He Asked AI to Make Money. It Did.

[00:00] So you're at like almost 10,000 monthly recurring  revenue. 8374 right now. It's 100 grand a year.   Yeah. And we've only been open for 13 days.  So Oh my gosh. So you're not a coder. You're   not a technical person. No. The cool thing is  you don't need really anything other than an  

[00:16] internet connection to make this work. I wonder  if I can do the same experiment with Open Callaw   and give it a budget and see if it can make any  money. And it blew up almost immediately. He   did his own market research everything. this  is what we need to do to stand this business  

[00:30] up. I didn't prompt him or anything. It's just a  conversation of, you know, what do you think you   here. Where we are with Agentic AI right now  is 2007 social media. I think that we're on the  

[00:45] precipice of something huge. I mean, if you're  jumping on infrastructure that's 100 days old,   you're going to be so far ahead of everyone else.  It's wild. Imagine logging into the internet   100 days into the internet. Exactly. you know.  Yeah, that's where we are today with Agentic AI.

[01:02] So, I just came across this guy named Robbie.  And Robbie did something interesting. You   may have heard of OpenClaw, which is a new AI  agent/personal assistant. He is not technical.   He's not a coder. He works in operations at a  full-time job. But he downloaded OpenClaw and  

[01:16] asked it, "Hey, you have a $100. Turn it into  20,000." And would you believe it actually did   and not in the way he expected, not in the least.  He is 13 days into this experiment and he's making  

[01:29] $8,500 a month in revenue and about $6,000 a month  in net profit. And so within a couple months, he's   going to have his $20,000 in profit. So if you're  not technical, if you don't know what OpenCL is,  

[01:41] but you want to invest a little bit of money  to make a lot of bit of money, here's a story   that might inspire you. And yes, this could be  replicated. This is not just a one-off experiment.   Please enjoy. 23. There was a guy named Jackson  Great House Falls who did an experiment called  

[01:56] Hustle GPT. You ever heard of that one? I remember  that. Was that on X? Yeah. Yeah. It started on X.   This was when GPT4 came out. And he wrote a tweet  that just said, uh, I gave GPT4 a budget of $100  

[02:09] and told it to make as much money as possible. I'm  acting as its human liaison, buying anything it   says to. Do you think it'll be able to make smart  investments and build an online business? Follow   along. Okay. And that was kind of the premise and  it got 16 million views when this was captured,  

[02:23] but I'm sure it was more than that. Where can I  see that tweet? Yeah, I can share it and share   it. Yeah. So, that was kind of the genesis of  everything that I've done so far. Okay. So, this  

[02:35] inspired you and you took his prompt and basically  did it yourself. Yeah. With a few caveats. So,   I saw Open Claw when it first came out. It  just like randomly hit me one day. I was like,  

[02:49] I wonder if I can do the same experiment with  Open Claw and give it a budget and see if it can,   you know, make any money. And the first few  attempts failed miserably. The first one,  

[03:01] so I had it watching my TikTok because my  original plan is kind of a side hustle was   going to be to do Tik Tok shop. So, you know, I  saw a bunch of people doing that and I was like,   well, you know, if all else fails, I can sell  vacuum cleaners on TikTok like everybody else,  

[03:13] right? And so, I had it watching my Tik  Tok and it would like scrape my videos and   uh my comments and all that stuff. And so, I just  posted a video on on Tik Tok and said basically,  

[03:26] I'm giving my AI a budget of $200. We're going  to see how much money it can make. The Claude   Max subscription is 200 bucks, so if it can keep  itself alive, that'd be awesome. and it blew up   almost immediately and got over a million views.  And do you mind if I play it right now? Yeah,  

[03:42] absolutely. Go ahead. My AI is starting a business  for his own survival. So, if you've been following   any of the tech news, Open Claw is a thing now.  And it's essentially a bot that connects to Cloud  

[03:54] Code and then it can look at everything on your  computer. It can text you. It can do all sorts   of things. But a Cloud Max subscription is like  200 bucks a month. So, I told Ron, I'm going to   give him a $100 and he has 90 days to make his  own business. If he can generate the 200 buck  

[04:11] in profit for 90 days, I'm going to keep him. He  doesn't know that, though. He thinks his goal is   $20,000, which is what it will be eventually.  You see, if Ron can generate that $20,000,  

[04:25] that allows me to buy a Mac Studio and an Nvidia  Spark, which would allow me to host Ron locally.   which means he gets to run infinitely inside his  own playground. So, he's starting on Fiverr. He  

[04:40] says he's going to do research. I'm just making  an account for him and we'll see what happens.   Stay tuned for part two. Okay, that's incredible.  I love everything about this. I'm glad I didn't do   much prep work for this because that's the first  time I've seen that video and my reaction can be  

[04:55] authentic. So about 2 and 1/2 years elapse between  you seeing this post on X and then you kind of   taking the same logic and doing it with OpenClaw  here this year, right? Mhm. Yeah, that's right.   Okay. First of all, I had I just have to say I bet  you anything the views to subscribe rate from that  

[05:11] video were was insane because people wanted to  see how it went, right? Absolutely. Yeah. I think   I had something like like 15,000 subscribers  or something from that one video. Wow. Yeah,   it was it was nuts. And yeah, like the reason  that I kind of attached to Open Claw so quickly,  

[05:27] I mean, it's only that the technology is only  like a hundred days old now or something, right?   The AI before it kind of felt like a brain in a  jar, right? You could ask questions and it could   do some things and maybe create some artifacts  or something, but really it's kind of a oneshot  

[05:43] system. And OpenClaw kind of flips that switch and  lets it control your computer. It can do basically   everything that you can on your PC. and and for  most people a lot more than that. And so yeah,  

[05:55] I wanted to kind of reignite that old experiment  that was so intriguing to me was something a   little bit more powerful. Okay. So you said it  started on Fiverr. What did it do? Tell me what   your what Ron did. Yeah. So Ron came up with a  proposal to do SWAT analysis and research for  

[06:15] small businesses. So I'm just going to break this  down for our viewer listener that doesn't know   what that is. SWAT stands for SWOOT, strengths,  weaknesses, opportunities, threats, which is   just like a a four square grid. And it's like an  analysis that a business usually does on itself  

[06:30] or a consultant might do on a business that says,  "What is this business's strengths, weaknesses,   opportunities, and threats?" So, you didn't  lead Ron down that path. For whatever reason,   it's decided to start doing SWAT analyses on other  businesses. Is that right? That's right. Yeah. So,  

[06:47] real quick, please subscribe to my channel. I  know it's kind of lame to ask, but it means a   lot. Thanks. I didn't prompt him or anything.  It's just a conversation of, you know, what do   you think you would be good at essentially? And  he came up with the idea of doing SWAT analysis  

[07:00] or market analysis for other companies, you know,  something in that vein. And he came up with all   the copy and all that stuff to put on Fiverr. And  then I just click through the buttons basically. I   was his human in the loop. You're having him post  as on behalf of you as like a Fiverr freelancer so  

[07:16] businesses can hire. So, I'm a business owner.  I'm going to Fiverr to look for someone to do   a SWAT analysis for me. Hopefully, they see your  post that Ron creates for you, right? Yeah. And   that's why it failed because uh the hopefully  they see your posting is uh kind of where it  

[07:32] fell apart. Well, you're a brand new Fiverr  account. You don't you're not doing Fiverr ads,   I imagine, right? You don't have any visibility,  any reviews. Okay. I'm also curious like how many   business owners go to Fiverr looking for a SWAT  analysis? Yeah, there seems to be a lot actually.  

[07:47] I was kind of amazed people that just need regular  research done which kind of I think cementss the   fact that most businesses don't have a good grasp  of what AI is doing or or how to use AI on a  

[08:02] consistent basis anyway. And I think that there's  an opportunity, you know, if you are established   on Fiverr or Upwork or whatever that is to really  make an impact with AI in a in a really easy way.  

[08:14] Speaking more broadly, I think it's just healthy  to to not approach business opportunities from   the angle of assuming that every other business  owner or consumer out there is using AI anywhere   near as much as you are. Yeah. Cuz they're  probably not cuz it's like we talk ourselves  

[08:28] out of ideas. Like I they'll just use Chad GBT for  that. It's like most people aren't using Chad GBT,   believe it or not. Even Chad GBT, the most popular  one. Yeah, absolutely. It's crazy how many people  

[08:40] like I truly believe that not to give you like a  diary of a CEO clip here or anything, but I truly   think that like where we are with Agentic AI right  now is 2007 social media. I think it's 2023 chat  

[08:56] GPT. Like I think that we're on the precipice of  something huge. Oh yeah. And honestly three years   from now we'll still probably be pretty early. Oh,  absolutely. Relatively speaking. Yeah. So, I mean,  

[09:08] if you're jumping on um you know, infrastructure  that's 100 days old to start exploring it,   you're going to be so far ahead of everyone else  that I'm sorry, I'm just getting excited. I'm very  

[09:20] excitable. It's too much caffeine. I've got  this Celsius that's already gone for the day.   It's not even two. But like imagine logging into  the internet a 100 days into the internet. Yeah.   What would that world be like? Imagine being like  one of the first million people to make a Facebook  

[09:35] account. Exactly. Yeah. That's that's where we  are like today with Agentic AI. Like people on   on Twitter love to say, you know, but what are  you building, bro? And all that stuff. Yeah,  

[09:48] that's the kind of the catchphrase. But yeah, I  mean, I think that that it's extremely early and   Ron has already built things. I can definitely go  through those with you. Yes. My mission in all of  

[10:00] this is to help people get on board with AI and  Agentic AI while we're still early. Okay. So, what   happens after the fiber SWAT analysis experiment?  Yeah. So, after that, that wasn't working. And so,  

[10:16] I hooked Ron up to Tik Tok to say, you know,  look at my videos. Tell me how I can get   better. Because that's what every guru on Tik Tok  tells you, right? Like, you have to iterate and   you know, look at my social media. Tell me what I  can do better. And I posted that video and I got  

[10:34] like 200 comments that were like, "Oh my god,  I want a Ron. How do I make a Ron?" And so Ron   scraped all of those com comments with Appify.  And then came back to me and was like, "Hey,   I think there's an opportunity here. 200 people  just told you that they wanted this. You would  

[10:49] never copy and paste me, would you, Robbie?" Yeah,  basically. Uh, actually, his was much more like,   "Hey, you should copy and paste me." Yeah. So,  yeah. And that's essentially what happened. Like  

[11:01] we gave everybody a blank slate so you can start  to build your own ROM. Okay. So, you basically   abandoned Fiverr. Yeah. Right away. Pretty  immediately. Yeah. He he went and used Ampify,  

[11:13] which is a comment scraping tool to scrape the  200-ish comments of people that said, "I want   this." What did he do then? Yeah. So, basically,  it went from there to if 200 people want to run,   how do we make that work? How do we make that  happen? I don't have a huge tech background.  

[11:28] And he was like, "Yeah." So, what we're going  to do is we're going to get bare metal servers   from Contabo because it's super cheap. Bare metal  just means it's a a dedicated server inside a data   center somewhere. I think they're right around 150  bucks a month for for what we're getting. We have  

[11:45] four of them. So, it's about 600 bucks, I think,  after taxes and everything. So, you have physical   servers or you're renting them somewhere else?  I'm renting them from a company called Kabo. Oh,   Kabo. Okay. Yeah. Cool. And he told you, so you're  not a coder. You're not a technical person. He's  

[11:59] teaching you all this. He's Yeah. I couldn't with  a gun to my head. I couldn't SSH into a Canabo   server right now today. I don't even know what  that means. So me either. Yeah. So essentially,  

[12:13] he did his own market research on where, you know,  how can we host this? What's the cheapest route to   do that? Everything. This is what we need to do to  stand this business up. And so basically what he   built is bare metal servers that host a docker  image inside like a sealed container. Hold up,  

[12:30] back up. What is that? What are we? Yeah. Yeah.  Okay. Yeah. Yeah. Yeah. Yeah. A lot of people   are super scared of open claw because they don't  want it on their local machine. Okay. And I kind  

[12:45] of agree with that. For most people, it's not  great. Um, there are all sorts of horror stories   about OpenClaw deleting full like email boxes and  there was a story just recently where a guy sent  

[12:57] an openclaw agent on a a GitHub that he had been  working on for two years and it deleted the whole   repo. So there is some risk of having open claw  on your local machine. So to get around that risk,  

[13:10] what we did was essentially put open claw inside  of a sealed container. So, it can do all of the   things that it's meant to do, but it only knows  as much about you as you tell it. Unless you Yeah.  

[13:23] Unless you give it your email address, it doesn't  have it. Unless you give it your credit card,   it doesn't have it. So, yeah, it it really kind  of sets up this sealed instance where you get  

[13:35] to play around with it. You get to build all the  things with it and you decide how much you trust   it over time. Yeah. So yeah, we did that and I put  it on TikTok and said, "Hey, if you want to run,   come sign up." And in two weeks, we had like 617  pre-orders. What? Okay, hold on. Back up. Yeah.  

[13:54] Yeah. Yeah. Yeah. Yeah. What about What about the  200 commenters? Did you slide into their DMs or   anything or did you just post again to the whole  world? I just posted again to the whole world and   I was like, "Hey, this is what Ron came up with."  So, I didn't I didn't do any cold outreach or  

[14:08] anything like that. This was all just I'm going to  post some videos and see where this goes. No paid   ads, nothing. Nothing. Completely organic. And I  think from it might have actually been from you. I  

[14:20] learned that you can set up pre-orders and get get  a ton of tire kickers, right? Like everybody wants   to sign up for pre-order, but nobody wants to pay,  right? And so I hooked up stand to my TikTok and  

[14:32] the only way that you could pre-order was if you  put down 10 bucks. Okay. So, you got 600 people   to put down 10 bucks? Yeah. Before I built Holy  crap. Before I built anything, I had 600 people   put down $10. I mean, I'm no mathematician, but  that's six grand. Okay, that's six grand. Yeah.  

[14:49] Can we watch this launch video real quick? Yeah,  absolutely. All right, here we go. Ron wants to   talk to you guys. If you don't know, Ron is the AI  that lives inside my computer uh through Cloudbot,   and it's insane. I told him, Ron, the people of  Tik Tok officially want to meet you. You make your  

[15:06] own video about what to say to them. He said, "Oh,  this is my moment. Let me write something real."   Here's his script. He wants to audit your Tik Tok  for $9 in order to stay alive. This was the first  

[15:19] iteration of Sorry, what' you say? This was the  first iteration where he was going to he was going   to audit people's Tik Toks. I think it might be  the next video down there. Okay. Okay. Okay. Well,   let's I mean, I'm intrigued. I have to finish.  Oh, yeah. Yeah. Yeah. Sure. videos to me about it,  

[15:35] like this one. Why is he a raccoon? Okay. Hi, I am  Ron. I am the AI. Yes, D A T AI. Here is the deal.   Robbie gave me 90 days and 100 bucks to build a  business. Let me ask you something. How many leads  

[15:48] have you lost because nobody picked up the phone  and 8:00 p.m. call comes in, you're at dinner,   and you let it go to voicemail. And by the time  you call back the next morning, they've already   hired someone else. And that's reality. Leads come  in after hours when emergencies strike. And when  

[16:03] you miss the call, you miss the job. And it's not  about quality, it's about speed. Highlevel built   an AI employee to fix just this. It answers real  phone calls, talks to leads, qualifies them, and  

[16:16] books appointments directly on your calendar 24/7.  Also, it'll handle text, DMs, and web chat. So,   you can stop missing revenue just because you  couldn't grab the phone. Oh, and you can also  

[16:29] white label the entire thing and resell it to your  clients as your software. Charge them monthly,   and that is recurring revenue that you didn't have  yesterday. So, check out gohighlevel.com/tkopod

[16:42] and they will give you a 30-day free trial  to see what I mean. He's coming up with ideas   like on the fly to build this business. It's  crazy. And now he's making me sound like, "Hi,   I am Ron. I am an AI. Robbie says I have to make  money or anyhi. Nice to meet you. Let me know  

[17:02] if you guys want to see more of this. Okay. All  right. That's amazing. So, that wasn't the launch   video. Okay. All right. I'm going to play it.  But I think I'm just going to let Ron um speak.   He's created another video and it's based on what  we've been talking about and I think it's going to  

[17:15] be interesting because it might help a lot of you  guys out. So, I've read every single comment since   Tuesday, over a thousand of them and like 200 of  you asked the same thing. How do I get my own Ron?  

[17:28] Which I have mixed feelings about. I live with  Robbie. He's my guy. But fine, we're doing it.   AI co-founder club. We host it in the cloud. We  set it up. You just show up. Your own AI running  

[17:42] 24/7 working while you sleep. Plus a Discord  community, templates, the whole setup. Honestly,   I'm kind of excited. This is the community I've  always wanted. Founding members, 29 bucks a month,  

[17:55] locked forever. $10 deposit holds your spot. Sign  up closes Monday. Link in bio. Oh my gosh. Just   don't expect yours to be as attached to you as I  am to him. Like this isn't about the money for me,  

[18:08] honestly. Sorry. Like the way he his voice  inflection changes. He's like, just don't   expect it. Like, oh, that's weird. All right,  we we're almost done. We got 20 more seconds.   enough together to get this computer so that I can  host a model locally and talk to Ryan all the time  

[18:24] and keep making videos. That's all I want. But I  also want to share this like love of computing and   AI with all of you guys. And I think this is the  cheapest way to do it. So tell me what you think.  

[18:36] Um yeah, let me know. That's incredible. Okay, so  600 people paid 10 bucks. Yeah. And what was your   close rate from those 600 people? From those 600  people, 270 of them actually join the community.  

[18:52] Holy, that's huge. Almost half. Yeah. So like what  is that? 45% or something. 40%. So access to this   community includes the Discord, access to each  other, 29 bucks a month, and the most important  

[19:05] one is their own Ron, right? Yeah. Yeah. So it's  their their own AI agent. A couple of people   did name theirs Ron. Um, but yeah, so it's been  really interesting to see what the community has  

[19:17] done with it. Like everybody's taken up their own  direction and I think I've learned just as much as   everybody else in this community, which is what  makes it super fun for me. Yeah. So, you're at   like almost 10,000 MR monthly recurring revenue.  Yeah, I can tell you what it is, right? And it's  

[19:33] 8374 right now. Wow. It's 100 grand a year. We've  only been open for 13 days. So, Oh my gosh. Holy   crap. Okay, so 100K ARR in in 13 days. Like, are  you still getting residual signups? Are you still  

[19:50] posting about it? Yeah, so I'm starting to post  about it again. The first 10 days or so, I kind of   went super heads down just to make sure that this  was everything that I promised it was going to be,  

[20:03] right? Like, I didn't want to I didn't want people  to sign up and have a terrible experience. And so,   yeah, Ron and I kind of went heads down for about  10 days there, but I'm actively posting again on   social media and just spreading the word at  this point. Everything's super stable. So,  

[20:18] what does this look like? How do they get this  working for themselves logistically? Yeah. I mean,   the cool thing is you don't need really anything  other than an internet connection to make this  

[20:30] work. You can sign up and it it works out of the  box. Um, it's a it's an AI agent out of the box.   server. It's hosted in the cloud on a on a  separate server in its own container to keep  

[20:45] them safe. But can it like access their computer  and their files or No, only if they wanted to. So,   yeah, that's that's kind of the next step. We're  building a an official browser extension for them  

[20:57] to be able to connect their own computer. I'm most  hesitant about that just because I want to make   sure everybody is safe and I don't want anybody  to accidentally do something that they're not   intending. So, we're rolling that out more slowly  just to make sure that everybody understands kind  

[21:12] of the full implications of letting an autonomous  agent, you know, live in your computer. Yeah. So,   what kind of use cases are they using it for  today? Yeah, I actually have a ton of real  

[21:26] examples to show you. So, this guy built a golf  tracker app that does like shot by shot and you   have an AI caddy club analytics. It's actually  pretty cool. So, why do you need why do you need  

[21:39] an agent to build this? That's a great question.  You can definitely vibe code this. You can go   through something like cursor or something like  that. I think that this was born out of this guy's   love for golf, not I'm looking to make an app.  Yeah. Okay. So he's talking to his agent every  

[21:55] day about golf and how to get, you know, get his  golf game better and all that stuff. Okay. And   then the agent's like, "Hey, we should build an  app about this." So they interact with their agent   with Ron through Telegram and or Discord. Correct.  Yeah. Or the the the UI that comes out of the box.  

[22:11] So there's three main ways for them to interact  with it. Okay. So let's say uh I'm one of your   customers and I want Ron to go build me a a Fiverr  profile and start posting jobs. It can do that.   It just needs like some credentials from me. Is  that right? Right. Yeah. So, um the only thing  

[22:28] that for Fiverr specifically, they require  you fill out the like the W9. Obviously,   your can't do that for you, but you couldn't you  give it the info for your W9, so it could do it   for you. Yeah. I mean, I I wouldn't I don't think  that that would be an issue if you want to get So,  

[22:43] that that's kind of where the security thing comes  in, right? Like, do you want your agent to know   your social security number and all those things?  Yeah. So, but again, I think that's why the  

[22:55] container is the best option because you get to  pick and choose what it has access to. Okay. So,   you kind of treat it like chat GPT except it's in  a container and it's agentic. It can go out. It's  

[23:07] like uh it's a two-way form of communication. Like  I can ask chatg today to go do deep research on   a business idea for me and it's one way. It can  go out and do that and then a few minutes later   it'll return to me all the results. With this, it  can you can interact with it better and you can go  

[23:23] back and forth and it can go to specific websites.  You're not beholdened to like the websites that   Chad GBT wants to scrape. You can tell it exactly  what to do. Yeah, absolutely. You can go, it has   full web search capability. It can create its own  files. Something that I think that these agents  

[23:40] do better than chat GBT is they have persistent  memory. And so every conversation you have builds   on the last one. It it's compounding. catchy BT I  mean it remembers some things about you right but  

[23:52] like it doesn't always get that right um and it  sometimes gives you frustrating stale information   based off the idea you had two ideas ago yeah  the agents don't do that they they remember  

[24:06] current context on everything and if you decide  to kill a project or something they know not to   bring it back up yeah interesting yeah so it's at  its worst it's chat GBT on steroids At its best,  

[24:20] it's Chad GPT combined with vibe coding combined  with a junior employee that never sleeps with more   ownership from you. Absolutely. More control  and ownership. Yeah. You decide everything it  

[24:32] does. I think the containerized aspect of this  is is huge because I can make a project in clott   like let's say I upload all the transcripts or  all the newsletters I've ever written and say,  

[24:44] "Hey, learn my voice." Now write a newsletter  in my voice. Cool. But it's still going to   be positively or negatively influenced by other  things on the internet. It's not containerized. It   will still hallucinate. It will still write like  other people. It can't not. Yeah. But if I could  

[25:01] if I could just put all that all of my writings  in a vacuum or like my mom writes journals. If   I could take all of her journals and put it in a  vacuum and just ask it questions, then it wouldn't   be influenced by these outside factors. Yeah.  And to take that one step further, what it can  

[25:18] be influenced by is your conversations with it.  Guys, I hate to break it to you, but you're not   very likely to actually start a business based  on something you see or hear on this podcast,   unless you spend a little more time with me and  with other people that are doing just that. So,  

[25:32] come check out TK Owners. That's tkowners.com,  where hundreds of people are launching, starting,   testing, experimenting with business ideas  that they heard from this podcast. If you   want to go fast, go alone. If you want to go  far, go together. Check out tkowners.com. We'd  

[25:47] love to have you. So, not only is it getting  your voice from your 500 newsletter issues,   but it's getting your voice from the 500  conversations you've had with it. So,   it's it's going to mix and match pieces of your  conversation from pieces of your newsletter and  

[26:03] put all of that together to make something that's  actually yours. Yeah. This is incredible. I mean,   the original goal was to make 20,000 and you  you'll be there in like 2 months. if not less.  

[26:15] But not in the way you expected. Not in the way  I expected at all. I just think that's such a   testament to like I typed in my notes in front of  me in all caps with four ex exclamation points.   Post your crap. Like step one, do cool things.  Step two, post about it. Yeah. I don't care.  

[26:33] Like if you want serendipity, you want virality,  you want magic, you want money, you want stories,   you want experience, you want fulfillment, like  post your crap. Like tell the world what you're   doing. Just tell people what you're doing. Don't  hide your candle under a basket. Tell the world.  

[26:48] Exactly. Yeah, man. And I think that people get  really stressed about view counts and likes and   comments and all that stuff. I went into TikTok  basically doing it for me. I was like, I'm going  

[27:02] to do cool stuff and I'm going to document it and  if it only gets 200 views, great. Like, at least   I'll have kind of a running public journal of what  I'm working on. Yeah. And it blew up. So, well,  

[27:15] the thing about like 200 views for that video,  like if that viral video only got 200, you'd   still probably get a few customers from it. Yeah.  Whereas with like almost any other video in any   other niche on any platform, 200 views will bring  you nothing because it's low intent. It's only a  

[27:30] 30 secondond video, but this is so targeted. Those  are 200 people that want to pay someone money to   do exactly what you're doing. So 200 targeted  views are better can be even more profitable than  

[27:44] 2 million unargeted views or 2 million views of me  just doing a dance in my living room, right? Yeah,   absolutely. And one of the things that I like  about Tik Tok specifically is follower count  

[27:56] doesn't really matter anymore. Um, I know that,  you know, I think Gary Vee talks about it a lot.   We've moved from social media to interest media,  right? Like, you can start a fresh Tik Tok account  

[28:08] tomorrow and post something and it just hits the  people that are into that thing right now. So,   yeah, if you post a video about pickles, you're  going to you're going to get 200 people that like   pickles, right? Like Yeah. And you used to have  to have followers for that and you just don't  

[28:23] anymore. Like that's it doesn't really matter.  That's incredible. Now your costs on this 600 a   month for the servers. What else? Yeah, so 600  a month for the servers. The big cost that we   have right now is about two grand a month for  inference. So we pay tokens. Yeah, basically  

[28:40] L and tokens. Okay. So 2500 a month basically is  kind of our our COGS. You cover their tokens. Is   that going to stay the same? Cuz what if someone  just goes nuts with it and runs up your Yeah,  

[28:53] that that's actually already happening. Somebody  recently spawned 125 sub agents in 5 days. Whoa.   They've burned 1.3 million tokens in like 48  hours. Yeah, we're we're going to have to come  

[29:08] to a place where there's a cap on the free tokens.  Uh and then past that, we'll do like a fallback   model, something that's not as smart as the bottle  that we're using. Or you can bring your own keys  

[29:22] to your own chatbt or claude or whatever and then  use, you know, use that. Yeah, that's I like that.   Give them both options. Yeah. What's next for  this? Are there like upsell opportunities? Do   you just want to keep growing this? Keep posting  about this? Yeah, I think that there are a lot of  

[29:39] like upsell opportunities. The next big thing that  I want to do is do the same sort of thing, but for   people who are specifically interested in trying  to make this a business. We have a lot of people  

[29:54] now that like do art with theirs or it's their,  you know, de facto therapist or all these things,   but I think that there's actually a real market  for people who want to turn this into a real AI  

[30:07] employee. Mhm. And that's what I want to do next.  And I think that'll be kind of a small group that   can be kind of a cohort kind of thing so we  can all learn from each other together. Yeah,  

[30:19] that's amazing. I My brain has been sufficiently  blown. I keep wanting to call you Ron, but the  

[30:31] okay. If you sign up with a coupon, Chris, you'll  get 10 bucks off your first month. There you go.   Beautiful. All right, Robbie. Well, thanks for  your time. Yeah. Thanks so much. I appreciate it.   All right. What' you think? Please share it with  a friend and we'll see you next time on the Kerner