But it's just like this grass is greener thing.
[00:31:24.360 --> 00:31:29.400] When you go from being somebody who does stuff constantly to being somebody who's like when you got money, but you're doing nothing.
[00:31:29.640 --> 00:31:33.480] The number one thing you want to do is start doing stuff again.
[00:31:33.480 --> 00:31:44.760] Well, but if your personality is built around you learning and doing things and suddenly you don't have anything like that, like your whole purpose in life, you're like, you're doing like, what is my purpose in life?
[00:31:44.760 --> 00:31:47.360] Yeah, was it to sell a company and make money and then do nothing?
[00:31:47.360 --> 00:31:48.800] Like, clearly not.
[00:31:48.800 --> 00:31:49.760] So, what's going to happen?
[00:31:44.840 --> 00:31:50.240] Let's talk about it.
[00:31:50.320 --> 00:31:57.680] Let's talk about this AI stuff because, like, I'm thinking, like, it just feels like there's a huge elephant in the room where half the stuff I build or consider building now.
[00:31:57.680 --> 00:32:01.760] I'm like, but isn't AI just going to make this obsolete in six months?
[00:32:01.760 --> 00:32:06.960] Like, we're building landing pages for indie hackers, like, basically a little profile.
[00:32:06.960 --> 00:32:16.400] And it's like, why would anybody ever use my website builder if six months from now or literally last week or two months ago, they could find some AI tool that will build them a landing page?
[00:32:16.720 --> 00:32:18.320] You got to build AI into it.
[00:32:18.320 --> 00:32:18.640] Right?
[00:32:18.640 --> 00:32:21.040] It's like, we have to sort of upgrade our stuff.
[00:32:21.040 --> 00:32:23.680] Like, we have to, or we're just going to be toast.
[00:32:24.000 --> 00:32:26.880] And it's like, it's so hard to wrap my mind around that.
[00:32:26.880 --> 00:32:28.800] Like, to some degree, I feel like a dinosaur.
[00:32:28.960 --> 00:32:32.320] I'm like, no, the old way of doing things is still here.
[00:32:32.320 --> 00:32:36.560] And the tried and true business fundamentals still apply.
[00:32:36.560 --> 00:32:39.760] But the technologist in me is like, no, it's all different.
[00:32:40.240 --> 00:32:41.920] Even just things in my life.
[00:32:42.080 --> 00:32:45.760] I want to spend a weekend just building an AI assistant for my own life.
[00:32:45.760 --> 00:32:46.960] I want it to text me.
[00:32:46.960 --> 00:32:48.000] I want it to email me.
[00:32:48.000 --> 00:32:49.120] I want it to call me.
[00:32:49.120 --> 00:32:50.160] I want it to remind me.
[00:32:50.160 --> 00:32:54.400] I want it to ask me questions about what I'm doing and then think and then tell me what I should do.
[00:32:54.400 --> 00:32:58.320] A year ago, this would have been extremely difficult to conceive of how it could even be.
[00:32:58.480 --> 00:33:01.280] It would be a huge, require a huge team to do something, right?
[00:33:01.280 --> 00:33:03.600] Well, now it's like, I can hook this up this weekend myself.
[00:33:03.600 --> 00:33:09.520] And like, I don't need a friend or a colleague or an expensive coach to motivate me or ask me what's going on.
[00:33:09.520 --> 00:33:12.800] Like, AI can do it extremely reliably.
[00:33:12.800 --> 00:33:14.480] It's the age of the entrepreneur, right?
[00:33:14.560 --> 00:33:19.360] Like, every repetitive job, everything that can be automated will be automated, right?
[00:33:19.360 --> 00:33:20.880] So, what stays?
[00:33:20.880 --> 00:33:21.200] Right.
[00:33:21.200 --> 00:33:25.520] And I think it's like the creativity in launching businesses.
[00:33:25.520 --> 00:33:27.840] Like, good luck trying to get that away with AI.
[00:33:27.840 --> 00:33:29.200] I don't think you can automate it.
[00:33:29.360 --> 00:33:30.360] And this is why I love AI.
[00:33:30.360 --> 00:33:33.160] Like, I don't like to do all these manual, boring stuff.
[00:33:33.160 --> 00:33:34.600] I don't want to make my own landing pages.
[00:33:29.840 --> 00:33:36.120] I don't want to write my own copy.
[00:33:36.280 --> 00:33:38.040] I don't want to do any of that.
[00:33:38.040 --> 00:33:41.880] I just want to build businesses, automate it, and move on to the next one.
[00:33:41.880 --> 00:33:46.520] So, I think that's why, like, as an entrepreneur, as an indie hacker, it's going to be an amazing time.
[00:33:46.520 --> 00:33:46.920] Agreed.
[00:33:46.920 --> 00:33:54.200] But, like, if you do a repetitive job, like financials, I think EBM just announced that they're going to fire 10% of their stuff.
[00:33:54.200 --> 00:33:55.080] And they said it's because of AI.
[00:33:56.280 --> 00:33:56.920] Yeah.
[00:33:56.920 --> 00:33:57.400] Because of AI.
[00:33:57.480 --> 00:33:58.360] Which I don't believe.
[00:33:58.360 --> 00:33:59.720] I don't think it's really because of AI.
[00:33:59.720 --> 00:34:04.280] I think that they needed to do layoffs and they're just what's going to happen to the SP.
[00:34:04.280 --> 00:34:10.040] Like, we're going into a recession, but all these companies just probably gonna save so much money in this.
[00:34:10.040 --> 00:34:15.320] Like, what is it going to do to the economy in like a stock market versus normal people way?
[00:34:15.320 --> 00:34:18.520] And I think, like, we, we are enthusiastic about it, right?
[00:34:18.600 --> 00:34:23.800] Like, but yeah, if you have an old job, yeah, I understand why you hate it.
[00:34:23.800 --> 00:34:26.680] So, there's this kind of hilarious book by this.
[00:34:26.680 --> 00:34:27.960] I think he was like an anthropologist.
[00:34:27.960 --> 00:34:32.360] He just died a couple of years ago, whatever, cognitive scientist, and it's called Bullshit Jobs.
[00:34:32.360 --> 00:34:39.720] And he's very like left-wing, so he's not coming at this from a perspective of like, you know, caring a lot about certain economic developments.
[00:34:39.720 --> 00:34:47.320] He really is just looking at our society and he's like, hey, capitalism creates a bunch of jobs that are horseshit, that aren't very meaningful, et cetera.
[00:34:47.720 --> 00:34:57.560] And his list of the jobs, I'll name a couple of them in a second, are the exact kinds of jobs that I think are in a lot of danger from AI.
[00:34:57.560 --> 00:35:07.240] And it makes me feel like there's going to be a lot of short-term pain and you know sort of a lot of job security is going to be gone there's you know this is going to be a very unstable place.
[00:35:07.240 --> 00:36:50.120] But in the long term, these types of jobs, like repetitive jobs, aren't the ones that I really imagine being that much loved lost for his book says over half of societal work is pointless and becomes psychologically destructive when paired with the work ethic that associates work with self-worth and so like that's like the like we're just looking this up now i haven't read this book but like this thesis of his book is that half the jobs we already have we shouldn't even have and then he has a list of like okay why what are these crappy jobs and so yeah and for context this is like 2018 so this wasn't you know he didn't see like he might have seen ai on the horizon but like you know he hadn't heard the term gpt before so flunkies right people that serve to make their superiors feel important goons who act a harm or to see others on behalf of their employer for example lobbyists corporate lawyers telemarketers duct tapers who temporarily fix problems that could be fixed permanently programmers repairing shoddy code airline desk staff who calm passengers whose bags do not arrive et cetera etc um and when i look at this list i'm just like oh that that is just the like the you know the first on the the victims that are going to get lined up against the wall from ai i disagree and it's like looking at his list like i don't like i think the things that he's saying are useless jobs are actually the jobs that are the safest the reason why is because a lot of the jobs that he's saying are useless are jobs that have to do with people's emotional state right like you lose your your bags at the airline you feel stressed you feel worried you're not sure if you're going to get it back people will pay money to feel better right or like flunkies he said of being a receptionist or being a door attendant or store greeter that just serve to make other people feel important people will pay money to feel important and feel good like i mean like there's some stuff in the ai landscape, like replica, like have an ai friend that you can talk to that is making some people feel good.
[00:36:50.120 --> 00:36:55.640] But I think that, generally speaking, for a while, people are mostly going to rely on other people to feel good.
[00:36:55.640 --> 00:36:59.400] And that's something ai is not going to be able to do, even when it can do everything else.
[00:36:59.400 --> 00:37:09.320] Which it seems like I'm pretty sure if an AI is gonna be nice to you and like tell you, oh, your bag will be back, like people would fucking hate it to be calmed by an AI in such a position.
[00:37:09.320 --> 00:37:19.720] I guess anyone who doesn't have a startup idea, they should get this book and just browse through it and be like, hmm, which one of these can I just like get a little piece of the pie on and automate out of it?
[00:37:20.120 --> 00:37:21.080] What are you doing for yourself?
[00:37:21.080 --> 00:37:28.920] Do you have any like behind the scenes Danny Postma AI tools that are making you more efficient and allowing you not to do repetitive tasks?
[00:37:28.920 --> 00:37:37.720] Mostly just rewriting my tweets, rewriting my content, because I'm not native English and I'm horrible in writing cohesion sentences for some reason.
[00:37:37.720 --> 00:37:44.920] So I will just write garbage in, like just do a brain fart of like 500 words for my tweets, put it in typefully.
[00:37:44.920 --> 00:37:46.840] Shout out to typefully.com.
[00:37:46.840 --> 00:37:50.600] Tell the AI to rewrite it in a cohesive way.
[00:37:50.600 --> 00:37:53.720] And then if the AI doesn't understand what I say, I know, okay.
[00:37:53.720 --> 00:37:57.400] Like if the AI doesn't understand what I say, no one is going to understand what I say.
[00:37:57.400 --> 00:38:03.480] Then I tell it to summarize it and then I tell it to make it a little bit more poppy and then I edit it for like 10 minutes.
[00:38:03.480 --> 00:38:05.240] So mostly those kind of things.
[00:38:05.240 --> 00:38:10.040] We should be doing this on the hacker scanning because we have a site where essentially people have to come in and write a lot.
[00:38:10.040 --> 00:38:13.320] They have to write posts, they have to write updates, they have to write comments.
[00:38:13.320 --> 00:38:16.120] We should be helping them write better content with.
[00:38:16.360 --> 00:38:19.000] I bet MicroAcquire already does this, for example.
[00:38:19.160 --> 00:38:24.120] If you go to microacquire.com, I think all their posts are written in the same way.
[00:38:24.120 --> 00:38:25.960] I wouldn't be surprised if they have.
[00:38:25.960 --> 00:38:26.360] Yeah.
[00:38:26.600 --> 00:38:36.680] You just put all these content in and then have an AI to generate it for them and then probably like one reviewer that makes some manual changes towards it.
[00:38:36.680 --> 00:38:37.080] Yeah.
[00:38:37.640 --> 00:38:39.000] I did this for Headline back in the day.
[00:38:39.000 --> 00:38:42.240] Like the first customers, they told me the output was garbage.
[00:38:42.240 --> 00:38:45.360] And you just look at the input and it's like, yeah, you input garbage.
[00:38:44.200 --> 00:38:47.200] Like input garbage in whoever's tied.
[00:38:47.200 --> 00:38:48.960] It's the same with Headshot Pro right now.
[00:38:44.600 --> 00:38:51.200] Like we refund 20% of the people because they upload garbage.
[00:38:51.840 --> 00:38:57.920] So for Headline, I was actually rewriting their inputs in the back end without them seeing it just to make it better.
[00:38:57.920 --> 00:38:59.280] So stuff like that is.
[00:38:59.280 --> 00:38:59.840] Yeah.
[00:38:59.840 --> 00:39:02.400] Well, this is one of the things that's already getting better.
[00:39:02.400 --> 00:39:04.960] Like the difference between GPT 3.5.
[00:39:04.960 --> 00:39:10.160] So GPT is like, you know, OpenAI's family of text-based AIs.
[00:39:10.160 --> 00:39:14.400] And you can use chat GPT to interface with it, or you can use the API.
[00:39:14.400 --> 00:39:19.360] But as technology gets better, you don't have to be as good of a prompt whiz kid.
[00:39:19.360 --> 00:39:29.520] Like right now, the people who are the best at using AI are just really good at understanding how the AI thinks, what kinds of things they should put in their prompt and their questions and their conversations with the AI to get it to do what they want.
[00:39:29.520 --> 00:39:32.560] But as it gets smarter, it gets more and more human-esque.
[00:39:32.720 --> 00:39:38.720] You could imagine five or ten years from now, maybe even a year from now, you could write absolute garbage.
[00:39:38.720 --> 00:39:40.080] Mid-Journey does it right now.
[00:39:40.080 --> 00:39:46.080] You can just type in and Midjourney actually optimizes your prompt and adds all these specifications to it so the AI better understands what you mean.
[00:39:46.080 --> 00:39:46.640] Boom.
[00:39:46.640 --> 00:39:47.200] There it is.
[00:39:47.520 --> 00:39:48.240] So yeah.
[00:39:48.240 --> 00:39:48.640] Yeah.
[00:39:48.960 --> 00:39:50.880] Are you guys using any tools I should know of?
[00:39:50.880 --> 00:39:52.720] Because it's actually quite stupid.
[00:39:52.720 --> 00:39:53.920] I only use one AI tool.
[00:39:54.880 --> 00:40:07.600] It's hard because I think being a builder and being an eddy hacker, I would always get stuck in these creative black holes where I would just be spending so much of my time creating that I would just get really behind on like using, right?
[00:40:07.920 --> 00:40:10.240] And I think I see like Peter Levels is like this too.
[00:40:10.240 --> 00:40:12.240] He's like, I'm using PHP, blah, blah, blah, blah.
[00:40:12.320 --> 00:40:14.320] He's like, he doesn't care how his code looks, right?
[00:40:14.320 --> 00:40:15.680] He's just trying to create stuff.
[00:40:15.680 --> 00:40:20.640] He's not trying to be like up to date with the latest technology and build systems and whatever as an engineer.
[00:40:20.640 --> 00:40:22.080] And I feel like that too as an Eddie hacker.
[00:40:22.080 --> 00:40:24.560] I'm like, I'm trying to like create value for other people.
[00:40:24.560 --> 00:40:30.840] So I'm not like sitting around like, you know, licking my finger and sticking it in the air and like trying to figure out what toys I should be playing with.
[00:40:30.840 --> 00:40:31.640] Like I'm building.
[00:40:29.760 --> 00:40:32.840] And that's kind of how you are too.
[00:40:33.160 --> 00:40:40.840] But like in a way, it's like we kind of need to strike this balance between like becoming more efficient builders and continuing to build.
[00:40:40.840 --> 00:40:49.720] I would say that that is something that like basically sort of, you know, what balance do you strike between like building in the ways that you know and exploring, right?
[00:40:49.720 --> 00:40:56.040] Like basically sort of having almost like a skunk works built into your own like way of working.
[00:40:56.040 --> 00:40:58.280] Basically what you should do is you should look at the times.
[00:40:58.280 --> 00:41:03.480] I think three or four years ago, hey, you know, this is the big, you know, the crypto wave is coming.
[00:41:03.480 --> 00:41:07.320] I think that, you know, some of the other waves that have come, no code, et cetera.
[00:41:07.320 --> 00:41:11.720] I think you can kind of poke your head out of the hole and go, ah, that's not really for me, right?
[00:41:11.720 --> 00:41:13.160] That's not revolutionary.
[00:41:13.160 --> 00:41:14.120] It's not a huge change.
[00:41:14.120 --> 00:41:18.600] I don't necessarily need to divide my time away from the things that I know are working and that I know.
[00:41:18.680 --> 00:41:20.600] None of those like made you more productive.
[00:41:20.600 --> 00:41:26.440] Maybe, maybe the creator economy thing where it's like, oh, hey, you can make money without writing any code.
[00:41:26.440 --> 00:41:34.760] You could just basically show your work, build in public, have an additional revenue channel that's lighter weight than building a full fledged SaaS app.
[00:41:34.760 --> 00:41:37.560] Maybe like that was like, it's not really making you more productive.
[00:41:37.560 --> 00:41:39.640] It's just like, oh, there's easier ways to make money.
[00:41:39.640 --> 00:41:43.960] But now it's like, oh, like the new sort of technological wave is actually making you more productive.
[00:41:43.960 --> 00:41:46.040] To me, it's AI and then the gig economy.
[00:41:46.040 --> 00:41:54.680] The gig economy, it becomes way easier, for example, to quickly hire a contractor or quickly get someone to whip you up a quick design, like a quick logo.
[00:41:55.160 --> 00:41:55.640] That kind of thing.
[00:41:55.640 --> 00:41:57.720] It's like, hey, you need to jump on that bus, right?
[00:41:57.720 --> 00:42:04.920] Like take a long weekend and just figure this out because it's going to, I don't know, that was, you know, 0.3x your productivity.
[00:42:04.920 --> 00:42:11.080] Now AI is here and it's like, dude, you might 0.5, you know, 1x, 2x your productivity, right?
[00:42:11.080 --> 00:42:11.800] Easily.
[00:42:11.800 --> 00:42:30.960] I wonder when someone is going to launch Fiverr with an extra A in the site, where five, like AI Fiverr, and it's just going to get all the jobs that are listed on their copywriting, checking, translating, is basically just going to have the same UI on it, have the gig works done by an AI, so it doesn't feel like an AI is doing it anymore.
[00:42:30.960 --> 00:42:36.560] I've seen at least five or six AI landing page generator tools out there right now, right?
[00:42:36.560 --> 00:42:38.240] It's like people are already doing this.
[00:42:38.640 --> 00:42:47.440] 100%, by the way, when this episode ends, I'll give it two days, and then there's going to be someone who launches and it's going to be called one or because Fiverr is about $5.
[00:42:47.440 --> 00:42:52.800] Now there's going to be an AI tool, it's going to be five times less the operating cost, and they're going to call it Wanner.
[00:42:53.120 --> 00:42:54.800] But I think that's what people are doing wrong.
[00:42:54.800 --> 00:42:59.760] For example, a friend of mine, Mark, they have this landing page generated, but they priced it $19.
[00:42:59.760 --> 00:43:06.880] But you need to understand, like, you are removing the value of a landing page designer that's $5,000.
[00:43:06.880 --> 00:43:11.040] Like, you don't have to price it cheap just because it is cheap.
[00:43:11.040 --> 00:43:19.840] Like, for Hetshot Pro, for example, like, it cost me a few dollars to do it, but I'm not going to price it at $10 because you're competing with a photo shoot.
[00:43:19.840 --> 00:43:20.400] Right.
[00:43:20.400 --> 00:43:30.560] And as long as they're not competitors and you're not going, like, as long as you're like a fancy tool and you know what content you're doing and you're good, like you can charge a high price now while it's still new.
[00:43:30.560 --> 00:43:35.200] Like probably in five years, this doesn't work anymore, but for now, be expensive, man.
[00:43:35.200 --> 00:43:36.880] Dude, this is the exact point.
[00:43:36.880 --> 00:43:40.560] This is the call back to the point about the benefit of being fast.
[00:43:40.560 --> 00:43:53.000] I mean, look, if the ultimate cost of like, you know, sort of whatever it is, right, a headshot builder for AI or whatever is, you know, say it's like, you know, a 0.5, you know, or it's five times less expensive to create it.
[00:43:53.000 --> 00:43:59.280] Look, in the fullness of time, a bunch of competitors are going to like, you know, sort of enter into the space and drive down the cost.
[00:43:59.280 --> 00:44:05.800] But if you're fast, then the pricing theory just becomes, right, what's the alternative to this, right?
[00:44:05.800 --> 00:44:07.560] What is the incumbent or whatever it is?
[00:44:07.880 --> 00:44:14.680] This is what Peter, this, what Peter did when he launched the Avatar AI, he charged $50, and he was the only one doing it.
[00:44:15.000 --> 00:44:18.840] So he could just charge whatever the fuck he wants because everyone wants to use it.
[00:44:18.840 --> 00:44:23.080] And then gradually I came in, we had to like lower a little bit of a price.
[00:44:23.080 --> 00:44:27.080] And then, of course, Lanza came out and we had to slice our price to 80% down.
[00:44:27.080 --> 00:44:31.240] I think we were at $40 and we had to go down to $10, $7.
[00:44:31.240 --> 00:44:40.280] But most of our revenue, I think Peter got like 90% of his revenue in the first weeks because he could charge so much money and then everyone had to go down.
[00:44:40.280 --> 00:44:43.320] So yeah, this is coming back again to your point and what I said.
[00:44:43.800 --> 00:44:48.920] If it's a hype, get the most money out immediately because you can.
[00:44:48.920 --> 00:44:50.680] You're not going to last.
[00:44:50.680 --> 00:44:56.520] There's so many reasons right now the best strategy is to have small products that you move fast on.
[00:44:56.520 --> 00:44:59.480] Because number one, you have no idea when you're going to be disrupted.
[00:44:59.480 --> 00:45:05.800] These incumbents might come by and crush you and whatever money you were able to make for six weeks or six months, like that's now gone.
[00:45:05.800 --> 00:45:08.600] You might be disrupted not only by incumbents, but by technology itself.
[00:45:08.600 --> 00:45:11.160] What you're building might become completely obsolete.
[00:45:11.160 --> 00:45:22.040] Like one of the things I see right now that's happening is everybody's kind of taking existing applications and slapping AI on top of it and getting like this 5x boost of like, hey, this is like five times better than what I had in the past, which is awesome.
[00:45:22.040 --> 00:45:29.160] But like, okay, we're still waiting for the wave of companies to come that are building products from the ground up with this AI in mind.
[00:45:29.160 --> 00:45:30.520] And that's going to be much better.
[00:45:30.520 --> 00:45:44.280] Like a community forum, for example, that helps you occasionally, you know, use AI, or like a social product like Twitter, that helps you occasionally use AI to like write better tweets, is not going to be as good as a social network that was built from the ground up with AI in mind.
[00:45:44.280 --> 00:45:45.600] And that stuff is coming, right?
[00:45:44.440 --> 00:45:47.040] Like that stuff's going to stand the test of time.
[00:45:47.280 --> 00:45:56.240] And that's going to lead to another second-order effect where a world full of all these AI products is going to create a whole bunch of new problems that we can't even imagine right now that are hard to imagine.
[00:45:56.240 --> 00:45:58.400] And there's going to be startups that solve those problems.
[00:45:58.400 --> 00:46:05.440] And so if you're building something that's going to take 10 years or five years to slowly come to fruition, I think you're dead.
[00:46:05.440 --> 00:46:10.080] Yeah, this is why as indie hacker, I think don't compete with the big companies.
[00:46:10.080 --> 00:46:12.480] Like I see so many people, I made the same mistake.
[00:46:12.960 --> 00:46:17.040] The first AI thing I made was like a stock photo, AI stock photo website.
[00:46:17.040 --> 00:46:20.720] But it's fucking dumb because Adobe has all the photos and libraries.
[00:46:21.040 --> 00:46:22.560] Why would I compete with it?
[00:46:22.560 --> 00:46:25.600] You have three months of time, and then someone else will come out.
[00:46:25.600 --> 00:46:30.160] And I see other people, like, they make this AI generator where you can generate pictures.
[00:46:30.160 --> 00:46:31.040] It's like an editor.
[00:46:31.040 --> 00:46:34.560] But Adobe just launched the Firefly thing and you're done for.
[00:46:34.720 --> 00:46:36.320] There's no way you can compete with it.
[00:46:36.560 --> 00:46:40.160] You should find a niche, ship really fast in that niche.
[00:46:40.160 --> 00:46:43.920] Because a big company is not going to outcompete a niche in that sense.
[00:46:43.920 --> 00:46:44.160] Nope.
[00:46:44.160 --> 00:46:45.360] It's not worth it to them.
[00:46:45.360 --> 00:46:59.920] And I think this raises an interesting question, which is like every time there's a technological shift, some percentage of it goes to the incumbents, the big players, some percentage of it goes to the startups, and some percentage of it goes like even a level below, like the anti-hackers.
[00:46:59.920 --> 00:47:02.240] Like I'm trying to make some money online right now, right?
[00:47:02.240 --> 00:47:06.960] Like with something like crypto, for example, the incumbents didn't make very much money from crypto.
[00:47:06.960 --> 00:47:15.120] All the biggest crypto companies were kind of startups because they moved really fast and it was like kind of niche enough that big companies were like, I don't know about this.
[00:47:15.120 --> 00:47:18.960] Whereas something like mobile, for example, all the incumbents took all the money, right?
[00:47:18.960 --> 00:47:21.360] Like Apple and Google created the phones.
[00:47:21.360 --> 00:47:23.600] Every big website made its own mobile app.
[00:47:23.600 --> 00:47:26.400] And like very few startups got in there and disrupted things.
[00:47:26.400 --> 00:47:28.480] I'm curious what it's going to be like with AI.
[00:47:28.480 --> 00:47:37.800] Like you were saying, like Adobe, like they can just slap AI into Photoshop and have a giant gallery of images and AI effects, and they're just going to kill everything because everybody already has Photoshop.
[00:47:37.800 --> 00:47:45.160] But from the bottom up, indie hackers can do what you're saying and target these super tiny niches that no incumbent's going to target.
[00:47:45.160 --> 00:47:51.480] And I think what's really cool is the fact that it's so fast and easy to build these AI apps now because you're basically writing prompts.
[00:47:51.480 --> 00:47:53.560] You're not writing a ton of code.
[00:47:53.560 --> 00:48:04.520] That there's no reason not to have super specialized apps for tons of little niches, which I think is going to make it harder for anyone to get a bigger foothold with a bigger, broader company that's targeting more people.
[00:48:04.520 --> 00:48:13.640] Like to your idea of like a Fiverr powered by AI, I can imagine 100 different profiles of AI copywriters, but they all have their own little different flavor because they all have a different prompt.
[00:48:13.640 --> 00:48:15.320] And essentially, you could pick a different one.
[00:48:15.560 --> 00:48:18.680] There isn't just going to be one AI copywriter to rule them all.
[00:48:18.680 --> 00:48:23.960] Why would there be if you could make a specialized AI that could be super duper good at each individual task?
[00:48:23.960 --> 00:48:30.040] Yeah, I think what I was shocked with on my Twitter is like, it feels like almost no indie hackers is doing SEO.
[00:48:30.040 --> 00:48:32.760] Well, SEO, like we indie hackers, we hate marketing, right?
[00:48:32.760 --> 00:48:43.400] And this is the easiest way to A, find a product idea, B, launch a niche product, and then C, don't have to do any marketing for it because it just stands there.
[00:48:43.400 --> 00:48:46.680] So that was like the lucky part I had with the ProvoPictor website.
[00:48:46.680 --> 00:48:48.280] Like there was no competition on it.
[00:48:48.280 --> 00:48:52.040] So I would just launch my website, get a backlink, and then just rank instantly.
[00:48:52.040 --> 00:48:54.520] So you can find like hyper-specific niches.
[00:48:54.520 --> 00:48:55.320] You don't have to compete.
[00:48:55.320 --> 00:48:56.280] You don't have to do marketing.
[00:48:56.280 --> 00:49:02.440] And you would just keep it because why would any other big company go down on that keyword for you?
[00:49:02.760 --> 00:49:04.200] By the way, I was going to ask you that.
[00:49:04.680 --> 00:49:15.600] You mentioned at the beginning of this call, actually, that you spend a lot of time on Google looking at keywords in relation to a lot of these, like the latest of your products on Postgrafts.
[00:49:14.840 --> 00:49:21.120] And then you also, you don't, you know, sort of, you're not building cool AI to help you with your own processes.
[00:49:21.280 --> 00:49:23.680] Like, you're not necessarily scratching your own itch.
[00:49:23.680 --> 00:49:30.240] Like, is Google is like, are you know, keyword search and SEO the thing that you use to find ideas?
[00:49:30.880 --> 00:49:33.280] Yeah, well, to find ideas, kind of.
[00:49:33.280 --> 00:49:38.800] So last year I started trying to do this with the program with programmatic SEO, like Fig components, tailbits.
[00:49:39.360 --> 00:49:42.080] These were just completely based on SEO research.
[00:49:42.080 --> 00:49:49.840] I was just searching all day, all week for ideas what I could build next without AI, because AI generation didn't exist yet and I had an On Copy.
[00:49:49.920 --> 00:49:52.000] So what am I going to do?
[00:49:52.400 --> 00:49:54.400] So yeah, I built based on that.
[00:49:54.400 --> 00:49:59.840] And this time, so with the Avatar race, everyone chose the avatar work, right?
[00:49:59.840 --> 00:50:01.120] But there's no searches on it.
[00:50:01.120 --> 00:50:04.320] So I chose Profile Picture specifically.
[00:50:04.320 --> 00:50:08.320] So I got the idea before as something I want to explore.
[00:50:08.320 --> 00:50:18.720] But then when I want to launch something, I also check, hey, which keyword is underrepresented, which is on the keyboard difficulty of 10, and then use that in your domain name, in your titles.
[00:50:18.720 --> 00:50:24.640] And this is like if you put my website in href, like I have a shit ton of traffic on Alders Profile Picture.
[00:50:24.640 --> 00:50:29.760] Well, I know, and this is why levels, YPETA went down, for example, because there are no searches on Avatar.
[00:50:29.760 --> 00:50:32.960] So once the hype is gone, there is nothing sustainable in it.
[00:50:32.960 --> 00:50:34.480] And the same with Headshot right now, right?
[00:50:35.040 --> 00:50:38.960] I get hype on TikTok, I get backlinks, but there is a lot of search on Headshot.
[00:50:38.960 --> 00:50:47.280] So you take advantage of the hype, and then you build for the longevity of it based on the keywords.
[00:50:47.280 --> 00:50:49.680] What do you think about the future of being like a founder, right?
[00:50:49.680 --> 00:50:52.960] Like it's cool because it's kind of like a golden era for indie hackers.
[00:50:52.960 --> 00:50:54.720] We're going to be much more efficient.
[00:50:54.720 --> 00:50:56.080] We can build much more.
[00:50:56.080 --> 00:50:57.920] But also, the competition is ramped up.
[00:50:57.920 --> 00:51:02.440] Way more people now can build apps that do very impressive things.
[00:50:59.760 --> 00:51:07.000] Any idea that's basic, there's going to be 10 million versions of it in like two days.
[00:51:07.880 --> 00:51:09.320] And so you're going to have to be super creative.
[00:51:09.320 --> 00:51:12.200] But even then, people can copy you super duper quickly.
[00:51:12.200 --> 00:51:14.120] So where is the defensibility?
[00:51:14.120 --> 00:51:14.840] Where are the moats?
[00:51:14.840 --> 00:51:17.320] Is it all just going to be these fast cash grabs?
[00:51:17.320 --> 00:51:21.320] Are there businesses that are going to be immune to this and that can stand the test of time?
[00:51:21.320 --> 00:51:24.680] Well, the thing is, I have a lot of competition with Hedgehog Pro and ProfoPicts, right?
[00:51:24.680 --> 00:51:25.560] But I think I have a mode.
[00:51:25.560 --> 00:51:34.680] Like all my competitors with Headshot, they have a lower quality because I figured out how to make my own custom model that looks more real life using some other techniques.
[00:51:34.680 --> 00:51:36.040] Like it's not just one AI.
[00:51:36.040 --> 00:51:38.840] It's like 10, 20 AI stacked on top of each other.
[00:51:38.840 --> 00:51:41.560] So you have like that is a moat.
[00:51:41.560 --> 00:51:45.720] On the other side, if you don't want competition, don't tweet too much about it.
[00:51:45.720 --> 00:51:47.000] Just keep your revenue numbers quiet.
[00:51:47.560 --> 00:51:52.600] A niche that's really boring that no one wants to jump into and just like be caching.
[00:51:52.600 --> 00:51:54.600] Yeah, don't come on the Andy Hackers podcast.
[00:51:55.000 --> 00:51:56.120] I mean, you've been through a lot, Danny.
[00:51:56.120 --> 00:51:57.320] You sold a company.
[00:51:57.320 --> 00:52:00.280] You have launching products, it seems like every other week.
[00:52:00.600 --> 00:52:05.320] You're at the forefront of like these new AI image generation photo editing tools.
[00:52:05.320 --> 00:52:11.720] What's something you've learned that you think other indie hackers who are just getting started, I mean, people listening to the show, some of them don't know how to code.
[00:52:11.720 --> 00:52:13.160] Many have never built anything.
[00:52:13.160 --> 00:52:16.600] Some people are two years into a product that's not making any money.
[00:52:16.600 --> 00:52:19.560] What do you think they can take away from your story?
[00:52:19.880 --> 00:52:29.320] Ship fast, build a lot of different things to find what works, and do SEO for marketing if you don't like marketing.
[00:52:29.320 --> 00:52:34.360] I think those are the three things to get you out of the slum.
[00:52:34.360 --> 00:52:34.840] Love it.
[00:52:34.840 --> 00:52:36.520] Danny, thanks again for coming on the show.
[00:52:36.520 --> 00:52:39.080] Can you tell people where they can find your stuff?
[00:52:39.080 --> 00:52:41.160] Is it just postcrafts.com?
[00:52:41.160 --> 00:52:42.680] Yeah, you can, mostly on my Twitter.
[00:52:42.680 --> 00:52:52.000] So if you go to twitter.com/slash Danny Postma with two A's in the end because the other one wasn't available, I post there with everything that I learned.
[00:52:52.160 --> 00:52:54.880] Yeah, I actually own the other, I own the other handle.
[00:52:54.880 --> 00:52:56.240] I just cannot get access to it.
[00:52:56.240 --> 00:52:56.720] It sucks.
[00:52:56.720 --> 00:52:57.200] Nice.
[00:52:57.520 --> 00:52:58.320] Yeah, just follow me there.
[00:52:58.320 --> 00:53:00.960] And there's like a bunch of links to all my latest projects, what I'm working on.
[00:53:00.960 --> 00:53:03.040] I'm sharing all my struggles, my learnings.
[00:53:03.440 --> 00:53:04.000] Yeah.
[00:53:04.000 --> 00:53:04.240] All right.
[00:53:04.240 --> 00:53:05.360] We'll put you in the show notes, too.
[00:53:05.440 --> 00:53:06.000] Thanks again, Danny.
[00:53:06.480 --> 00:53:07.520] Cheers, guys.