00:00:00In the last month I've gained over 38,000 followers on YouTube,
00:00:0350,000 followers on Instagram and 11,000 followers on Tik TOK.
00:00:08And that is in huge part due to my Claude code content system.
00:00:12And today I'm going to break it all down the custom skills I've built my daily
00:00:16workflow and show you how I've used Claude code to automate my entire content
00:00:20system. So you can too.
00:00:22So what we're looking at here is the seven Claude code skills that are the backbone
00:00:26of my content system. And they have driven 10 million views over the last month.
00:00:30As one main team, no editors, no VA's nothing.
00:00:33Now those 10 million views are across 90 pieces of content
00:00:38over 30 days of those 90, 30 are long form videos.
00:00:43It's essentially been a long form YouTube video every day in March and 60
00:00:47short form. So those are shorts reels, Tik TOKs, all of that.
00:00:51And the real number you should be focusing on. Isn't the followers.
00:00:54As I said in the intro, even the 10 million views, it's the 90 videos,
00:00:5890 videos in one day for one person, like not to toot my own horn,
00:01:02but is a pretty impressive amount of volume.
00:01:05And the only way I was able to do that was with a sustainable,
00:01:10sustainable repeatable system. And that's what we're going to go over here today.
00:01:14Because again, I'm doing this as one person,
00:01:15but I'm not locked to the computer 12, 16 hours a day, right?
00:01:19The only way I can maintain this is again, if it's sustainable,
00:01:22if this is something that makes sense, furthermore,
00:01:25when it came to those 10 million views,
00:01:27there wasn't a single piece of content that got over 400,000 views.
00:01:31So this isn't a situation where also the 10 million number came from like two
00:01:35viral hits and the rest were just duds, right?
00:01:37This was a win with like 90 jabs and no haymakers.
00:01:40So I think that's also good to know.
00:01:42Like we aren't trying to just create some random lucky viral hit.
00:01:45This is small consistent wins that I think anyone can repeat.
00:01:48So how are we able to do this?
00:01:50How are we able to use cloud code to create the sort of sustainable system that
00:01:54does create content people actually want to consume? Well,
00:01:56first we need to understand the content creation process as a whole.
00:01:59Then we need to break down that process into individual parts and then assign
00:02:04specific cloud code automations and skills to those parts, right?
00:02:08That's how we methodically break this down.
00:02:10And I would break the content process into four real phases.
00:02:13The first is research. The second is ideation.
00:02:17The third is scripting. And the fourth is distribution.
00:02:22And it is from these four phases that we pull out different cloud code skills.
00:02:27And some of these phases have multiple skills because there's a lot going on.
00:02:31Take scripting, for example, right? That's going to encompass hooks.
00:02:34That's going to encompass the actual script,
00:02:36the outline of the video as well as some packaging stuff like titles and
00:02:40thumbnails. But let's start with the first two phases research in ideation,
00:02:44because I think it's important to talk about both of them in parallel because it's
00:02:47very much a cycle, right? You research some stuff,
00:02:50you come up with ideas from your ideas, you need more research.
00:02:52And then oftentimes from that research, you come up with more ideas.
00:02:56So one and two are very closely tied. Now,
00:02:59the big skill for me is my YouTube pipeline skill.
00:03:03And this brings in notebook LM. Now, every single skill you see here today,
00:03:08as well as the Twitter research engine,
00:03:10I'm going to show you in my GitHub script can be found inside of chase AI.
00:03:14Plus a link to that is in the comments. Chase AI plus is also home to my cloud code,
00:03:19masterclass, which is the number one place to go from zero to AI dev.
00:03:22This gets updated every single week. So if you're trying to figure out,
00:03:25how can I actually master cloud code and have like an actual pathway forward?
00:03:29Well, definitely check us out again, links in the comments now,
00:03:32back to the YouTube pipeline skill,
00:03:34which I think is the most powerful out of all seven of these skills. Well,
00:03:36that notebook LM pie skill allows us to bring the power of notebook LM into
00:03:41cloud code. So I can give notebook LM, whatever I want,
00:03:44whether that's like YouTube, URLs, PDFs, documents,
00:03:48anything I could do in the normal notebook LM, you know, web app,
00:03:51but I can do it through my terminal.
00:03:53And this is great because notebook LM is really good at handling some of this
00:03:56content that can be kind of a pain with cloud code,
00:03:59namely things like YouTube videos and all this is offloaded onto Google servers,
00:04:03right? We're not using cloud code tokens to do the analysis.
00:04:05We're having notebook LM and Gemini do it for us. And then we just bring it back.
00:04:09And I get access to all the notebook LM deliverables, right? Videos,
00:04:12slide decks, images, anything I can do here, I can do via the terminal now.
00:04:17And that skill uses the notebook LM PI CLI tool to create that
00:04:21bridge between cloud code and notebook LM.
00:04:24Now this repo includes its own skill.
00:04:27So the YouTube pipeline research is essentially a skill that calls additional
00:04:32skills. It's a higher order skill.
00:04:33And so what the YouTube pipeline skill does is it takes the notebook LM
00:04:38PI CLI tool and skill and essentially automates the
00:04:43sourcing of it.
00:04:44So it grabs a bunch of YouTube URLs based on your conversation and includes
00:04:49the analysis part.
00:04:50So it uses this as the bridge and then automatically sources and then
00:04:54automatically analyzes all in one command.
00:04:56But using the skill implies you already have a source of information, right?
00:04:59You've already figured out what you want to talk about or what you want to do
00:05:02analysis on, which begs the question,
00:05:04how do we even figure out what to talk about in the first place?
00:05:08How does cloud code help us there?
00:05:09And that kind of goes beyond what you see in the skill breakdown, right?
00:05:13What we need to figure out is like step zero. You know,
00:05:16we got to figure out what is the fountain head of knowledge for your particular
00:05:21niche for tech. It's pretty obvious, right? For all this AI stuff,
00:05:23everything kind of comes from a few places,
00:05:25either comes directly from GitHub repos or it spawns on Twitter, right?
00:05:30And then eventually it makes its way to YouTube. Every once in a while,
00:05:33something will like originate on YouTube,
00:05:35but it's usually Twitter and GitHub, right from there it goes to YouTube from
00:05:39YouTube. It gets spread out.
00:05:40So we need to figure out what your fountain head of knowledge is because if it's
00:05:45not tech and it's not AI, you need to understand like,
00:05:47where does the information originate?
00:05:49So you can kind of be first on the ground to talk about it. And so in my case,
00:05:53since we're saying, Hey, it's coming from GitHub or it's coming from Twitter,
00:05:57how do I use cloud code to help there? Well, when it comes to Twitter,
00:06:00I just had cloud code build me a Twitter scraping web app.
00:06:04So that's what you see here. It goes to telegram. And every 30 to 45 minutes,
00:06:08I get some sort of tweet based on a number of keywords and a number of authors
00:06:12that pops up and says, Hey, here's what they're talking about. Here's the lights.
00:06:16Here's like a velocity score.
00:06:18And it also allows me to reply to them if I want because I also hooked up my
00:06:21Twitter API and here's sort of the breakdown of how this web app works. Again,
00:06:25it was pretty easy to create inside of cloud code yet it's relatively
00:06:28sophisticated and it's very customizable.
00:06:31So every 45 minutes or so it's kind of on a randomized timer.
00:06:36It scrapes 40 to 90 tweets. It uses an app of high tweet scraper.
00:06:39That's pretty cheap.
00:06:40And then it filters it and scores the tweets.
00:06:43It finds based on a number of scoring signals. So it looks at velocity, authority,
00:06:48timing, opportunity, and repliability, because I have the ability to, like I said,
00:06:52to reply to these tweets. If I want to,
00:06:54all the tweets it gets gets sent to super base.
00:06:57So to make sure I'm not always getting the same tweet from the same person and
00:07:00that it also kind of just like diversifies it a bit from there.
00:07:03It scores them and then it chooses them based on the score. It uses soft max.
00:07:07So it applies a probability score to each.
00:07:08So I don't always get the number one score each time. Again,
00:07:11we want some randomization there from there.
00:07:13It gets pushed to telegram and it also has the ability to give me potential
00:07:16replies. So I have Brock tied to that.
00:07:18Now if you've been on Twitter for any amount of time at all,
00:07:21you know it is absolutely plagued with AI bots posting there.
00:07:24So all of the replies do go to super base and essentially get
00:07:29scored. And that way I have insight into the kind of responses I'm giving it,
00:07:34because I also can do custom responses and over time,
00:07:36it kind of becomes a system that improves upon itself. And then lastly,
00:07:40it shows up in telegram. Now let's talk about fountain head number two,
00:07:43which is trending GitHub repos. Yes, there's a trending page on GitHub,
00:07:47but why can't I just get this information automatically along with some nice
00:07:50insights as to the velocity of these trends, right?
00:07:53How many stars have they gotten since they first got created and also going to
00:07:57have it filtered, right? I just kind of want to see AI stuff. Well,
00:07:59Claude code did all that for me.
00:08:00It created a script that runs every single morning that brings me the GitHub
00:08:04trending repos in the AI space and it puts it inside of my obsidian vault.
00:08:08So what I'm able to see is the top 10 trending repos that were created in the last
00:08:12seven days. Every day I see the stars, the language,
00:08:16I get a link and I get a quick description on top of that.
00:08:19I also can see the top five trending for that month. And then it gives me,
00:08:22you know, it's suggestion each day. And why?
00:08:24And so between this GitHub script cloud code created and this Twitter engine,
00:08:28I'm able to solve for this step zero problem,
00:08:31which is like how do we even find things to talk about in the first place that
00:08:34aren't just a repeat of, you know, what's been on YouTube for the last week,
00:08:37right? We need stuff that is new and this allows us to do it. And again,
00:08:41the nice thing with Claude code is you don't have to use GitHub.
00:08:44You don't have to use Twitter.
00:08:45You just need to identify what those are for you and your niche.
00:08:48Then have cloud code build them because once you do have that right,
00:08:52the step zero of like the fountain head, then you can plug into here,
00:08:57in this whole skill breakdown setup, right?
00:08:59Then once I have the idea I found on GitHub or the idea I see someone talking
00:09:03about on Twitter, then I can throw the YouTube pipeline search at it, right?
00:09:07This is called YT pipeline, but it doesn't have to be YouTube, right?
00:09:09This can be whatever it is. And then that does the analysis on notebook LM.
00:09:13And like you saw with GitHub,
00:09:14this is also all being done inside my obsidian vault. So yes,
00:09:20I'm going to have my terminal up with Claude code talking to it,
00:09:22but everything Claude code creates is in a Markdown file inside of my vault.
00:09:27So it's very easy for me to see what's going on as well.
00:09:30And to take a look at reports and see connected articles, right?
00:09:33Just gives me better insight and keeps it all organized, right?
00:09:36Because especially if you're doing content,
00:09:38like if you're doing this every day, doing multiple types of research, this,
00:09:42if this is just in like a code base and you don't have any like obsidian in there,
00:09:44it can kind of get away from you as the human, like cloud code can handle it fine,
00:09:48but you you'll struggle. So we understand where to find ideas at the ground level.
00:09:52And we just kind of talked about the YT pipeline skill,
00:09:55how we can point it at those ideas. We found somewhere,
00:09:59send it to notebook LM and have it do a bunch of research and analysis next
00:10:03becomes kind of like ideation and strategy.
00:10:06And so this is taking that research and then figuring out how can we position the,
00:10:10these ideas with desire mapping?
00:10:12How can we take these ideas and actually turn it into content that someone would
00:10:16actually care about at a high level.
00:10:18And so what ideation is going to do is it's not going to redo the research,
00:10:22but it's going to take a look at the research in terms of the competitive
00:10:25landscape. And like, what are other people saying about this? What are the gaps?
00:10:29What are potential things that no one's talked about that might resonate with an
00:10:33audience, right?
00:10:33So this is taking the research out of a vacuum and then placing it again in the
00:10:38competitive landscape you inhabit. So let's take a look at this one in action.
00:10:42So I've been doing some research on rag and Claude code planning on putting it on
00:10:45some content. That's like the seven levels of Claude code and rag,
00:10:49because it's a space that actually has been changing a lot over the last really
00:10:52year. So we're doing, we're invoking the ideation skill.
00:10:56Take a look at our recent rag and Claude code research and come back with sort of
00:11:00the landscape. So here's what Claude code came back with. Again,
00:11:03it's pulling from research we've already done.
00:11:05So first thing it gives us is the competitive landscape, saturated angles,
00:11:10open gaps, and then performance outliers, right?
00:11:14What have other people talked about that kind of like went nuts after it gives us
00:11:17that context, it goes into video ideas, right? Titles angles,
00:11:21the kind of desire we're hitting, then the formats and competitive gaps.
00:11:25And it does this for a ton of different videos, right?
00:11:29And all it gave us nine different options and then it ranks them.
00:11:32And I think what you see here is important because it's repeated throughout all of
00:11:36these skills in the system. I use,
00:11:38when we talk about using Claude code and the automation,
00:11:40like what we're really talking about is turning Claude code into a collaborator.
00:11:44Right? At every step of this journey,
00:11:46I want to have some sort of input, right?
00:11:49I don't want Claude code to automatically go to GitHub and then I don't even see
00:11:53it. And then at the end it just gives me, Hey,
00:11:55here's the full script you're going to do today. By the way,
00:11:57I created the thumbnail and I created the title and everything's ready to go.
00:12:00You just need to say these words.
00:12:01You don't want that because it's going to be terrible. Okay?
00:12:04If you're doing anything with AI that has any creative bent whatsoever,
00:12:08you need to stay in the driver's seat. Now,
00:12:13obviously throughout all of this, Claude code is doing a ton for us,
00:12:16but it's doing analysis and it's coming up with potential plans and potential
00:12:20ideas. You still need to be there along the way to like check it off and say, Hey,
00:12:24I don't like this. I don't like that.
00:12:25That's how you can actually get a good output at the end.
00:12:29And that's how you can maintain your voice because no matter how well you train
00:12:32this thing,
00:12:32if you expect it to go from stab zero to the full script and no point in between,
00:12:37were you there to be like, let's do this idea. Let's change that.
00:12:39Let's change that it's going to be generic and it's going to suck.
00:12:42But the nice thing about this is, Hey, if you did want to automate it like that,
00:12:45you can, but every step of this journey, right?
00:12:48The expectation is that you take a look at the outputs of Claude code before you
00:12:51move on to the next phase. And what it's really doing,
00:12:53what this is really buying you is all the leg work to do this
00:12:58sort of analysis from scratch and seeing stuff like this and seeing
00:13:03its ideas helps you refine what you think you'll go with.
00:13:06Because I would say nine times out of 10,
00:13:09I ended up doing some variant of what it gives me.
00:13:11I don't usually do the exact thing, right?
00:13:12Because we always have something different we want to kind of put in there,
00:13:15but that's it for the ideation section, right? So we did step zero,
00:13:18found the knowledge. We did step one.
00:13:21We did some research with pipeline and we brought in notebook LM we've completed
00:13:25ideation. You know,
00:13:26we kind of understand where this potential content idea is within the context of
00:13:31what everyone else is doing. And of course,
00:13:34all of this is being done inside of obsidian instead of our vault.
00:13:36And if the obsidian stuff is kind of going over your head,
00:13:39I'll put a link above a video I did where a deep dive on sort of obsidian and
00:13:43notebook LM. And that brings us into phase three, which is the scripting section.
00:13:47Now, when it comes to scripting, I will say for myself,
00:13:50I'm not a huge script guy.
00:13:52I will script the hook like the first 30 seconds.
00:13:57So what you saw in the intro of this video, where I was like, yeah,
00:13:5938,000 followers and 11,000 people on Tik TOK that was scripted, right?
00:14:04I went back and forth with Claude code quite a bit using this hook skill and
00:14:07figured out, okay, what exactly am I going to say?
00:14:09Because when it comes to content and social media,
00:14:12the hook is very, very important. Packaging is very, very important.
00:14:14So I want to nail that and it's only 20 seconds, but everything else,
00:14:17it's outlines, it's concepts with bullet points.
00:14:19I kind of know what I'm going to talk about, but not really.
00:14:20We're just going to do it live.
00:14:21So the outline skill that I give you, again,
00:14:26it's just like that. It's big picture stuff, although the hook really nails it.
00:14:30And the hook stuff comes in large part from Callaway.
00:14:34I take a lot of his, his ideas, shout out to all of his content.
00:14:37His stuff is brilliant.
00:14:38So I essentially did this whole setup on a bunch of Callaway's videos and then
00:14:43incorporated that into how I have called code approach hooks and outlines and
00:14:47titles and that sort of thing. But let's see this in action.
00:14:50And we're going to run the hook skill,
00:14:51the outline skill in the YouTube title skill on this potential Claude code rag
00:14:55video and see what it comes back with. So I told it,
00:14:57let's use your recommendation.
00:14:58Its recommendation was a context engineering type bent and said, Hey,
00:15:03let's run the hooks outline and YouTube's title skill on it.
00:15:05So here's what it brought us for the hook section, five variations.
00:15:09And then for each hook, it breaks it down into a spoken hook, a visual hook,
00:15:12as well as a potential text overlay. If we want to add that as well.
00:15:15Now the text overlay specifically are more for short form type stuff.
00:15:19So this isn't something I would implement with the long form hooks then moved to
00:15:22the outline and includes the target length.
00:15:24Some of the related documents that are also in our obsidian vault that we may want
00:15:28to reference. And then it has the hook. And then again, outline is just a section.
00:15:32So like general idea, you know, the core idea context engineering is this,
00:15:36we're gonna explain what context engineering is as well as the talking points.
00:15:39It also includes potential visual aid. So Hey,
00:15:42if I wanted to add some sort of Excalidraw diagram, here's what you could do.
00:15:45And then some of the source material if I wanted to reference that on screen as
00:15:48well. And it just repeats that for every single section.
00:15:51And then lastly moves into title options.
00:15:53And the nice thing with the title skills, it's not looking at it in a vacuum.
00:15:56It actually looks at all your previously performing titles to get a sense of like,
00:16:00okay, what's actually working for this guy.
00:16:02And then it breaks them down into tiers. So first tier context,
00:16:07engineering just made prompt engineering obsolete. And then it tells you, Hey,
00:16:10here's what I'm basing it on, right?
00:16:11Here's like the previous video that did X amount of views.
00:16:15This is why I think this title would work. And it does this for all of these tier,
00:16:18two titles are calculated risks. So these are a little out there,
00:16:21which is nice to know because chances are you can do some like ABC testing.
00:16:25So every once in a while it's worth throwing something crazy out there instead of
00:16:28doing like three tier one titles that are all kind of samey.
00:16:31And then it follows it up with thumbnail text options. And again,
00:16:34kind of the same system here. And so between these three skills, hooks,
00:16:38outlines in titles, we pretty much have like 90% of our video mapped out, right?
00:16:43The packaging is almost there in terms of the title and the hook and what's going
00:16:47to be on the thumbnail.
00:16:48And then the video outline essentially carries the actual content.
00:16:52The only thing that's not in here obviously is something related to building the
00:16:56thumbnail itself, but that's a personal preference.
00:16:58I really don't think AI is great at thumbnail creation in a vacuum.
00:17:02It's one thing if I show up with a specific idea,
00:17:04but it's so visual and it's so subjective. That's something I do purely manually.
00:17:08And once you're in this place and you're happy with how this was all created,
00:17:11now it's time to actually film the content, right? And that's purely manual.
00:17:15Like I'm not someone who does AI avatars or anything like that.
00:17:18I don't really think it's worth it in 99% of cases.
00:17:20So there's no real clogged good automation for the actual creation portion of
00:17:24this. And so that moves us into phase number four, which is distribution, right?
00:17:28And distribution has a few layers to it. Now there's the most obvious form of
00:17:32distribution, which is like, Hey,
00:17:33we want to post this video to something like YouTube or Instagram or Tik TOK.
00:17:37Very easy to create something like that inside of cloud code,
00:17:40like an automatic distribution system.
00:17:42You can tie it to a particular folder inside of your Google drive and create an
00:17:45automation that like on a trigger when something gets added, it does that.
00:17:48To be totally honest, I use cap cut for my video editing.
00:17:53And so like posting it to YouTube from there,
00:17:55posting it from Tik TOK there is very easy.
00:17:58And I honestly just manually post it to Instagram.
00:18:00Is that the most efficient thing in the world? No,
00:18:02but it works for me because it takes two seconds and I'm fine with that.
00:18:04Especially since I do for Instagram trial reels and trying to automate that
00:18:09portion of it is annoying. I don't even think it's possible,
00:18:11or at least it wasn't when I last tried. So for me, when it comes to distribution,
00:18:15I'm thinking more of repurposing,
00:18:18repurposing in terms of taking a video from say YouTube and turning that into text
00:18:22content on my website as a blog and then text content on LinkedIn and Twitter and
00:18:27short form repurposing, right? If I have a long form piece of content,
00:18:30how can I turn that into short form? And I'm not talking about just clipping it.
00:18:34I'm talking about, okay, how do we distill 30,
00:18:3640 minutes of me talking to someone on YouTube into a 30 second,
00:18:4060 second 90 second clip on shorts or Instagram or Tik TOK, right?
00:18:43So these two skills, my content cascade and my short form skill do this.
00:18:48Now the content cascade is all about that video to text distribution, right?
00:18:52I'm taking YouTube video. I'm turning into LinkedIn again, like all skills.
00:18:55This is very, very customizable.
00:18:58You may not have some sort of like a YouTube content fountain head, right?
00:19:02You could switch that up for anything though.
00:19:04You could point this skill at just an article or someone else's YouTube video,
00:19:07something you want to talk about in a text format.
00:19:09And this will take that and turn it into a blog, Twitter and LinkedIn, right?
00:19:15Obviously this skill in particular is tuned to my voice,
00:19:18but it's not too hard to change that.
00:19:19Especially if you use something like the skill creator skill,
00:19:22which will run tests on it. So when I run the content cascade skill,
00:19:26it automatically grabs the transcript from YouTube.
00:19:29It turns it into a blog post automatically posts that turns it into a Twitter
00:19:33thread with like seven different replies. Again,
00:19:35automatically post that once I approve it and then gives me a few variation of
00:19:39LinkedIn posts. Now I'll be first to say I'm kind of lazy when it comes to LinkedIn,
00:19:44but I don't automate the LinkedIn posts because I use something like lead shark.
00:19:48Well,
00:19:48I do usually jerk to kind of have the whole a lead magnet thing set up with that.
00:19:53So this does a great job of doing that because again,
00:19:57there's so many platforms, there's so many social media platforms.
00:20:01It's not realistic to be like, all right,
00:20:03now I'm going to like take this content and write, you know, these posts on my own.
00:20:07You know, I know myself, I'm more of a video content guy.
00:20:10So any way I can automate the tech side of it is great.
00:20:14And here on my website, I'm in the blog section and you can see it automatically,
00:20:17obviously creates the entire blog,
00:20:20but it also embeds the YouTube video has a bunch of SEO stuff.
00:20:24It's SEO, um, like forward.
00:20:28So the whole idea is this blog is less about, Oh,
00:20:31these articles are so good on my blog. And more of that,
00:20:33like as my content repository continues to grow, so does the blog,
00:20:38so does my visibility on things like Google search, right?
00:20:40Everything just ties into everything.
00:20:42Cause I sure as hell and not writing those blogs on my own.
00:20:45Although I did give it a ton of my own writing so it could see how I wrote,
00:20:48right? It stays away from things like the chat GPT isms, right? It's not X,
00:20:52it's Y right? So part of the skill,
00:20:55it takes a look at all the AI writing tropes and avoids them.
00:20:59And then last but not least is short form repurposing. Now,
00:21:02the short form repurposing is pretty basic.
00:21:05Essentially it's redoing all this stuff like the hooks, the outlines,
00:21:10right? And then it's just putting, you know, that into a 30, 60, 90 second format,
00:21:15right? It's giving you hooks to use.
00:21:16It's giving you potential captions in terms of like what pops up on the screen at
00:21:20the beginning. So this is just a condensed form of what we've already done.
00:21:23And because my short form is already being pointed at a long form video,
00:21:28all the work's kind of already done, right? It's just cutting a ton of the fat.
00:21:31But what that allows me to do is, you know,
00:21:33essentially take what I create on YouTube.
00:21:36And this just becomes like this monster that I can post a blog on my website.
00:21:41I can post on Twitter. I can post on LinkedIn. It becomes a short,
00:21:45it becomes an Instagram reel and it becomes a Tik TOK, right?
00:21:48Six platforms,
00:21:52six different pieces of content from one main thing I've created on YouTube,
00:21:55right? That's where the sort of the content cascade name comes from.
00:21:59And that's the beauty of this system because it doesn't just end at a YouTube
00:22:02video. The YouTube video itself becomes its own right little founded head of
00:22:06knowledge that we talked about before, but it's for you.
00:22:09So that is my cloud code content system.
00:22:11It is essentially my collaborator on steroids. Like I've said before,
00:22:14every step of this process, like I'm going back and forth with Claude.
00:22:17I am not expecting it to give me a perfect product at the end,
00:22:20but I offload so much legwork to it. All the analysis,
00:22:24all of the competitive research, all of like the hooks,
00:22:27all the baseline ideation, like it does all of that.
00:22:29And it lets me kind of focus on the high leverage things. Furthermore,
00:22:32once I create a piece of content,
00:22:34it gives me a very simple to X execute path towards
00:22:39distributing it in different forms of multiple platforms, right?
00:22:42Which is what leads to something like a 10 million month as a single person
00:22:46without any quote unquote viral posts.
00:22:48So if you want to get your hands on all the exact skills,
00:22:50the Twitter research engine, the GitHub script and the cloud code masterclass,
00:22:55make sure to check out chase AI. Plus again,
00:22:57there's a link to that in the description in the comments.
00:23:00There's also a link to my free JCA community in the description.
00:23:04If you want a bunch of free resources and just getting started with AI.
00:23:07But other than that, let me know what you thought and I'll see you around.