[0:00]So we know AI agents can save us lots of time, but can they actually make us money? Well, one developer recently found out that that answer is yes.
[0:10]He was able to develop an AI agent which has generated him nearly half a million, over half a million views on Tik Tok and nearly $1,000 in monthly recurring revenue on his app.
[0:22]Generating tons and tons of content, consistently iterating and improving on the content and essentially replacing an entire marketing team. Uh just using AI.
[0:33]And this case study, I think is very indicative of what we can expect to come for essentially all of our careers.
[0:40]So today we I want to do a quick case study on what this developer did to leverage these open this Open AI agent to completely explode his business.
[0:51]And again, you may say, well Mark, $1,000 a month in recurring revenue isn't a ton.
[0:55]But guys, that's in five days. He did this in five days.
[0:59]And what I think you need to understand is that these AI agents learn and iterate and become better and better and better.
[1:08]So there's this momentum that builds. Anyway, let's do this case study and I I think you're going to be blown away by what this guy did.
[1:15]Because the reality is a lot of guys out there say that, oh, AI is fake and it's always an Indian and it's it's just a bunch of smoke and mirrors.
[1:23]The big tech guys don't really have Guys, this is for real.
[1:27]I I don't know what to say, but every day I'm blown away. So I want to do more case studies of what's actually working because it's it's kind of blowing me away.
[1:33]So, so what happened here? So a developer named Oliver Henry used his open claw agent to create an AI agent named Larry.
[1:43]And essentially, Larry, this autonomous AI entity, if you would, uh ran on an old gaming PC.
[1:50]So he had a dedicated PC that ran this. By the way, I was looking the other day, actually just today, I was looking at an old PC that I have sitting on the ground.
[1:59]And I was like, huh, I wonder what trouble we could get into with that old computer there.
[2:05]Anyway, uh Open claw. What is Open claw? I covered this in more detail in a video. I'll post a link to the video here.
[2:10]Um, but Open claw is an open source software that gives AI a persistent identity, memory, and access to actually real tools.
[2:21]So Open claw, this entity can have access to your social media, to your calendars, to your email, basically anything that's on a computer.
[2:30]Your open claw agent can gain access to, which is exciting and terrifying at the same point.
[2:35]In any event, in this case, uh he connected his open claw agent to his clawed AI account.
[2:44]And here's what Oliver did when he set up this open claw.
[2:47]So, uh, these are the exact steps that Oliver took with setting up Larry, his Open Claw agent.
[2:53]So, he gave it its own personality and memory so that it could persist between sessions.
[2:59]It could get smarter every time it opened up Tik Tok, every time it opened up an app, every time it looked at the data and the results.
[3:07]It got smarter and smarter. He also gave it the ability to read and write files onto Oliver's local machine.
[3:13]Again, you're using Open Call, you probably want a dedicated computer uh separate from your the computer that you're using daily because you want it to have basically autonomy.
[3:23]It's like your own employee working on this computer.
[3:27]Uh he gave it the ability to generate images through open AI's API.
[3:33]Basically, that's chat GPT. Um custom code that he wrote himself. He uh added in order to text overlays onto those images.
[3:41]So, you know, you've got an image, it's going to create images, it's going to include text over that.
[3:45]Um, he gave it access to an app called posted, uh, which essentially manages his Tik Tok account.
[3:52]It's an API that connects to his Tik Tok account.
[3:55]He gave it skill files that teaches his AI agent specific workflows and he granted it memory files where it logs every lesson learned.
[4:06]And this is a really important thing, guys. Anytime you're using AI, any frankly, when you think about implementing a new AI operating procedure, let's say or a new AI software.
[4:17]You have to think of it kind of as like a science experiment. You want to hypothesize, you want to test, you want to iterate, you want to observe, you want to see if you can replicate and iterate and improve.
[4:27]This is really important. You can't just say, oh, I set up my you know, my my AI agent or my AI GPT and now it's it's done.
[4:35]Well, no, that's the first step. You got to continue to iterate.
[4:38]And that's why these um adding the memory files to his AI agent is so important because he doesn't have to go in and manually adjust it all the time.
[4:45]This thing is learning on its own.
[4:48]And how does Larry actually communicate and talk to it? He just has a WhatsApp chat. He just chats with it in WhatsApp and I've I've heard in some cases guys, um you don't have to use WhatsApp.
[4:59]You can use Telegram. There's different, you know, messaging apps, but he just has conversations with it in Telegram.
[5:03]Hey, you know, what's up? We're going to talk about that here in just a moment. Um as well, but anyway, you go, so he's having these conversations.
[5:11]Larry will be his AI agent will be like, hey, posts are done. Go ahead and go ahead and get them posted up and he does that.
[5:17]And I've even again, heard people using Telegram, of course, and just having phone calls with their AI agents.
[5:24]You hook up your AI agent to 11 Labs, well, now you can actually get phone calls from your AI agent and have real conversations, which is is kind of bizarre having your AI agent give you a call.
[5:38]That's pretty, pretty wild, right? Anyway, so, he's got his content machine created up. Uh created how does he actually turn this into viral content that drives traffic to his app and drives monthly recurring revenue?
[5:51]Well, his AI agent, Larry creates carousel images for slideshows and it overlays text on top that image.
[5:58]And he's found at least right now that six slides per post with a text overlay with a story style caption and five hashtags is the key.
[6:09]And he also needs to go and add music on top of that. I'll cover that here in just a moment.
[6:13]His goal with this Tik Tok channel is to drive traffic to his app called Snugly.
[6:18]And basically what Snugly does, is it helps people design rooms.
[6:22]You have a blank room, it will help put up a design, you know, kind of like what would it look like with different furniture, different paint, and so on and so forth.
[6:29]So basically what Larry does is Larry writes these detailed prompts based on previously discerned criteria.
[6:37]It's going to keep testing new things. And I think that, guys, is the most remarkable thing.
[6:44]You know, in the past when you're using automation, like whatever you tell the app to do, like I've done a lot of work with Zapier, whatever you tell it to do, it's going to keep doing that over and over and over again.
[6:55]It's not going to iterate on its own. You can go in and you can learn and give it new directives, new prompts, new instructions.
[7:02]But what Larry does is it continues to iterate.
[7:07]It continues to adjust. Um and so anyway, once these posts are completed, Larry uses this app called Post to schedule social media posts on Tik Tok.
[7:20]And then it will uh notify him that it's time to push live.
[7:29]So, uh, the Open AI agent generates the images, adds the text overlays, writes the caption.
[7:33]Larry uploads everything to Tik Tok, like I mentioned, using post, and saves it as a draft.
[7:39]It does not post it. Once the draft is saved, uh the open open claw agent or the the Open Claw agent sends a message to Oliver, the creator via WhatsApp.
[7:46]Oliver opens up Tik Tok, picks a trending sound, adds it, paste the caption, hits publish, done.
[7:54]So he's doing 60 seconds worth of work to get these posts out. And then what happens is his Open AI Open Claw agent, Larry learns with every post.
[8:03]It iterates on every post. It improves with every post.
[8:06]It's going to test a lot of different things, testing a lot of different things at first, and a lot of stuff was flopping.
[8:10]And then all of a sudden, he found this kind of formula of like, if we put this type of text into the into the post, we know that that has a a viral opportunity.
[8:23]Um and it's just going to start maximizing for views, for engagement. Every failure becomes a rule, every success becomes a piece of its formula and improves constantly.
[8:29]And by the way, I do this on YouTube as well.
[8:32]I know what type of videos are going to hit. I know what type of intros can in hooks are going to engage viewers and bring them in.
[8:39]And I'm I test different things out because we don't want I don't want to be a one-trick pony.
[8:45]There's a lot of things that I want to talk about, but regardless of the fact, what we're seeing here is we're seeing an evolution in the way that engagement happens online.
[8:54]You know, for someone like me, I'm in the marketing world, it's all about traffic. How much traffic can we get?
[9:00]How well can we convert that traffic? What type of offers are going to work the best?
[9:47]And, you know, leveraging AI agents to do that at a high level and improve is like completely mind-blowing.
[9:54]Now, with all that said, this technology, these AI agents, open claw, is so new.
[10:01]And as we've seen over past few days, it can be pretty bizarre.
[10:05]It can be unpredictable. We covered a um a tool that an Open AI developer created called Lobster Wild, which essentially gave away all of his Solana, all of his, all of these, all of these meme coins on the Solana blockchain away just because somebody asked.
[10:23]So, like these AI agents can be pretty unpredictable, which is kind of scary.
[10:27]I haven't, you know, between you and me, I haven't jumped into this open claw madness yet because I don't, you know, I'm concerned about privacy.
[10:35]I'm not a developer. I'm a regular guy. I'm in marketing and I don't know that I'm ready to take that leap.
[10:43]However, this stuff is moving really fast and there is a first mover advantage.
[10:46]So, I want to bring this to your attention. What do you think about Open Claw? What do you think about these AI agents?
[10:51]You know, is this all hype? Do you think that it's too early to get in?
[10:55]Do you feel like if you don't step in and start using these tools, you'll be late to the party?
[11:00]Because the reality is in a year, everyone's going to be using this stuff.
[11:06]I didn't cover this in a video, but chat GPT, Open AI, Sam Altman, went out and got the developer, the primary developer of Open Claw.
[11:16]Poached him, brought him to Open AI so they could start developing these AI agent tools over there.
[11:21]So, the world is moving so fast. I, you know, I'm not going to I'm not going to sit here and say that nobody's going to have a job.
[11:28]But what I will say is that everyone's job is going to change dramatically.
[11:32]And the guys that understand how to use these tools are going to be well on their way to being relevant, respected, and financially free.
[11:39]Period, full stop. I'm about to get home and I'm like this close to pulling the trigger on setting up this open claw on a computer.
[11:46]I know it's there's risks involved, but it's it's so insane what's possible if you can build out these tools.
[11:53]Anyway, guys, I hope you enjoyed the video. Thanks for smashing the like button. If you're subscribed to the channel, you're going to know this stuff is happening before everyone else.
[12:00]So make sure that you're subbed and I'll see you in the next one, guys. Yeah, peace.
