Why I Ditched Make.com and N8N for GCP (you're not that guy pal)

Hey there, I'm Parker Rex. I just recorded this, but my microphone messed up. So, we're back and it should be good now. But yeah, lots to cover. Let's just jump into it. This is a daily upload where I go through my AI strategy and I cover some of the news if it is relevant. A lot of times it's not just But let's get into it. So, first off, I actually recorded a video early today. I had to delete it because I was basically doxing myself. There was like it was for my wedding and I wanted to test a bunch of stuff out. I basically recorded the whole thing and then went through all of it and made these markers and I was like, I'll just blur this stuff out. And I was like, no, this is I don't even know if this is that interesting. But what I did do was I copied this guy's stack. His name's Cheyenne and he makes product called User Jot and Log Snag. You can check those out. The guy's pretty badass. I like the way he yaps on the internet about things that he's building. So this is a great excuse to go try some new tech and built the whole thing about an hour and a half with tanstack router. So if you're used to directory or file-based routing that's what next is. It's all I've known with the next pages and next app router. It's just a little bit different. So in next you have you know app and then prompts and then page and layout. In this case, you actually have one file where you determine all of them. And there's just different ways of doing it. But Vit in this stack just absolutely murders it on speed. And again, I just built something really simple. Just set up a landing page, but absolutely enjoyed it. It flew. And I know Next has their reasons why it's slow. We want to explain it and all that, but devs don't care. They just want fast stuff. So I'll be definitely exploring these as a stack in the future. I just think it's cool. So I did that and then the way that I walked through it was just here's all the jobs that need to be done and I went through and did the design or sorry the setup from scratch all in Vim. So 15 minutes of that it was like 13 minutes design work making logos picking out fonts color schemes and all that using like coolers.co and using chat.com to generate a crest and color vibe like a bunch of mockups and then the page structure took a little bit of time because I just never done it. Getting shots in tailwind all these oily plate things and then the taskmaster was the majority of finishing it up. So it's like hey we want to have this and that and this and that. Good job. You guys want a video like this then comment below live builds and I'll do something similar. So I basically just recorded it and I sped it up. So, I wasn't actually yapping where I am right now and coding it because it slows me down and I wanted to knock this out, but I just recorded it and then afterwards I sped it up a little bit and then I talked over it. So, if that's something you want to see, let me know. And then on to the Q&A. So, I had a really good comment that I had such a good answer for before, but basically on the other channel, this guy said, "Yo, this is the best thing ever. I'm so happy to see somebody totally nerding out going into GCP, not just MN." And I was like, dude, this is the best comment ever. This is why I love YouTube because he said, I love this so much. Automation and exploring the GCP infrastructure for us smalls of not using NBS suggestion. Generate some workflow text files LLM generated for ease. So we can digest with an LM ourselves to then get a more detailed play through of your processes. Detailed play through of your processes. Cool to see this level geek out. Something greatly needed on YouTube. Awesome comment. And I went into it and I basically said, "Look, like, first of all, awesome. Thank you." Second of all, I have this as a task to do. I'm just working through the phases of the content strategy that I outlayed about a month ago and it's working so far, by the way. Like, if I show you my daily stats, I'm just sticking to the plan. I'm not negotiating it. But you can see I had a school and I made it free and I was like, I don't want to do this for free. Like, no one even cares. When you join a free one, it's basically like a glorified Facebook page. And I don't plan to be on school forever, but I need to make it paid because then people will care more. I will care more. It just makes the world go around. So that's when I kicked everyone out. It was the first of this month, so about 20 days. And then you just see how it correlates with the amount of marketing to do. So when I make videos, I get more people in the community. And then we'll eventually transfer over to Vibe with AI, which will be great. And I was actually working on that at the beginning video, but it got cancelled where I was using augments to debug something because I don't like the way that the RBAC works with one tap. If you're unfamiliar with one tap, if you go to a site like eBay and you're not logged in, but it has a Google session, which most people do, then it'll say, "Hey, continue as so and so." So that's what I added and it was causing some issues. So I'm having deal with it. on the point of his question with NA and what you're trying to learn on like which one to choose. It just comes back to what are you trying to get done and what is your goal? If your goal is to try to print money fast with selling this as a service and I think it's a great idea to an goal is actually more like hey I want to make big impact at scale by helping people through software and so I look at the hard skills. So I think this is actually a helpful exercise for pretty much anyone. It's like cool like what do you want to do? What's your goal? If it just like print cash, then you go and and this kind of goal thing, it like trickles down to everything you do. So I look at this and I'm like, okay, what are my skills? And I list them out. So it's like, cool, one of my skills is educating people because I've spent a lot of time being the connective tissue between engineers, managers, non-technical people, sales, ops, marketing, all those things. I've had to wear every single one of the hats. And that makes me good at teach like basically putting on a hat for a different audience. Like this is a very technical audience. So I talk a certain way. I don't really slow myself down. But if I went to talk to someone on the street, I'm going to be a different person. Um not like I'm schizophrenic, but like you you tune the message to the right person based on where they are. So I look at skills. I'm like education, coding, systems. Systems means system pattern, system thinking, system design, and marketing. like I love all these things because of my different experiences across my professional career. And so for me it's like okay well GCP it's gonna help me in a lot of ways because it ticks all these boxes where it's like it's going to be slower for me to get up and running but I need this thing in order to do this thing. It all ladders up to that then I need to use code and I need to use media at scale and it needs to be really high quality. I need full customization. And then each time that I go and I work on one of these things, it's going to make me level up on all these things. And then at the end of the day, all these charts, like if this is one month, it's going to go bananas if I just do a good job and then I don't negotiate on my goals. So when I look at that, I'm like, okay, it's obvious that if I'm following my interest, because it literally started with me being like, oh, I need nad in order to accomplish this. But then I mounted a bucket that had a pabyte in it. I was like, whoa, what else could I do? I just follow that path and it's a necessary thing for you to do because right now I'm in phase two of this roll out. Phase one, daily videos. This started about six weeks ago. Got that down. Phase two, add three main channels. Did that this week, missed it the second uh the first week. Now phase three is shorts. And that's going to be a doozy, right? That's when I need automation because doing this 10 videos a week across a year, 520, right? But then doing shorts where it's like, okay, I need this existing content to get chopped out and I can't use web hooks because literally no one does it. So, do I need to roll something? Not sure. But in that case, phase three would be okay, three videos a week. And if I do that across 3.1, which is splitting those three across three other channels, then what? That's actually nine a week. And then 9 times 52, it's 468. So I was hallucinating in the last one. But you're going to be at least 468 on the shorts, probably more. So, I don't know. It's a lot of lot of video content. And then from there, you kind of split it out even more. It's like you're going to have a terabyte worth of yap that you can then turn into written format. And it's across everywhere. So, in blogs, I'll look at like the href SEO templates and that would be like a phase one is run it through that. And then I do another one which is AI humanizer humanizer and this thing is through surf shark it's called something shark but there's a bunch of services. So yeah in order to do all this stuff like I might find that when I go to use the service that it's actually just this prompt plus another thing. So it's like two edge functions and instead of having another SAS I got to learn and I get to roll it myself and it's just going to be dope. So that's for me why GCP. But if I were if I had a different Y and it was just like just strictly like print than N. But I follow curiosity. I really am a strong believer in icky guy Japanese philosophy where it's like okay got these three things these three circles you want to be right here. This is fulfillment. So what are the things that you find fun? For me I don't find any very fun. It's not technically challenging. It's like fisher price. And then I have the value. So it's like if I'm gaining value, capturing value and then skill. So I'm probably like right here in terms of skill on GCP. Maybe like right here. I don't know where it would be. I'm like high here and high here. Basically I want to be here with with GCP. I'm like 50 hours into doing it. So learning how to learn is actually the skill. That's why you got to use AI. Ask a zillion questions. Follow the interest. So that's all it on the questions. Let's see. Then on to some of the news. So some of the stuff I was looking at this AI podcast thing and I went into it and it probably wasn't that interesting. So I recorded yapping like through this codebase. Somewhat interesting I guess. But what I do is whenever I'm trying to learn something that is in the code world, I try to find other projects. So like this one for instance, this guy rolled his own version of AI podcasting where it will accept all files, but it'll accept a PDF or a block of text or a MP3 or video and then it'll turn it into an AI podcast and it uses 11 Labs. So this would just be a cool example because it has a bunch of stars. I think that it's probably pretty good. Then what I do is I just spend about a half hour like digging around in here and asking questions. So literally like what the hell does this do or WTF is a block in this context and then I understand a little bit better like you go and you see like this guy spent a lot of hours like his walk thing for him tracking like I spent three hours 40 minutes across all code bases today not very much time but he probably spent like 50 hours in this thing so go learn from him and I learned oh cool this create project form thing is like pretty important these little routes in here these are pretty important too these are the fundamental things he uses upload thing for uploading things. So, it just saves you a whole bunch of time. And I would then add that like way way down here where it's like, okay, I've done all this other stuff. We're at like 6,000 pieces of content. Let's add an AI podcast. Probably wouldn't publish it. Next up is I need to plan out these lessons for cursor on fiber AI. So, a lot of people join because they see these taskmaster videos and like, whoa, how do you do that? And I think that a lot of people forget that what they're good at they find easy and they devalue it. So I find cursor just super duper easy to use. Taskmaster even easier to use because I was a product manager for so long. And then I've also been an engineer using cursor since it came out. And I've also been an engineer since using like GPT35 and the AI SDK. Like literally the day it came out I was like holy smokes this is cool. So you forget that there's all this stuff that you know and that's like a good message for anyone watching this that if you're good at something you're going to probably devalue it because it's easy for you and that doesn't mean that it's not valuable. So what I'm doing now and I announced this in the school is basically I'll do like a little minieries. It's not like a full course. It's like oh like how do you set up cursor? You do the shortcuts. What about this feature? What about that feature? You know all the little knickknacks in here that I use to be able to fly, right? So it's like oh like these noobs probably don't know they can do splits and they probably don't know that you can open it that way or open it this way or open it this way. you know, all these things and uh it's pretty helpful, right? So, even like Vin motions like I'm getting everybody. It's the jam. Sorry. But yeah, I'm going to be doing that and then something like this like, oh, let's open some splits there or how do you do these commands? How do you do those ones? I think people will find that helpful. It'll also go to the extensions. So, it's like, oh, this augment thing pretty helpful. You should use it for debugging and you should pair it with B1, which is a debugger. say, "Oh, B1, that's the one. You're an instant incident response engineer." Then boom, you paste in whatever the issue is. It does. It runs an indexer on it. Probably need to know about containers because if you're not using them, you're not it, dude. You're not that guy. They're just really helpful. And then climb. So, how to use the memory bank if you're upskilling on a new project. All these tools are independent right now. Eventually they'll be together but this is just the market. So client very specific use case for getting product context test text uh tech context and basically scaling up on something and then getting all these markdown files so that anyone can read them. It's really good at that. It's really good at a lot of stuff like but kind of kind of stuck to taskmaster now to be honest but yeah we cover that. cover docker cover else I use if it's down here I don't use it right but project manager I don't use everyone should know about git by now yeah I'm excited that should be helpful also just extensions there's little things in your settings that people don't realize that like autosave is messing with it and like what about this if there's like a VSS code thing is that's going to conflict with my llinter so I'm excited to do those there like 15 videos all be like three minutes because I wouldn't video that was about like a really narrow topic for more than that time. And then lastly, yeah, new version came out. They posted it today as of uh yeah 420 Easter, but it has approved PRD parsing, which means it can be multiple ones. It has chain of thought reasoning, which makes that better. It should be better. What's ironic is that like a week ago it was better to do a global install, but because it's been an entire week now, you shouldn't do that. So, kind of tricky to keep up, but that's just where we're at right now. If you guys have any ideas that I didn't cover in this video that you want my take on, cool. Uh, I have infinite content ideas, so I don't need ideas to be honest. But I want you guys to be like, "Hey, what's your take on this?" I guess that's that's the the thing in these videos moving forward is like, "Yes, ask any questions that you have in here." But like, I love a good hot take. So, I'll see you in the next video. Make sure you like the video. It's right there. That little thumb. Look at me, Mr. YouTube. Asking for a like and a subscribe.