Bản chép lời đầy đủ của 322 - Joe Masilotti

let alone what will happen in the future. 00:05:35.680 --> 00:05:39.839 Um, but I think the one safe thing to say is that things are gonna be different. 00:05:40.000 --> 00:05:43.680 We don't know how, but they are, I mean, they're already different. 00:05:43.839 --> 00:05:44.079 Yeah. 00:05:44.240 --> 00:05:48.480 Um, the the job market for developers is different than it used to be. 00:05:48.720 --> 00:05:54.319 Um I think it's different for business too, both in positive and negative ways. 00:05:54.480 --> 00:06:02.879 And the question that I've been asking myself, and it sounds like you and Colleen have been asking yourselves as well, is what do we do? 00:06:03.199 --> 00:06:06.079 What are the best moves in this new world? 00:06:06.560 --> 00:06:08.000 Yeah, exactly. 00:06:08.240 --> 00:06:13.040 A big a big angle on it is like we don't have bosses, we're our own bosses. 00:06:13.120 --> 00:06:16.480 Uh, name of the podcast is permission not required. 00:06:16.720 --> 00:06:19.519 Idea being we don't have to ask anyone for permission. 00:06:19.600 --> 00:06:34.720 Like, we are our own bosses in this case, and that comes with a ton of freedom to explore whatever we want to explore, but also at a ton of risk, where if we make the wrong decision, we cannot provide for our families anymore. 00:06:34.800 --> 00:06:39.120 You know, it's like it's not like, oh, the business goes under, I'll get like whatever uh I'll get severance. 00:06:39.279 --> 00:06:43.839 It's like, no, that that is the end of the revenue stream for for our family. 00:06:44.160 --> 00:06:58.319 So that is just like this interesting take on all of this, where so much is riding, so much is on the line right now, you know, and figuring out how to a make it work and hopefully thrive, you know. 00:06:58.399 --> 00:07:05.279 That's that's the goal, but it kind of feels like a lot of people are scrambling, and uh I'm one of them right now. 00:07:05.600 --> 00:07:11.680 Yeah, yeah, and the risk is real, like it's not just a remote possibility, I don't think. 00:07:12.079 --> 00:07:27.360 Um because I'd you know, people are getting laid off, people are out there and not finding work, and and it's it's not exactly the same for consultants and employees, but it's just two points on a continuum. 00:07:28.000 --> 00:07:39.920 You know, um freelancers and agencies and and these these sorts of workers are facing a lot of the same challenges as employees. 00:07:40.079 --> 00:07:46.240 It's it's hard for a job seeker to find a job, and it's harder for agencies to find work. 00:07:46.399 --> 00:07:56.160 Uh clients expect more out of the agencies now because they know about AI and they believe that AI can make people more productive. 00:07:56.879 --> 00:08:06.319 So I I imagine I haven't talked with many like freelancers or consultants in in in the last little while, but I assume you're dealing with a lot of the same stuff. 00:08:06.639 --> 00:08:07.360 Oh, yeah. 00:08:07.519 --> 00:08:25.839 It's been it's been a big point of it's taken up a lot of brain power over the past couple months of where does my consulting business fit in in this new world where code has become a commodity already, you know, let alone what it's going to be in three months. 00:08:25.920 --> 00:08:28.240 Like code is practically already a commodity. 00:08:28.399 --> 00:08:39.440 And I've built a consulting business on building mobile apps for people, where my deliverable at the end of the day was written down as code that you now can put in the app store. 00:08:39.919 --> 00:08:49.840 And a lot of my the last couple of weeks, a couple of months has been me figuring out how do I break out of that and go a level higher. 00:08:50.159 --> 00:08:55.200 And how do I no longer like I was always delivering more than just code. 00:08:55.360 --> 00:09:00.879 I was helping you with, you know, app store review and dealing with Apple and advising on which native features to build. 00:09:01.120 --> 00:09:06.879 But instead of selling that, I was selling code, and now I'm swapping that. 00:09:07.120 --> 00:09:11.200 It's more about like I will help you figure out what to build. 00:09:11.440 --> 00:09:17.360 You need a mobile app, talk to me, like let's figure out what features you need to build, how to get into the app store. 00:09:17.519 --> 00:09:18.799 Do you need in-app purchases? 00:09:18.960 --> 00:09:21.440 How to save 15% on fees? 00:09:21.679 --> 00:09:37.360 Like, I was always doing that, but now I'm actually leading with that in my positioning and my branding and my offering because that is what people have more value in, or think they have more value in, because like they could just spit up the app themselves. 00:09:37.519 --> 00:09:39.120 You know, that's what that's what I'm hearing. 00:09:39.279 --> 00:09:44.080 And they can spin it up, but they're still coming back to me and being like, now how do we get it in the app store? 00:09:44.320 --> 00:09:48.720 Because Apple wrote us this email and denied us, and I don't know how to how to fight this. 00:09:48.799 --> 00:09:53.519 And it's like, okay, well, I've done this 25 whatever times, like I'll help you. 00:09:53.679 --> 00:10:06.559 Uh, so it's been interesting for me, especially, where I'm like off I'm doing the same stuff, but I'm totally swapping how I market it in this in this new world. 00:10:07.200 --> 00:10:11.919 I want to get a little clearer on what we mean by code is now a commodity. 00:10:12.320 --> 00:10:15.120 I I like partly agree and partly don't. 00:10:15.360 --> 00:10:27.519 And there there are some terms that are fuzzy, like uh we might use the terms like coding, programming, engineering, all that stuff interchangeably. 00:10:28.320 --> 00:10:33.440 Um, but there's like distinctions that can be made among those and stuff like that. 00:10:33.679 --> 00:10:48.320 So like uh an LLM can write code and it can get it to work, but that's not really the same thing, of course, as system design, and

it's uh certainly not the same thing as product design. 00:10:48.559 --> 00:10:48.799 Yeah. 00:10:48.960 --> 00:10:56.000 And so there's still some like um valuable human discretion and taste and stuff like that in there. 00:10:56.159 --> 00:11:12.720 So I sp I suspect that what you mean when you say code has but become a become a commodity and tell me if I'm getting this wrong, but I suspect you mean like that lower level where it's like you decide what you want, and then the LLM can write the code. 00:11:12.799 --> 00:11:18.639 You're still the programmer, but the code part is is abstracted away, you don't have to do that anymore. 00:11:19.279 --> 00:11:22.480 Yeah, yeah, I think that was a really good distinction you made at the beginning of that. 00:11:22.639 --> 00:11:37.679 The difference between code and engineering and code and product and code and strategy, like the physical typing of keys on the keyboard is has been commoditized, has been given over to the robots, and if not now, it will in the next couple of months. 00:11:37.840 --> 00:11:40.000 Like I'm pretty confident about that. 00:11:40.240 --> 00:11:45.120 Just and I don't even say like across the world, I mean for my workflow. 00:11:45.360 --> 00:12:07.200 Like I am letting the the robots do the actual hands-on keyboard stuff, and I'm reviewing it and I'm architecting it, and I'm coming up with like how to build those things because I've done it so many times and I know what should be done, but I'm not typing you know, def function and like you know, that kind of stuff anymore. 00:12:07.279 --> 00:12:11.840 Like I'm less in my editor and more in like the review diff viewing phase. 00:12:12.159 --> 00:12:13.919 Same, exact same for me. 00:12:14.159 --> 00:12:30.559 What do you think is the likelihood that AI basically ingests all the same stuff as what's in your book and it like knows as much as your book does, uh, it has all that same knowledge and even more. 00:12:30.720 --> 00:12:33.360 What do you think is the likelihood of that happening? 00:12:33.919 --> 00:12:43.840 For across the world for the every developer, um low, I would say, because it's still relatively new and the models haven't caught up with it yet. 00:12:44.159 --> 00:12:45.440 No, I just mean eventually. 00:12:45.679 --> 00:12:47.600 Oh, eventually, a hundred percent. 00:12:48.080 --> 00:13:00.159 Um, I mean, like, for me, my workflow, and I encourage anyone who has purchased any PDF of a book, like feed, give that to the LLM as a resource as you work on those type of things. 00:13:00.399 --> 00:13:02.000 I've written that book, right? 00:13:02.159 --> 00:13:04.720 Um, Hotwire Native for Rails developers. 00:13:04.799 --> 00:13:09.360 I've written it, but I wrote it now when I actually started writing it over almost two years ago. 00:13:09.679 --> 00:13:12.480 And I have forgotten some of those things. 00:13:12.639 --> 00:13:21.840 So what I do whenever I kick off a Hotwire Native project or a mobile app, I will have that as a text-only resource for the LLM that they can consume. 00:13:22.000 --> 00:13:30.159 And it's like, not only am I getting my own knowledge, I'm amplifying it because the LLM kind of knows the rules that I already know how to build these apps with. 00:13:30.559 --> 00:13:34.639 Yeah, Joe, I just realized that I am a stupid idiot. 00:13:34.799 --> 00:13:44.639 Um, because I wrote my book, Professional Rails Testing, with, you know, it's hundreds of pages of exactly the way I think things should be done. 00:13:45.519 --> 00:13:58.159 And uh LLMs, to be frank, I think are shit at writing tests because they're they're going off of the examples they see in people's repos, and those examples are not very good examples. 00:13:58.399 --> 00:14:00.720 And I I've gotten frustrated with it. 00:14:00.879 --> 00:14:04.240 Um, and it never occurred to me just to give it my book. 00:14:04.879 --> 00:14:05.519 Yeah. 00:14:05.759 --> 00:14:06.159 Yeah. 00:14:06.320 --> 00:14:06.960 It's huge. 00:14:07.120 --> 00:14:11.840 Well, first of all, you can't say that you're stupid and then finish that sentence with the book that I wrote. 00:14:11.919 --> 00:14:14.720 Let's just let's just put that down for sure. 00:14:14.879 --> 00:14:17.360 Um, that requires some level of competency. 00:14:17.519 --> 00:14:20.159 Um, but yeah, it's like a huge breakthrough. 00:14:20.399 --> 00:14:24.639 And yeah, us for for us authors, it's distilling knowledge we already knew. 00:14:24.799 --> 00:14:41.679 But if you're an individual out there and you're working on uh a code base that needs a bunch of new tests or is like starting up the testing, and you own Jason's book on testing, toss that into the LLM as a resource, and now you have Jason sitting by you that you could ask questions to. 00:14:41.919 --> 00:14:53.440 It's like the ultimate chat with PDF app that has everyone has built, you know, and launched at the beginning of the AI surge, where instead of you're just like chatting with it, you have the author sitting next to you. 00:14:53.679 --> 00:15:05.279 And and that framing is huge because you get a point of view that is more knowledgeable than you, that has like seen it, has been around the block a few times, but it's in your code base. 00:15:05.440 --> 00:15:10.080 So you could have it say, Okay, chapter three has this, these like concepts. 00:15:10.240 --> 00:15:13.279 How do I apply those concepts to this code base? 00:15:13.840 --> 00:15:28.000 And now you're like not only getting better results, but you're also learning what is being architected because it's coming from the book, and and that's been a huge level up for me in just like learning things. 00:15:28.639 --> 00:15:29.519 Interesting. 00:15:30.000 --> 00:15:36.879 I'm I'm gonna do that as soon as I can after this, put put my book in there and see see what happens. 00:15:37.360 --> 00:15:49.679 Um yeah, something that I think about sometimes is like what what's the AI gonna get good at and

what is it not gonna get good at anytime soon? 00:15:50.159 --> 00:15:58.480 And and and my estimation of how smart it's gonna get, how quickly, keeps going up. 00:15:58.799 --> 00:16:03.600 I keep realizing that I've underestimated how good it can get, how fast. 00:16:04.159 --> 00:16:13.600 The level it's at right now is already it it it's not like ten times as smart as I thought it would be by now, but it's definitely smarter than I thought it'd be by now. 00:16:13.679 --> 00:16:15.840 I'm I'm impressed by what it can do. 00:16:16.639 --> 00:16:31.600 I rarely run into a AI wall, like where the where the I'm using Claude 100% of the time, uh where Claude can't do something, or it's like gets stuck or can't figure it out. 00:16:31.679 --> 00:16:43.600 And like sometimes I'll you know, it'll hit a roadblock and I'll have to direct it or guide it in a way because I've done this thing a hundred times, but it's rare that Claude is like, sorry, can't do that, or you know, whatever. 00:16:43.759 --> 00:16:45.840 Like, this is wild. 00:16:46.159 --> 00:16:51.759 Um you're familiar with CI, you're you know, building Saturn CI, you know how all those workflows work. 00:16:52.000 --> 00:17:09.440 I have this little tool that I'm working on right now that does um automated iOS app builds, where essentially you click a button, it pings your server, does some stuff in your Rails app, and then kicks off a GitHub build, GitHub action build that packages the app, uploads it to test flight for you. 00:17:09.599 --> 00:17:12.319 And like there's 17 webhooks involved with that. 00:17:12.400 --> 00:17:19.279 It's talking to App Store Connect, it's doing your you know PA identification, signing certificates, it's doing all that. 00:17:19.519 --> 00:17:28.960 And the LLM has has like orchestrated all of that when all I was saying was like this is the end goal, let's walk through the plan. 00:17:29.200 --> 00:17:33.839 Let's figure out oh, Apple doesn't have a webhook to like let me know when a build is ready. 00:17:34.000 --> 00:17:35.279 Cool, what are our alternatives? 00:17:35.440 --> 00:17:38.960 Let's do polling, let's do um like a manual refresh. 00:17:39.200 --> 00:17:54.559 And just like keeping momentum going has been so huge in that where it doesn't get stuck very often, which is to go back to what you said before, like I keep underestimating how awesome it can be. 00:17:55.119 --> 00:18:11.039 And I think that the last frontier for me, well, the next frontier, because I'm sure there'll be a new one, is more of the visual design space, where I still can't get Claude to do good design for me. 00:18:11.200 --> 00:18:20.880 I still can't get Claude to um make stuff look and feel good that isn't disconnected from like the rest of the app. 00:18:21.039 --> 00:18:28.880 And and I think that hopefully Wathen, you know, Adam Wathen is like figuring that out with the UI.sh stuff that he's working on over at Tailwind. 00:18:29.279 --> 00:18:38.079 Um but that is still like something that I I'm I do by hand still, I guess is the right way to put it. 00:18:38.480 --> 00:18:39.599 Right, yeah. 00:18:40.000 --> 00:18:47.359 Yeah, but I can't think of any reason in principle why AI can't eventually get good at that. 00:18:48.240 --> 00:18:48.799 Yeah. 00:18:49.440 --> 00:19:00.319 And and something that is uh I I've noticed that seems different between now and some months ago is this the level of apparent like common sense. 00:19:00.640 --> 00:19:28.559 Like before it seemed like AI had like no common sense, it was basically just a sophisticated database where it was like, hey, do this, and it just knows a bunch of stuff, so it can write the code for you, and that's great, but it didn't it if um if it ran into a problem where the solution was obvious or whatever, um it wouldn't necessarily be like, oh, this is true and this is also true, so clearly I should do this. 00:19:28.720 --> 00:19:35.039 It didn't really have that, but now it does, like it makes sensible decisions on its own. 00:19:35.839 --> 00:19:36.880 Counterpoint. 00:19:37.119 --> 00:19:41.519 I saw someone ask all of the latest models the other day. 00:19:41.759 --> 00:19:45.200 Uh the car wash is 50 meters away. 00:19:45.359 --> 00:19:46.960 Should I walk or should I drive? 00:19:47.200 --> 00:19:53.119 And every single LLM responded with that you should walk to the car wash. 00:19:53.279 --> 00:19:56.240 And it's like, you know, who knows how fabricated it was. 00:19:56.319 --> 00:20:12.799 I didn't go and test it, but it's like stuff like that that we're so deep and we're so like entrenched in what AI can do that when you ask it questions like that and take a step back that like the every person is asking, you're sometimes people are like, this is stupid. 00:20:13.759 --> 00:20:16.400 Of course I should drive I need to drive to the car wash. 00:20:16.480 --> 00:20:17.680 Like, what are you talking about? 00:20:17.839 --> 00:20:21.359 And we're like, oh no, that's a logic, that's not a logic question. 00:20:21.440 --> 00:20:27.839 You shouldn't ask an AI that and um we'll see how long it takes to figure out the car wash problem. 00:20:28.240 --> 00:20:32.240 Yeah, that's interesting because it's what you might call an ill-formed question. 00:20:32.319 --> 00:20:35.119 Uh it doesn't have enough information in it. 00:20:35.279 --> 00:20:40.400 Um, and what the AI should say back is why are you going to the car wash? 00:20:40.559 --> 00:20:48.079 Like, do you do you work there and you have to pick up your paycheck or something, if that's how paychecks even work these days? 00:20:48.319 --> 00:20:51.759 Um or are you getting your car actually washed? 00:20:52.000 --> 00:20:53.680 And I'm actually gonna do that right now. 00:20:53.839 --> 00:21:00.640 Um the car wash is five minutes from my house. 00:21:00.960 --> 00:21:04.160 Should I walk there or drive? 00:21:04.480 --> 00:21:08.079 I'm gonna see if these people on Twitter are uh just trying to

pull me. 00:21:09.200 --> 00:21:11.200 I need to get my car washed. 00:21:12.000 --> 00:21:14.559 Like I'm going to the car wash to get my car washed. 00:21:14.720 --> 00:21:16.079 Should I walk or drive? 00:21:16.559 --> 00:21:17.599 Oh, interesting. 00:21:17.759 --> 00:21:26.079 Because I remember seeing it and I thought that that part wasn't in it, but there's a good chance that I just there's a good chance that I just didn't read it very carefully. 00:21:26.400 --> 00:21:28.000 Well, what is what is your what is your prompt? 00:21:28.319 --> 00:21:31.119 Let's uh let's let's live test the AI. 00:21:31.599 --> 00:21:32.559 Ha ha ha. 00:21:32.960 --> 00:21:34.720 The answer was drive. 00:21:34.799 --> 00:21:36.240 You need the car there anyway. 00:21:36.559 --> 00:21:38.000 Hey, there you go. 00:21:38.319 --> 00:21:38.559 Yeah. 00:21:38.880 --> 00:21:39.680 Who were you asking? 00:21:39.759 --> 00:21:40.880 Or what are you asking? 00:21:41.440 --> 00:21:45.119 Um Claude Opus46, I believe. 00:21:45.359 --> 00:21:46.240 Yeah, okay. 00:21:46.640 --> 00:21:47.279 Yeah. 00:21:47.680 --> 00:21:48.720 Yeah, interesting. 00:21:48.799 --> 00:21:54.799 Yeah, and I wouldn't put it past these fucking Twitter people to uh to just give us something fake. 00:21:55.039 --> 00:21:55.440 I know. 00:21:55.680 --> 00:21:57.680 That's the other thing about all this AI stuff. 00:21:57.839 --> 00:21:59.759 There's so much BS out there. 00:22:00.880 --> 00:22:01.519 So much. 00:22:02.400 --> 00:22:03.759 It's it's it's wild. 00:22:04.079 --> 00:22:06.799 It's better just to not pay attention at all. 00:22:07.200 --> 00:22:14.720 Anyway, um, I'm curious about this product that you're working on, Purchase Kit, and how long have you been working on it? 00:22:14.880 --> 00:22:15.839 What is it? 00:22:16.079 --> 00:22:21.200 How are you involving AI and not involving AI and all that stuff? 00:22:21.680 --> 00:22:31.680 Yeah, so Purchase Kit is a full solution to uh in-app purchases for your iOS and Android apps. 00:22:32.000 --> 00:22:45.599 So you have an iOS or an Android app in the App Store, let's say it's built with Hotwire Native, and you want to start charging people subscriptions or in app, you know, like consumables, you need to use in-app purchases to follow Apple and Google's guidelines. 00:22:45.759 --> 00:22:48.400 And they're just like a pain in the butt to implement. 00:22:48.640 --> 00:22:50.960 They don't interact well with your Rails app. 00:22:51.200 --> 00:22:56.079 The data lives on Apple's servers or Google's servers, so you never know who's subscribed and when. 00:22:56.240 --> 00:23:08.799 And Purchase Kit will do all of this for you with like a series of different packages and endpoints and SaaS tools, uh, all and then put all the data of the subscribers in your Rails database. 00:23:08.960 --> 00:23:11.920 So you can do the queries on like how many people are subscribed. 00:23:12.079 --> 00:23:13.920 Okay, this customer, what plan are they on? 00:23:14.079 --> 00:23:16.079 Should they be show them the paywall or not? 00:23:16.240 --> 00:23:21.279 And it all lives in your Rails database instead of distributed across three different servers. 00:23:21.599 --> 00:23:26.640 So is this specifically for people who have a Rails app and a mobile app? 00:23:26.880 --> 00:23:27.440 Exactly. 00:23:27.599 --> 00:23:27.839 Yeah. 00:23:28.079 --> 00:23:28.400 Got it. 00:23:28.640 --> 00:23:41.440 It is a gem, an iOS package, an Android package, and then like a SaaS that sits on top of it that connects the platform I call it, that connects all the moving pieces together. 00:23:41.759 --> 00:23:51.359 And the platform itself is like, you know, dealing with webhooks from Apple and Google, passing webhooks along to your Rails server, writing stuff to your database via the gem. 00:23:51.839 --> 00:24:02.799 It's a lot of cord, it's like a coordination tool that means you can add in-app purchases to your app with like four or five lines of Ruby and no native code. 00:24:02.960 --> 00:24:16.880 Instead of hundreds of lines of Ruby, you know, hundreds of lines of iOS, hundreds of lines of Android, um, and all with the stuff that makes it robust, like exponential back-off for webhook failures. 00:24:16.960 --> 00:24:20.480 And if whose server goes down, how do we resolve that? 00:24:20.640 --> 00:24:24.960 And webhooks out of order, like all that complicated stuff you don't want to have to deal with. 00:24:25.119 --> 00:24:36.240 It's it's what I've built for clients, like four or five times, and this is the productized version of that to make it easy for anyone to add in app purchases in like an afternoon. 00:24:36.640 --> 00:24:39.440 And where are you at from a business perspective? 00:24:39.599 --> 00:24:43.839 Are are you working on getting your first customer or are you already past that point? 00:24:43.920 --> 00:24:44.720 Where are you at there? 00:24:45.200 --> 00:24:47.279 Yeah, working on getting my first customer. 00:24:47.359 --> 00:24:57.279 Um, outside of like me using it for my I don't call it internal tools, but like my public-facing stuff will use Purchase Kit to do a net purchases. 00:24:57.359 --> 00:25:00.480 So it's being used, it's just like not no one's paying for it yet. 00:25:00.640 --> 00:25:10.480 Uh, which has been really interesting because almost all of my products in the past are very quick, like, oh, I need that thing, let me buy it. 00:25:10.720 --> 00:25:12.960 And then I'll use it later. 00:25:13.119 --> 00:25:19.839 Like my bridge component, I have a library that adds native features to your Rails app for running on iOS and Android. 00:25:19.920 --> 00:25:22.000 And it's like a one-time purchase, a couple hundred bucks. 00:25:22.079 --> 00:25:26.079 It's if you need those features, you buy the thing and you you integrate it later. 00:25:26.400 --> 00:25:34.480 Purchase Kit is very much like a long sales cycle, which I never really thought about when I was building it.

00:25:34.799 --> 00:26:01.279 Where I've had onboarding calls with multiple people where we've like gotten it set up in their app and it's working in the sandbox, and now they're ready to buy, but they are working with Apple to like figure stuff out, to figure out uh where the in app how much the in-app purchases should cost, which regions it should be available in, like all of this stuff that is outside of the product that like slows down the sales cycle. 00:26:01.440 --> 00:26:02.799 Um, and I'm impatient. 00:26:02.880 --> 00:26:05.519 Like, I just want someone to hit buy and start paying me. 00:26:05.839 --> 00:26:08.799 A, I want to validate it before before I put more time into it. 00:26:08.880 --> 00:26:12.880 But B, it's like, come on, you know, move faster, people. 00:26:13.039 --> 00:26:21.359 And and then this is this is new to me because like I said, everything that I have ever built has always been this like they see it, they buy it, they use it in like a day. 00:26:21.519 --> 00:26:26.000 And purchase kit has been it takes a couple weeks for someone to even consider it. 00:26:26.079 --> 00:26:33.920 And then like I have to have an onboarding call and then a sales call, and it's like, is this what I want to be building and spending my time on? 00:26:34.160 --> 00:26:40.720 Um, you know, if I was at 10k MRR, that's a different story, but like to get it off the ground has uh has been has been tough. 00:26:40.880 --> 00:26:41.680 Has definitely been tough. 00:26:41.920 --> 00:26:43.119 Yeah, interesting. 00:26:43.440 --> 00:26:50.480 And are are there existing products that are kind of analogous or is this like a unique novel thing? 00:26:50.799 --> 00:26:53.119 There are unique there are uh existing products. 00:26:53.200 --> 00:26:55.839 Uh Revenue Cat is the big one. 00:26:56.000 --> 00:27:03.440 Uh the difference is that all of the tools around in-app purchases are focused on mobile developers. 00:27:03.680 --> 00:27:08.640 They're like, make in-app purchases easy from a mobile developer perspective. 00:27:08.960 --> 00:27:13.519 So what the the features that they offer are like dynamic paywalls. 00:27:13.680 --> 00:27:19.200 Oh, go on our website and configure your prices and like show a paywall that's all native code. 00:27:19.440 --> 00:27:22.720 There's nothing for Rails developers except Purchase Kit. 00:27:22.880 --> 00:27:26.880 And the big distinction is that the feature set is very different. 00:27:27.119 --> 00:27:38.079 To build a dynamic paywall where you're showing different prices to different users, you render an HTML page to your customers, just like you do in a pricing experiment that would link to Stripe. 00:27:38.240 --> 00:27:40.079 That problem is not something I need to solve. 00:27:40.240 --> 00:27:41.839 Rails does that for us. 00:27:42.079 --> 00:27:49.200 But the problems that do exist are like how does the Rails app tell the iOS and the Android app to make a purchase? 00:27:49.440 --> 00:27:54.240 How does

the iOS app tell the Rails app that the purchase was made and like validate receipts? 00:27:54.319 --> 00:28:04.400 So I'm solving very different problems because it's built for Rails backends and Rails developers, not for iOS and Android developers. 00:28:06.960 --> 00:28:19.440 Is there anything that you could imagine something you could learn at some point that would make you decide, nope, I'm I'm not gonna pursue this this product, I'm gonna go do something else? 00:28:21.039 --> 00:28:32.640 I think that so I have a a lead right now that wants to use Purchase Kit, but is migrating would be migrating off of Revenue Cat, which is really exciting. 00:28:32.720 --> 00:28:39.920 They're doing like 40k a month into that purchases, so it's it's non-trivial, you know, it is their business. 00:28:40.319 --> 00:28:51.759 And what would make me not pursue this anymore is if when we get down to the migration, if I realize that I have overlooked 90% of the surface area. 00:28:52.240 --> 00:29:02.319 Like, sure, this product works for the clients that I've worked with in the past, but actually, when it becomes a SaaS and it's self-serve, there are all these other problems I didn't think about. 00:29:02.559 --> 00:29:05.119 That is where I probably would throw in the towel. 00:29:05.519 --> 00:29:09.759 And they're not code, they're not answers that can be solved with code. 00:29:09.839 --> 00:29:19.599 It's like, oh, well, you know, we have to deal with compliance, and that's gonna require our a custom contract, and like we have to deal with where the data lives. 00:29:19.680 --> 00:29:26.720 It's like all of those things that I like don't want to deal with, that's where I would say, all right, you know what, this isn't gonna, this isn't gonna work. 00:29:26.960 --> 00:29:31.200 Um I it's gonna take a while to get there though, because I need to talk to a lot of people. 00:29:31.759 --> 00:29:40.000 I feel pretty good about it because like again, I've built this solution for five, six clients in the past, and this is the exact architecture that they follow. 00:29:40.079 --> 00:29:41.680 It's just generalized. 00:29:42.160 --> 00:29:44.480 Um but I was always in there with the code. 00:29:44.559 --> 00:29:46.960 If something broke, I could change their Rails code. 00:29:47.119 --> 00:29:53.119 Now I only have a gem to deploy, and I have to rely on people updating their gem, like for security updates and stuff. 00:29:53.279 --> 00:29:55.680 So it's a very different problem. 00:29:56.880 --> 00:29:58.240 Yeah, interesting. 00:29:58.480 --> 00:30:08.240 Um with Saturn CI, there's a there's kind of a comforting aspect of it where it's not a novel kind of product. 00:30:08.400 --> 00:30:08.720 Yeah. 00:30:08.880 --> 00:30:10.559 It's just a CI platform. 00:30:10.799 --> 00:30:16.799 If you know about GitHub Actions and Circle CI and all that, you understand what my product is. 00:30:17.039 --> 00:30:30.559 And there's the the comforting thing is if I'm not getting sales, that can only mean one thing, and that can only mean that I'm just not doing good enough at it yet. 00:30:31.440 --> 00:30:32.000 Yeah. 00:30:32.400 --> 00:30:41.759 Um it it can't be because the product is the wrong thing or something like that, because there's already people being successful with CI products. 00:30:41.920 --> 00:30:42.160 Right. 00:30:42.240 --> 00:30:47.440 Um so it can only be a lack of effort or doing the right things and stuff like that. 00:30:47.599 --> 00:31:04.640 But I've worked on products before where it was a novel idea, and it's like tormenting because it's like, am I not getting sales yet because I'm just not doing the right stuff, or is it because this product is doomed to failure and nothing I could ever do could make it work? 00:31:05.119 --> 00:31:06.799 Yeah, that's a that's a tough one. 00:31:06.960 --> 00:31:18.079 And now that you can build one can build things so much more quickly, it it becomes this trap of, well, this didn't work for like three weeks, so should I just try something new? 00:31:18.400 --> 00:31:20.480 Versus, well, no one even knew about it. 00:31:20.559 --> 00:31:23.440 You had a hundred people check out your website and you gave up, you know. 00:31:23.599 --> 00:31:26.000 So where do you find that balance? 00:31:26.720 --> 00:31:34.079 Um Yeah, I I do want to ask, like, the Saturn CI stuff, uh, are you running on a Linux box? 00:31:34.240 --> 00:31:38.000 Like, what is the actual can you run on a Mac box? 00:31:38.160 --> 00:31:39.680 Like, why how does that all work? 00:31:40.160 --> 00:31:59.039 The runners are hosted, so there's not a way to run your own runners, you just use mine, and the way that works is I just have a giant VM on Hetzner, and it uses Kubernetes for the orchestration, and it runs that way. 00:31:59.359 --> 00:32:00.000 Nice. 00:32:00.160 --> 00:32:01.039 That's really cool. 00:32:01.200 --> 00:32:07.119 That's that's still something that I have not experimented with at all, like Kubernetes, Hetzner. 00:32:07.359 --> 00:32:12.640 I mean, with the recent Heroku news, I'm I'm I that might change soon. 00:32:12.799 --> 00:32:19.119 Um I'm slowly migrating stuff off of Heroku, but I'm just going to render because it's the most familiar. 00:32:19.279 --> 00:32:35.839 Uh but I'm spinning up a new product right now that I need a Rails website and a demo server, and like those are 25 bucks plus a month each, just to have a Rails app running before it hits the memory issue, you know? 00:32:36.160 --> 00:32:41.599 And having those on a single Hetzner box feels a lot better. 00:32:41.759 --> 00:32:52.960 I just like haven't spent the time to learn all of the stuff that I need to learn and come all and all that, and then it's I'm really just delaying the inevitable because I know that's coming soon. 00:32:53.279 --> 00:32:58.799 Yeah, and obviously uh AI can really grease the wheels for you. 00:32:59.039 --> 00:32:59.440 Mm-hmm. 00:32:59.599 --> 00:33:00.079

Mm-hmm. 00:33:00.640 --> 00:33:02.480 It's like, do I want to do that though? 00:33:02.640 --> 00:33:04.640 That's like that's what all this has come down to. 00:33:04.720 --> 00:33:15.920 It's like AI can do anything for me, but is that what I want to spend my time prompting and validating and testing and managing and maintaining in the long run? 00:33:16.240 --> 00:33:23.359 And having something on render is a very much of a on a on a on a platform as a service, I should say, is a safety blanket. 00:33:24.160 --> 00:33:38.000 Yeah, it's kind of like the this AI stuff with regard to productivity, it's kind of like the industrial revolution where people predicted that people would just start working like a five-hour work week or something like that because you can be so much more productive. 00:33:38.160 --> 00:33:39.440 And that's not what happened. 00:33:39.599 --> 00:33:46.799 People still use up all their available time to work, they just accomplish more during that time. 00:33:46.960 --> 00:33:49.039 And that's exactly what we're seeing with AI. 00:33:49.119 --> 00:33:53.440 It's not as though people are working less, they're working the exact same amount, maybe more. 00:33:54.079 --> 00:33:58.559 Um and they're just producing more output. 00:33:58.720 --> 00:34:00.480 I don't want to necessarily say value. 00:34:00.640 --> 00:34:08.559 Some people are creating more value, some people are just generating a bazillion lines of code, and whether that's more value is is dubious. 00:34:08.880 --> 00:34:12.079 Um, but it's that's interesting to see how that plays out. 00:34:12.639 --> 00:34:15.039 I am 100% falling into that trap. 00:34:15.280 --> 00:34:26.559 I am working so much more the past couple of weeks than I have in the past, and producing so much more, like you said, lines of code. 00:34:26.800 --> 00:34:28.719 Is there the is there value in this? 00:34:28.880 --> 00:34:29.920 I don't know yet. 00:34:30.079 --> 00:34:37.119 I'm excited to find out, but my output has been off the charts the past couple of weeks. 00:34:37.440 --> 00:34:39.599 And I don't know if that's a good thing. 00:34:40.000 --> 00:34:41.039 I'm curious. 00:34:41.199 --> 00:34:46.960 So I had this episode with uh Gail Lackman McDowell. 00:34:47.679 --> 00:34:49.280 Yeah, I think I got that name right. 00:34:49.440 --> 00:35:13.920 Um, the author of uh Cracking the Coding interview um a a few dozen episodes back, and and we talked about the irony that like the job market, the the job application process is kind of destroyed now because people get spammed with AI applications and stuff like that, and you're the real applications just get lost in this sea. 00:35:14.559 --> 00:35:23.679 Um and so it's really hard for job seekers now, and basically what we are having to do is go back to just networking. 00:35:24.559 --> 00:35:32.159 And in this hyper-technological age, uh the personal relationships are actually more important rather than less. 00:35:33.280 --> 00:35:50.159

Um and I think that's true with a lot of stuff, and with Saturn CI, that's that's been my like initial that's like my first layer is like going to first to people I know, people I already have a relationship with, and and going after those people. 00:35:50.320 --> 00:35:59.119 And then once I kind of went through most of those, now I'm like meeting new people and making new relationships and stuff like that. 00:35:59.519 --> 00:36:06.000 I'm not prioritizing like some kind of automated mass marketing and stuff like that. 00:36:06.400 --> 00:36:14.480 So I'm I'm and and I don't necessarily mean like bad stuff, I just mean like blog posts and and that kind of thing. 00:36:14.559 --> 00:36:19.360 Like I'm working on that now, but I only just started that like quite a ways in. 00:36:19.599 --> 00:36:28.719 Yeah, and I'm curious for you, what are you doing for sales and marketing, for your consulting, for your products, whatever you want to talk about? 00:36:28.880 --> 00:36:32.880 Has anything changed for you and now that we're in this different world? 00:36:34.079 --> 00:37:02.880 I think that the approach is the same for me, for my consulting business, and that is share as much as I possibly can, give away as much as humanly possible, paywall the things that are only the highest of leverage and the highest of value or the deepest of you know content, and and hope that pray that inbound leads will come. 00:37:03.519 --> 00:37:08.880 And that is what's worked for me since I've been independent for like going on six years now. 00:37:09.119 --> 00:37:16.239 I think that I've tried cold outreach, I've tried, hey, you know, do you want to do a mobile app or or stuff like that? 00:37:16.400 --> 00:37:20.800 And I don't think I've ever had a single client from that kind of stuff. 00:37:20.960 --> 00:37:30.480 Every single one of my clients has been responding to a newsletter, seeing my blog post, talking to me after a conference talk, uh, hallway track type stuff. 00:37:30.800 --> 00:37:35.519 And I'm starting to realize that I keep fighting that. 00:37:36.000 --> 00:37:46.000 I keep trying to do cult, do outreach, and like find new clients, or like tell the world that I have opportunity to work with them, and like it never works. 00:37:46.159 --> 00:37:47.679 So why do I keep trying? 00:37:47.920 --> 00:37:57.599 So I'm leaning into the fact that if all of my leads are inbound and they find me through X, Y, and Z, do more X, Y, and Z. 00:37:58.239 --> 00:38:00.559 And X, Y, and Z are content. 00:38:00.960 --> 00:38:06.480 So that's making sure I have something interesting to post every single day on Twitter. 00:38:06.800 --> 00:38:11.519 Having a newsletter article on sub on my Substack every single week. 00:38:11.679 --> 00:38:19.280 Uh, I now have a podcast that's coming out every single week, so like getting notifications in Apple Podcasts or Overcast or whatever. 00:38:19.599 --> 00:38:21.440 The book, of course, is huge. 00:38:21.519 --> 00:38:24.880

Uh I try I'm gonna try to speak at a conference at least once a year. 00:38:25.199 --> 00:38:30.159 All of these things I I've been doing forever. 00:38:30.400 --> 00:38:36.159 But now I'm looking at them as these are not one of the ways to make money. 00:38:36.320 --> 00:38:39.440 These are the only ways to bring in new clients. 00:38:39.679 --> 00:38:47.679 So I should be doing if I want to make have a successful business and make more money, I should be doing more of the things that bring in leads. 00:38:47.840 --> 00:38:50.719 Like, you know, when I think about it that way, it's so simple. 00:38:51.039 --> 00:38:54.480 Uh I feel like I I'm just like always trying to find like the easy out. 00:38:54.559 --> 00:38:56.800 It's like, what if I send like 10 cold emails and I get a client? 00:38:56.880 --> 00:38:57.519 That'd be so easy. 00:38:57.599 --> 00:39:01.519 It's like, no, just write, just write an article and you'll get someone in two months. 00:39:02.079 --> 00:39:02.880 Yeah, yeah. 00:39:02.960 --> 00:39:10.079 Well, uh a lot of um developers are neurodivergent, uh autistic, ADHD. 00:39:10.239 --> 00:39:15.199 Uh I think I'm well, I know that I'm autistic and probably ADHD also. 00:39:15.679 --> 00:39:23.199 Um, and I've fallen into this trap, as so many people do, of like, okay, I did something and it worked. 00:39:23.360 --> 00:39:24.320 What do I do next? 00:39:24.400 --> 00:39:26.079 What's what's the different thing that I do? 00:39:26.239 --> 00:39:30.079 I have to do something new and I'm always like chasing novelty. 00:39:30.800 --> 00:39:37.679 Um and I've kind of uh figured out how to uh push back on that and not succumb to it so much. 00:39:37.840 --> 00:39:42.079 It's like, wait, if I did something that worked, I should do that same thing again. 00:39:42.320 --> 00:39:47.280 And it's like not it it doesn't tickle that part of my brain that like has an itch, you know. 00:39:47.519 --> 00:39:57.199 Um but you have to like go I have to go against my natural inclinations, uh which is which is hard. 00:39:57.760 --> 00:40:04.079 I have ADHD, um, was diagnosed not long ago, like only a couple of years ago. 00:40:04.239 --> 00:40:08.320 So it's still uh it's still pretty new to me. 00:40:08.639 --> 00:40:13.039 And the the novelty thing is something that I'm constantly fighting as well. 00:40:13.199 --> 00:40:14.400 It's like this worked. 00:40:14.639 --> 00:40:15.760 Try something new. 00:40:15.920 --> 00:40:17.679 It's not good, keep doing it. 00:40:17.920 --> 00:40:20.880 Um my writing is a really good example of that. 00:40:21.039 --> 00:40:26.159 I cannot write in the same editor for more than like a few months at a time. 00:40:26.400 --> 00:40:28.639 I need to change my editor, and I've leaned into that. 00:40:28.719 --> 00:40:30.400 So it's like, okay, cool, I have seven editors. 00:40:30.480 --> 00:40:36.239 I have Typora, I have MacVim, I have notes, I have text edit, and it's like I just cycle through those. 00:40:36.320

--> 00:40:38.000 I write in Substack directly. 00:40:38.159 --> 00:40:43.440 I cycle through those every couple of months because it's novel and it feels new and exciting. 00:40:43.599 --> 00:40:45.760 Right now I'm on a voice recording kick. 00:40:45.840 --> 00:40:55.280 So when I'm out on a walk, I have my little mic, you know, hooked up to my chest and I record the content from my blog post, and then I transcribe that, and that's my first draft. 00:40:55.440 --> 00:40:57.920 Because that is novel and exciting to me right now. 00:40:58.079 --> 00:41:00.639 In another week, I'll be back to pen and paper. 00:41:00.719 --> 00:41:01.920 Like, who knows? 00:41:02.159 --> 00:41:11.840 Um, and those are the things that I can switch up that like are okay to have the novelty because the end result is the same, it's a polished piece of work. 00:41:12.159 --> 00:41:22.480 But the novelty in where my focus lies or what is working for my business, those things that are working need to be doubled down on. 00:41:22.559 --> 00:41:30.480 And it's so easy to say and so hard to do that, yeah, end of sentence. 00:41:30.719 --> 00:41:32.320 Like that's all I got. 00:41:32.800 --> 00:41:34.159 Yeah, it's so hard. 00:41:34.400 --> 00:41:40.000 Um, and I've been trying to get some blog posts out on the Saturn CI website for so long. 00:41:40.159 --> 00:41:42.960 I think I have like one blog post finished. 00:41:43.119 --> 00:41:53.599 I used to crank out a blog post like every week, um, but for some reason it's harder now, and maybe it's harder because it's uh it's connected to a product. 00:41:53.760 --> 00:41:57.440 It's not just like writing a blog post because of whatever. 00:41:57.679 --> 00:42:09.920 Um, I feel like it I maybe I'm holding myself to a higher standard because I feel like it has to be good because it has to um drive traffic. 00:42:10.480 --> 00:42:19.760 I'm I'm not sure what the hang-up is, but I keep starting blog posts, and then I'm like, oh no, this is like too ambitious and broad scoped. 00:42:19.920 --> 00:42:25.360 Let me let me try something that I can finish within a reasonable amount of time instead of like three months. 00:42:25.599 --> 00:42:29.199 Um and then I start a new one and I'm like, oh no, this topic sucks. 00:42:29.599 --> 00:42:30.880 Start a new one. 00:42:31.199 --> 00:42:37.679 It's uh pretty it's pretty hard, but yeah, I'll I'll I'll get through it. 00:42:38.159 --> 00:42:42.800 I think that that is part of the writing process that is that is good friction. 00:42:43.679 --> 00:42:47.519 Uh it's something that you could have published that first post. 00:42:47.599 --> 00:42:51.920 You could have like wrapped it up in a crappy way and and not actually hit your point. 00:42:52.159 --> 00:43:06.880 And is having two blog posts on the blog now better than having like two, or sorry, having four now that are all kinda meh versus having two that are really good and really get your point across in a couple of months? 00:43:07.039

--> 00:43:12.559 Like, you're with blogs you're and and content like that, you're you have to play the long game. 00:43:12.800 --> 00:43:15.360 Writing a post is not going to bring you a customer in a week. 00:43:15.599 --> 00:43:16.960 This is just not how it works. 00:43:17.119 --> 00:43:19.679 It's gonna bring you a customer in like six months. 00:43:19.920 --> 00:43:26.880 And if you're looking at that timescale, it doesn't matter if you publish now or in two weeks or in even three weeks. 00:43:27.119 --> 00:43:30.639 Like earlier the better, yes, but to a degree. 00:43:31.039 --> 00:43:39.039 And especially in the age of like generated content, like quality is so much more important than quantity. 00:43:39.360 --> 00:43:42.880 So I'm not like I think that's good friction. 00:43:43.280 --> 00:43:58.800 That's another thing that makes it harder, uh, because it used to be that you could just write a tutorial on some like here's or or or just like here's an explanation for how this uh configuration setting in Capi Barrel works or something like that. 00:43:59.039 --> 00:44:01.039 But now it's like Why would I read that? 00:44:01.119 --> 00:44:03.360 I can just ask AI when I need it. 00:44:03.519 --> 00:44:06.719 Um, so that kind of content doesn't make as much sense anymore. 00:44:06.800 --> 00:44:22.800 What does make sense, of course, is like opinions, things you you like take a stand on something and say, like, hey, I think this uh that that's the kind of stuff that's a lot more valuable that you're not gonna get from an AI, at least not yet. 00:44:23.679 --> 00:44:33.840 I I'm a sadly uh to admit, like I'm on Hacker News probably ever I'd say every day, and I and I always click through five, six articles a day. 00:44:34.159 --> 00:44:38.960 I finish close to zero of those articles that I click through on. 00:44:39.199 --> 00:44:45.440 And it's mostly because they feel like such an endeavor to work through. 00:44:45.599 --> 00:44:52.639 They're so long, they're like, you know, 2,000 words or something, and yes, attention and all that stuff, you know. 00:44:52.960 --> 00:45:04.079 But I read a couple posts, or I read a post the other day that was, I think like 250 words, and it was like there was no preamble, there was no like, hey, I'm this person writing about all this stuff. 00:45:04.159 --> 00:45:08.719 It was just like, here's what I'm thinking, here's why that worked, and here's why it maybe didn't work. 00:45:08.800 --> 00:45:10.159 And then it just ended. 00:45:10.400 --> 00:45:21.280 And it was so refreshing to read an article without having to like take a break from it and like check Twitter or just like, you know, attention go, you know, things going off, or like whatever. 00:45:21.440 --> 00:45:24.320 I was able to just like sit down and read that article start to finish. 00:45:24.400 --> 00:45:26.800 And I left and was like, oh, interesting take. 00:45:27.039 --> 00:45:28.239 And I moved on with my life. 00:45:28.320 --> 00:45:35.280 And I feel like we're moving towards that where it's like a hundred percent human, the articles like that. 00:45:35.440 --> 00:45:39.039 Here's my opinion on something, here's some thoughts about it, end. 00:45:39.440 --> 00:45:46.880 And it's not 2,000 words uh of a preamble that at the end is really just like, oh, and by the way, you should buy my product. 00:45:48.079 --> 00:45:50.480 And I'm curious to see how that works out. 00:45:50.880 --> 00:45:52.159 Yeah, interesting. 00:45:52.400 --> 00:45:56.639 I have a handful of posts like that, probably a good amount of posts like that, actually. 00:45:56.880 --> 00:45:57.199 Yeah. 00:45:57.440 --> 00:46:07.920 Um, but it's okay, so my target is to to for to to write my blog posts and then post them to the Ruby subreddit. 00:46:08.800 --> 00:46:14.639 Um, because that is that is where they can become visible and get traffic. 00:46:14.800 --> 00:46:24.320 And I know that Peter Cooper looks at the Ruby subreddit to get his content for Ruby Weekly, and so that often gives me a boost. 00:46:24.639 --> 00:46:38.719 Um and so when I start to write a blog post, I'm like, okay, what can I write that will be sufficiently relevant to Ruby that I can post to the Ruby subreddit and not get yelled at by the moderators and stuff like that. 00:46:39.039 --> 00:46:39.679 Yeah. 00:46:40.000 --> 00:46:40.159 Yeah. 00:46:40.559 --> 00:46:43.280 And it is it's hard. 00:46:43.599 --> 00:46:50.880 Because most of the most of the thoughts I have are like technology agnostic principles. 00:46:51.360 --> 00:46:51.679 Right. 00:46:52.719 --> 00:47:02.480 And so I I I feel like I almost have to uh do something contrived in order to make it relevant to Ruby or Rails or whatever. 00:47:02.639 --> 00:47:04.400 That's that's part of my friction. 00:47:05.920 --> 00:47:22.400 I think that a lot of that comes also down to what your goal of those content pieces is, and being more explicit with yourself about is this a blog post to raise folks' awareness and become more comfortable with Jason? 00:47:23.039 --> 00:47:33.440 And maybe that leads to trusting him more, and then in the future I will hire him as a consultant or an advisor, or is this like I'm trying to build my audience? 00:47:33.599 --> 00:47:38.559 Or is this I'm trying to sell and become make folks aware of my product? 00:47:38.960 --> 00:47:43.920 And those are very different goals, and I know I struggle with this a lot. 00:47:44.000 --> 00:47:52.000 It's like I have an article and I'm like, okay, I want this to bring in clients, but also make people aware of this product and also trust me. 00:47:52.079 --> 00:47:54.960 And it's like, if I try to do all of those, the article flops. 00:47:55.039 --> 00:48:04.079 But if I try to just like focus on one, or or at least keep my guardrails up, that's like the point of this article is to just build some trust around my name. 00:48:04.320 --> 00:48:12.400 There's no CTAs, there's no like links anywhere, it's

just like here's what I'm thinking, maybe someone will read this and resonate with it. 00:48:12.639 --> 00:48:24.320 Those are the articles that I have the most fun writing, and those are the articles that usually end up doing like the rounds more so than the ones where I have a specific ask in them. 00:48:25.599 --> 00:48:29.519 Yeah, for me it's a hundred percent for Saturn CI. 00:48:30.239 --> 00:48:43.360 Um, and I don't need anything terribly specific out of it, other than people to land on the site and look up at the URL and see SaturnCI.com and maybe be like, oh, what's what's this website? 00:48:43.440 --> 00:48:44.639 What's this thing all about? 00:48:45.199 --> 00:48:46.559 Yeah, that's all I'm shooting for. 00:48:46.639 --> 00:48:48.480 It's a pretty low bar. 00:48:49.039 --> 00:49:06.719 Um yeah, and a big part of it is just attention, spending time and attention on the thing and not letting yourself get distracted by again novelty, just making yourself focus and do the thing. 00:49:07.199 --> 00:49:07.679 Mm-hmm. 00:49:08.239 --> 00:49:08.800 Yeah. 00:49:09.119 --> 00:49:11.119 How uh how is Saturn CI going? 00:49:11.199 --> 00:49:16.800 Like, are you you have you have folks that are using it, folks that are paying for it and stuff? 00:49:16.880 --> 00:49:18.480 Like where where are you in that journey? 00:49:18.800 --> 00:49:20.960 I have a folk who's paying for it. 00:49:21.360 --> 00:49:21.920 Nice. 00:49:22.159 --> 00:49:38.800 Um, yeah, and then another so that is a solo developer, um, and then I have a an actual business um that I used to, I mean, not to not to make it I I don't want to say that this paying customer is not an actual business. 00:49:39.039 --> 00:49:46.800 Um but I have like a bigger, more established, uh older business um getting onboard onboarded onto it also. 00:49:47.039 --> 00:49:49.119 It's a company I used to work for. 00:49:49.519 --> 00:49:55.360 Um and it's also, at least in this particular case, an extremely long sales cycle. 00:49:55.840 --> 00:50:02.320 It's been in the works for I don't know, four or five, six months, something like that at this point. 00:50:02.719 --> 00:50:06.320 And it's kind of a high switching cost and all that stuff. 00:50:06.639 --> 00:50:07.199 Oh yeah. 00:50:07.519 --> 00:50:09.760 So that's that's taking a long time. 00:50:10.000 --> 00:50:23.679 Um, and then in parallel to that, I'm just talking with as many people as I can and trying to get people to to um try it out, and I'm doing demos for people and stuff like that. 00:50:24.239 --> 00:50:31.519 Um so yeah, it's going about as I was gonna say as well as one could hope. 00:50:31.599 --> 00:50:41.280 It's I could hope for a lot more, um, but it's going decent it as decent as I could reasonably expect, maybe is the way to put it. 00:50:42.480 --> 00:50:48.239 Yeah, and you're not relying on that to be your sole source of income to support. 00:50:48.800 --> 00:50:49.599 Absolutely not. 00:50:49.920 --> 00:50:50.239 Yeah. 00:50:50.400 --> 00:50:58.880 So there's like less pressure to to make that sale and then beg someone for money and stuff when you're doing your your your your main revenue stream outside of it. 00:50:59.599 --> 00:51:00.000 Right. 00:51:00.159 --> 00:51:05.840 I've mentioned many times that one of my principles is don't bet more than you can afford to lose, because I've done that before. 00:51:06.079 --> 00:51:14.239 But in this case, there's no time pressure other than the like psychological time pressure that I choose to impose on myself. 00:51:14.639 --> 00:51:15.360 Right, right. 00:51:15.519 --> 00:51:16.719 Yeah, that makes sense. 00:51:17.039 --> 00:51:34.880 I feel like with folks getting less and less or more and more uh disenfranchised with GitHub in general, uh having an alternative for running actions and CI and stuff uh is more and more appealing to people. 00:51:35.360 --> 00:51:42.559 Um it's the switching cost that I imagine is probably the hardest part of all of it for potential customers. 00:51:42.880 --> 00:51:52.000 Yeah, yeah, and there's kind of an aspirationslash intention that I have of gaining sort of a critical mass, at least in the Ruby community. 00:51:52.320 --> 00:51:54.159 Because I feel like you've done that. 00:51:54.400 --> 00:52:03.039 Like in in general, in the Ruby community, say you if if you say to somebody like, hey, who's the hot wire native guy? 00:52:03.199 --> 00:52:05.679 They're gonna be like Joe Masolotti, duh. 00:52:06.320 --> 00:52:07.920 Like a lot of people know that. 00:52:08.159 --> 00:52:11.760 Um, and even if they say, like, hey, who's the Rails testing guy? 00:52:11.840 --> 00:52:13.360 They'll they'll say me. 00:52:13.920 --> 00:52:20.079 Um but what I don't have yet is uh Saturn CI having the recognition like that. 00:52:20.159 --> 00:52:35.440 Like I'm curious, uh, if if I ask you, Joe, um what's the like prominent Rails uh admin panel app that's been like do you do you know which one I'm talking about? 00:52:35.679 --> 00:52:37.840 Yeah, like Avo and Yeah, yeah, exactly. 00:52:38.159 --> 00:52:38.639 Yeah, yeah. 00:52:38.800 --> 00:52:56.559 So Adrian has done a a great job of like getting that critical awareness of Avo in the in the Rails community, and so I feel like that is something really good to aspire to, to have that it it's kind of like a binary thing, like either you have it or you don't, almost. 00:52:57.039 --> 00:53:02.719 You could view it either way, but yeah, I feel like that's something worth having. 00:53:04.559 --> 00:53:05.519 I agree. 00:53:05.760 --> 00:53:12.480 Uh I am cautious about that in general. 00:53:12.800 --> 00:53:15.280 Um, and I'll give you two anecdotes here. 00:53:15.519 --> 00:53:23.119 I am known as the Hotwire Native guy right now, which has been great for my branding and positioning as I'm doing Hotwire Native projects. 00:53:23.440 --> 00:53:31.519 I am starting to level up from that and trying to be like the mobile app guy in the Rails

ecosystem. 00:53:31.840 --> 00:53:39.199 And I'm slowly starting to release, you know, more high-level content, more around mobile apps, less around Hotwire Native specifically. 00:53:39.360 --> 00:53:46.480 Hotwire Native is still a tool that I will reach for, but like if your company wants to do React Native, great, I can do React Native. 00:53:46.559 --> 00:53:50.159 I'm not gonna say no to it just because it's not hotwire native, right? 00:53:50.639 --> 00:53:57.199 And breaking out of that mold, I know will be challenging and is already starting to be challenging. 00:53:58.239 --> 00:54:03.119 What I so that's just like one thing that has been interesting for me. 00:54:03.599 --> 00:54:12.559 What I caution you in is if you're trying to change like Jason Sweat's byline, right? 00:54:12.719 --> 00:54:26.159 Which is essentially, you know, the Rails testing guy to be the Saturn CI guy, you better be confident that Saturn CI is going to be what is bringing you the most revenue and going to be the core of your business. 00:54:26.400 --> 00:54:44.559 Like, if you lose the the test, the Rails testing guy, and you are the Saturn CI guy, but Saturn CI is bringing you in 2K a month, and you no longer are getting consulting gigs or like getting job offers that you want because you're no longer known as the Rails testing guy, that is scary to me. 00:54:44.719 --> 00:54:47.920 Like I feel unc I feel anxious about that for you. 00:54:48.239 --> 00:54:57.360 And it's something it's so it resonates so in like my core because I'm always struggling with consulting versus products. 00:54:57.760 --> 00:55:05.119 And consulting pays the bills, but products are so fun and they energize me and they bring me so much joy. 00:55:05.360 --> 00:55:12.079 And how do I balance like what do I share versus who I am versus how I make money in the world? 00:55:12.320 --> 00:55:17.840 Um so that I I that's a little bit of a ramble, but like, do you do do you get what I'm saying there? 00:55:17.920 --> 00:55:18.960 Am I getting my important thing? 00:55:19.119 --> 00:55:20.000 Absolutely. 00:55:20.800 --> 00:55:21.360 Yeah. 00:55:21.679 --> 00:55:26.960 Well, luckily, you know, you and I haven't talked in a little while, so you might not be aware. 00:55:27.199 --> 00:55:31.519 I I sold my soul and I now have a job at a big company. 00:55:31.679 --> 00:55:38.800 And so I I don't need to worry about consulting anymore or being perceived as the Rails testing guy or anything like that. 00:55:39.280 --> 00:55:49.840 But right now, like that's just right now, and you've had how many years of of building that persona up that is it is like just hand in hand with with your name? 00:55:50.079 --> 00:55:52.880 And you know, jobs are great, but they don't last forever. 00:55:53.039 --> 00:55:54.880 And what if you do want to move back into consulting? 00:55:54.960 --> 00:55:59.599 And now you're like, you're only not only, but like Saturn CI is what you're not. 00:55:59.760 --> 00:56:04.559 You know, it's just like I always look at careers as a very long-term game. 00:56:04.719 --> 00:56:17.199 Uh, like maybe I get a full-time job next year for a little while, maybe I continue to do consulting, maybe I go back to it, but um jobs are easier to change than it is around like your whole persona online, I think. 00:56:17.519 --> 00:56:18.559 Very good point. 00:56:20.000 --> 00:56:36.960 Yeah, and that's another question of like if you have a product, to what degree do you want to attach your name to that and have those two things be kind of synonymous versus like getting the name of the product out there and you don't really worry about whether your name's so connected to it or not? 00:56:37.119 --> 00:56:38.880 I I think it's kind of inevitable. 00:56:39.039 --> 00:57:00.480 Like, yeah, I you know, I'm just guessing, but who knows if Adrian Marin like uh consciously chose to like have his name out there, Adrian Marin, rather than just trying to get Avo out there just because it happens to be Adrian giving the talks at conferences and stuff like that, like naturally we're gonna be like, oh yeah, Adrian. 00:57:00.719 --> 00:57:06.079 Um, so maybe it comes kind of comes with the territory or whether whether that's your intention or not. 00:57:06.320 --> 00:57:06.800 Mm-hmm. 00:57:07.360 --> 00:57:25.280 Yeah, it's a tough one, and it's one of those existential things where it's like when it's tied to your identity, even if you pretend that it's only your work identity, I still feel that identity when I'm not working and when I'm out or like at non-tech meetups, it's like, well, what do you do? 00:57:25.360 --> 00:57:29.039 Well, oh, I'm the hot wire native guy in the Rails community, and they're like, Well, what the hell does that mean? 00:57:29.280 --> 00:57:35.599 It's because I have like that is my identity, even if it is only supposed to be my career identity. 00:57:36.079 --> 00:57:45.920 So I don't know where I don't know like what the point is of that, but it's just it's a it's a it's hard, it's easy to conflate the two, I guess. 00:57:46.320 --> 00:57:54.000 Yeah, yeah, and for me it's it's kind of hard to decide what I want my identity to be. 00:57:54.159 --> 00:57:57.920 I've I've been working on my next book for a long time now. 00:57:58.719 --> 00:58:06.639 I'm I'm now calling it uh software design from first principles, and it's gonna have nothing to do with Ruby or Rails. 00:58:06.880 --> 00:58:09.679 And it's like, where does that fit into everything? 00:58:10.159 --> 00:58:14.880 Uh maybe it doesn't, but maybe I don't get to pick, you know? 00:58:15.440 --> 00:58:22.079 Maybe that that goes out there and if that book, well, if if the book is not successful, then who cares? 00:58:22.159 --> 00:58:24.000 Because it's not gonna change anything. 00:58:24.480 --> 00:58:28.880 But if it takes off, then it's like, oh, now people know me primarily for this book. 00:58:29.039 --> 00:58:31.920 I didn't exactly pick

that, but okay. 00:58:32.800 --> 00:58:39.440 Yeah, letting the the the masses decide who you are versus like setting in stone for yourself. 00:58:39.679 --> 00:58:41.599 I I think it's always a mix of both. 00:58:41.760 --> 00:58:42.320 I really do. 00:58:42.400 --> 00:58:45.360 I don't think you can have it all one way or the other. 00:58:45.519 --> 00:58:48.960 Like I think that you're you want to be the Saturn CI guy. 00:58:49.199 --> 00:58:52.239 Well, the your audience has to uh agree with that. 00:58:52.320 --> 00:58:58.239 Like I'm trying to be the mobile app guy, like maybe I'll always be known as the Hot Wire Native Guy, no matter how much I try. 00:58:58.400 --> 00:59:02.800 It's always gonna be a push and pull of those identity things. 00:59:02.960 --> 00:59:04.960 Like it's it's it's culture, right? 00:59:05.119 --> 00:59:12.800 It's not yeah, you can have it in your byline, but it's really the culture of what people reference you and call you and think about you when they when they when you hear your name. 00:59:13.519 --> 00:59:16.000 Um unfortunately, we're just about out of time. 00:59:16.159 --> 00:59:18.800 I want to make sure to mention one other thing first. 00:59:18.960 --> 00:59:26.639 Um, RubyConf 2026 is happening July uh 14th through 16th. 00:59:26.719 --> 00:59:27.760 I should know these dates. 00:59:27.840 --> 00:59:31.199 I'm the co-chair of the uh RubyConf 2026. 00:59:31.280 --> 00:59:33.760 Uh it's happening in July in Las Vegas. 00:59:34.000 --> 00:59:38.079 Um I don't know if you're thinking about going, Joe, but I I I hope so. 00:59:38.239 --> 00:59:42.480 And by the way, our CFP has just recently opened. 00:59:42.639 --> 00:59:51.840 Um and so, dear listener, if you're interested in uh submitting a proposal, Joe, if you're interested in submitting a proposal, our CFP is open now. 00:59:52.079 --> 00:59:58.320 Um, but in any case, I hope, Joe, that we can hang out at a conference in person sometime soon. 00:59:58.480 --> 01:00:04.559 Um before we go, anything you want to share with people in terms of links and stuff like that? 01:00:05.199 --> 01:00:10.000 Uh yeah, I first off, I I will be submitting a proposal. 01:00:10.400 --> 01:00:12.800 I'm very excited to make it down to Vegas. 01:00:12.880 --> 01:00:17.039 I'm in Portland, Oregon, so it's cool to have another West Coast conference. 01:00:17.440 --> 01:00:29.119 Um and I'll be submitting something in the you know, mobile app, Rails, Ruby space, still kind of figuring out what that's gonna be, um, if it's hotwire native or if it's adjacent or something. 01:00:29.199 --> 01:00:31.920 Um, so I'll definitely be submitting something soon. 01:00:32.239 --> 01:00:37.119 And in terms of links, uh, I have everything on masilotti.com. 01:00:37.280 --> 01:00:43.679 I got my newsletter, my podcast, uh, the hotwire native stuff, my book, all the products we talked about. 01:00:43.760 --> 01:00:46.960 That's all there, and um that's the best place to reach me. 01:00:47.199 --> 01:00:47.760 Excellent. 01:00:47.840 --> 01:00:52.079 Uh well, Joe, we'll put that stuff in the show notes, and thanks so much for coming on the show. 01:00:52.320 --> 01:00:53.599 Yeah, thanks for having me, Jason. 01:00:53.760 --> 01:00:55.440 Good uh fun talk as always.