Ep. 87: We Built an Entire Creative Team With AI Agents. It's Getting Weird.

0:00:00Fast Hours hits 100K subscribers

Play from 0:00:00 0:00:00 What's up everybody? Fast Hours podcast, episode 87. Drew Brucker of Rory Flynn. What do we what do we have going on today? We just hit 100,000 subs on YouTube, Rory Flynn. at you guys. Look at you guys out there. Hell yeah. That was a we we made a concerted effort to get there. It was a it was an all hands on deck between us and the AIs to get to get there.

Play from 0:00:26 0:00:26 I yeah, for sure, dude. I think the AI definitely had something to do with it. Also, man, shout out to shout out to all the people that have been listening from the beginning, from the early days, from the maybe even the LinkedIn live days, if you know what I'm talking about. If you know, you know. shout out to our guests. Like, we've had a lot of great guests along the way that helped us get there. it's just been a lot of fun, man. I think like we probably would.

Play from 0:00:52 0:00:52 still talk and bullshit and text and have these calls regardless. But I don't know that we'd still be doing this without the support from everybody and just like those those classic like, ~ listen to your guys' pod, like love what you guys are doing, blah, blah, blah, blah, blah. Like, you know what I mean? And then it's from somebody unexpected or you never even knew, listen to it. You know, like moments like those. I know you've gotten those. yeah, it's that's the best.

Play from 0:01:18 0:01:18 It's like I ~ there's people that say I like your content, but when people say you like the podcast, it's like that's I like that better 'cause that's that feels like it's us, It's not just ~ my content, which feels like a separate side of who I am. No, it's just my writing and whatever I create, not like the actual personalities. I like people ~ I like dude the dad stuff's super relatable or like, you know, when when I s when people still bring up the fact that I ~ s you know, stepped on a nail, ~ Which is coming up on Yep.

Play from 0:01:46 0:01:46 it's I realized was too I I opened my it is just about September twenty fourth, twenty twenty four. So I I opened my my doctor's portal the other day and saw that message in there. It's like I stepped on a nail stepped on a rusty nail. Yeah, I had to I had to go get a I have to go for my annual checkup. So I was just like in there booking an appointment and saw that message and just laughed. ~ but if any whoever's been here since those days, you're real one.

Play from 0:02:13 0:02:13 Tier one. Tier one listener, right? ~ but yeah, dude, it's just like you know, now we can get one of those cool plaques. I've always wanted one, not just be just because like I don't know, I just want a plaque. I never get a plaque for anything. So that was yeah. I'm excited for for what we do from here too. I mean, I think like we've we've obviously been more consistent. We've got ~ we've tried to bring different guests on, get a lot of different perspectives.

Play from 0:02:40 0:02:40 Last week touched on something new. We've got some great guests coming up the end of the year. but like growing with you guys has been fun. And then maybe just switching topics, like we're we have. A lot to get into. I'm excited about today because, we don't have a guest, there's the exciting part of having a guest on, and then there's the exciting part of not having a guest on. And when you have like a week or two with guests, and then you have an open week like this, it's like, yeah.

Play from 0:03:07 0:03:07 Cause there's there's been a lot that has taken place since the last time you and I were here. And I mean, I we got we got two pages full of like just notes of things that we want to get into today. I'm excited, dude. I'm pumped for this one. two years ago we might have been ~ on an on a week we would have been nothing happened, what are we gonna talk about? And just make something up. Now it's like I don't even know what we ~ how we can leave things out because it's so it's so fast and there's so much.

Play from 0:03:36 0:03:36 And we say this every time, you're like Roy, you say this every every freaking week, every podcast. ~ it'll it's it's real now. If you don't feel it, you you haven't opened a social media account in two, three weeks. it's even weirder now. You go away for a couple of days. I spend a week my weekends are now I I don't work like I used to because, newborn here and it's just chaos. And I come back on Sunday night and I just look at X, I'm like, what did I miss? And it's just I don't know, everything.

Play from 0:04:03 0:04:03 Like million text messages about people talking about stuff. I'm like, Man, it can't even take two days to like tune out before I feel like I'm out of the loop. So it's fast, man. You know, now we got flying cars and shit. So Flying cars and shit. What a segue. Cause I was just thinking about my segue for that. Because ~ you I you mentioned something where it's just like, yeah, things are changing fast, right? ~ there was a point in time when we were kids where we thought flying cars would happen.

0:04:21Flying cars are getting ridiculous

Play from 0:04:31 0:04:31 And then there was ~ we all grew up and we're like, this is not gonna happen. And now we might have to rethink that, It's like this opens the door for that. And it sh it it's so. Shocking and unexpected in a lot of ways. Where now it's like, well, could you take these cars on the road right now? And there's gonna have to be a lot of like things that change as a result of this, right?

Play from 0:04:56 0:04:56 It's like, but but like I I I hope everybody's seen this. I would imagine like Elon's like pumping the gas, so to speak, on the algorithm with this video. I would imagine everybody's seen it by the time that we're showing it here, but ~ it's nuts. This is freaking nuts, man. One of the most got we got a car climbing a wall.

Play from 0:05:22 0:05:22 It's so absurd. Like, can we give a shout out to the production team also who made this look epic? Just like at night with the red lights and everything. Now it's, you know, upside down. Things just crazy that this is real. I know we have this. I know these cars existed because I've seen the cars that, you know, can stick to the walls and drive upside down, but I I didn't see the ~ you know, compressed gas like landing fly off the ramp thing coming.

Play from 0:05:48 0:05:48 I think. That's the the one thing that's that's interesting to me about all this is the ~ is that when you picture flying cars in your head, I think everyone went like the traditional route when you saw flying cars. It wasn't like that. ~ It wasn't like, ~ I'm gonna use rocket boosters, like rocket booster technology to land this thing. It was like, ~ it has wings or it's like a helicopter, like traditional like ways we think about stuff.

Play from 0:06:14 0:06:14 And then this one just comes out and's like, I I can kick flip now with like compressed gas. you're like what? So I mean this one's an AI video, obviously, ~ rendition, but I just thought This was interesting too, because it's like, okay, I could actually see that. you could actually have a dedicated ramp that goes straight up to the top of a building or something like that. You don't have to spin around 20 flights of a parking deck.

Play from 0:06:39 0:06:39 ~ I could see that. You know what I mean? I don't know. It's just, dude, this crazy times. I I I don't know if I've saved the full video. I know I sent it. but it had the kick flip that you were talking about. It did this zero to sixty. en point eight seconds, point eight. It's nuts. Well drive you you can see how it's possible driving an elect Drew and I both drive electric vehicles now. Something I never thought I'd say. Like I just never thought electric would be for me, but it's awesome.

Play from 0:07:08 0:07:08 Like I love it. The charging is weird, but the driving experience is great. And you can see how, you know, my 7,000 pound SUV gets zero to 60 in like 2.5 seconds. It's nuts. It's just like it just doesn't make sense. And there's no gear shifting and like there's no catching, it just ~ smooth, whoosh. And you're like, all right. So you can see how this can be achieved. But also I I read a cool paper, not paper.

Play from 0:07:34 0:07:34 I think it was post. It was a a a post about a paper. No, like a post about some sort of like scientific paper that was like the the thing they couldn't solve for was friction. Like to get zero to sixty in less than one second. Like the friction didn't catch like the tires didn't catch fast enough or it would just burn too much rubber. Yep. have it get there. So they had to that's why they solved it with the suction situation, which is just crazy tech.

Play from 0:08:00 0:08:00 And Grok is working on this, right? they're working on ~ Grok is being utilized in all of this. It has to be. because like why why wouldn't you? You have all the compute and you know in the world over there, and they can just throw money at the problem because they have so much of it. So like, hey, I didn't see it anywhere, but there's no way that Grok's not part of it. Right. Like it has to be. yeah. And ~ ev if you're and if you're a ~ a Tesla owner, having AI, you've got Grok in the vehicles, which I don't I don't know what other vehicle has that right now, because you've got this interesting hybrid of the person that owns Tesla that owns grok.

Play from 0:08:42 0:08:42 And it's embedded in the vehicles, right? So it's like you could you could do and then you've got grok bot, right? So it's like you can literally work while you're driving. And then you've got full self-drive. So you could literally just be sitting there thinking, doing work, while your car's driving. ~ I like I I just before I was a Tesla, I wouldn't be thinking about that. ~ it's just too far removed from me. But now that I'm I'm in it, it's like, dude, this not enough people I don't think know that's even possible right now.

Play from 0:09:12 0:09:12 And once you once you get into this electric side and and some of that stuff too, it's like how do you go back? ~ it it would be cool to own a second car that is gas. I was looking at a I was looking at a ninety-one Camaro the other day is like had the twenty-seven thousand miles on it. Or no, no, no, was a Corvette. It was a Corvette. And I was like, it's only 26,000. ~ I'd love to, I'd love to just have that, just as a second, ~ thing.

Play from 0:09:42 0:09:42 But I I I just don't know how you go back, right? Rory? we were just talking about the before we hopped on. Tesla and and Rivian rolling out these ~ Halloween, specialty ~ effects in the car, right? Like so you you've got that whole Ghostbuster thing going on. And ~ like Tesla's got this thing where it kind of turns your car into like a ghost on the road and everything else looks like pumpkins and goblins and all that. And there's like a trick-or-treat mode where you can turn it on.

Play from 0:10:09 0:10:09 So when you get outside the vehicle and somebody's approaching the vehicle. It it'll scare them and make all these Halloween sounds and it's playing Halloween music. So I I'm just thinking about this the other day. I'm like, my wife's like, So what are you gonna do? ~ it ~ we don't really get that much trick or treaters here. We usually take our kids elsewhere. I was like, Well, we're just gonna take this to wherever we trick-or-treat to. I'm gonna park it, turn on the Halloween mode, and then I can watch people get scared by the vehicle in my dash cam.

Play from 0:10:37 0:10:37 it's recorded. So it's like, yo, it's it's nuts. It's nuts, bro. It's ridiculous. And ~ again, I've this this video I was just showing on the screen. This was I I woke up in the morning and it was like, Do you want to go into Ghostbusters mode? I was like, Okay. ~ Show me sh tell me more, ~ what is that? And then I saw Rivian and they like all you did this commercial with Bill Murray. this whole spot. they they did a really good job with the promotion of it.

Play from 0:11:04 0:11:04 Yeah. And they did ~ but it's cool, the way they integrated it. they did a lot of like this whole retro video walking through what goes in here. just fun, a way to interact with it, right? And it's like, the car is now the Ghostbuster the LIDAR. They have little games, light modes, all these different things, which is this this goes this goes two different ways, Number one, super cool that they can run a partnership like that. Number two, I feel weird that my car is now a marketing tool.

0:11:26Cars are becoming software platforms

Play from 0:11:33 0:11:33 It feels weird that they can just be marketed to me with a switch of a button. Yeah. in it that feels gamified that you can interact with, right? cars might have had I I'm not a car guy, ~ traditionally. So I I don't know. I would assume there is software in the vehicles for a while, but this is a different type. Like sorry guys, I might be way out of my depth with this. You can call me an idiot, but the software is now visible to the driver.

Play from 0:12:01 0:12:01 Right. it's an interactive thing. Whereas before it's just okay, there's a soft there's software in here, but with Tesla Rivet, you have software updates and then you get things like this, and then you can interact with the soft and then there are games on your car. ~ it's just a totally different thing now. And it's just like that's not even being talked about enough. ~ update. It's like, we've updated self-driving. that's not, ~ software update for you know your standard gas vehicle.

Play from 0:12:27 0:12:27 That's like, I can drive better now by myself autonomously because I just downloaded a software update. which is kind of nuts and in itself, but it's it's a weird thing because like I said, I never would expect myself to be a electric car owner. Here we are. it's awesome. I love it, but you see this Tesla stuff. I said this to you before we got on. and I I just don't feel like anyone else in the US, manufacturing wise, is pushing the hardware as much as the software, right?

Play from 0:12:55 0:12:55 you look at China, Korea, Japan, ~ technologically they have much smarter devices and ~ they have robot. You go into, you know, I walk into the the airport in Korea. And a robot bumped into me and was Can I take your bags? I'm like, What the hell is this? ~ this is this was four or five years ago. And I'm like, they have robots that help you in the airport. And like we you know, we don't have you know, we have like, you know, metal detectors that don't work at our airports kind of thing.

Play from 0:13:23 0:13:23 And ~ it's like the hardware it's so different. Like I don't know if we don't put an emphasis on it, if it's just something that, you know, we're resigned to just import. But like you look at Tesla and like that flying car, that there's no other Car manufacturer that's innovating that here. There's no one, no other like real hardware company that's doing that. ~ software, yeah, we're all over it. But it's like when does hardware catch up? ~ cause that's the interesting thing.

Play from 0:13:50 0:13:50 Like, I'm sure China has you know, China look at the look at the phones. You know, you see all the things on social media about the phones the Chinese have, you know, Huawei, all that. like the different types of technology that they have from personal devices and, smart in their, smart homes, things like that. But, ~ it maybe it's just not a thing here, but yeah. from a creator standpoint or a democratization standpoint. If you could start because now a lot of the software, right?

Play from 0:14:15 0:14:15 It's you could clone it. ~ the moat's dried up in a lot of ways, right? ~ I created my own Calendly ~ a year ago. You know what I mean? ~ I so there's that part of it. And then even last week, we're just like, well, why? Why do you want to continue to go this path when you could find something or build something that is customizable for your specific situation? It's not this ~ all in one built for a big audience that wide ranging.

0:14:40Why AI is destroying traditional software moats

Play from 0:14:41 0:14:41 It's it's really now about the niche and the customization and the tailored aspect of it. And, you know, I think like hardware, just think about that, man. Like if we were to, you know, we have all these things that we've created or built or done with with AI. And then, you know, it's like I probably could have productized some of this stuff and you could have productized some of this stuff, but it's like it's probably the lunch is gonna get eaten soon.

Play from 0:15:06 0:15:06 So it's like I don't really want to spend days, weeks going down one particular path, not really knowing that it feels bulletproof or airtight in any way. But if you were to take that and then find some sort of hardware application for it, built into a hardware sort of environment, I could see that taken off. Here in the next ~ year or two. I could definitely see that being a So that leads me.

Play from 0:15:34 0:15:34 I'm just gonna bring this one up because this was this like took over over the last week. If anyone hasn't seen this, it's called it's a GitHub repo. It's called REA and I'm sure a lot of people have been doing this. But it's reverse engineer anything. This repo specifically, and it was able to break down a lot of software and ~ rebuild it and rebuild it really well, and rebuild video games and rebuild, you know, different pieces of you know, saw someone that like rebuilt Netflix to run it faster.

Play from 0:16:00 0:16:00 so it's like Things like this now, you have agents forms that can decompile software and rebuild it in its way. And I think this one I think is worth if I can find it. probably one of the best like responses, the CEO of Xbox came out, and like probably one of the most self aware and like brand aware statements I've read so here this is a note that she sent to the team.

0:16:09AI can reverse-engineer entire software products

Play from 0:16:25 0:16:25 And it was basically like, we no longer have an advantage, essentially, is like everything that we've done can be done now with AI. So, you know, is they're asking Xbox Labs, a small group of game makers and engineers who come whose job is to turn emerging ideas into things our players and developers can experience. we'll build them in the open, learn from what people want to do with them and measure success by what people are playing and creating. And it's also they want to take old games and redo them in new ways and give people the opportunity to interact with the development of the games.

Play from 0:16:57 0:16:57 So it's like a way to say, ~ you know, we've held sort of the monopoly on this, but we're not like living in this world where it it's like this is it. Like we s we own it and we're not gonna adapt to what's going on. Like they're gonna ad I would I would recommend everyone Take a look at this post, because it was from from Asha, the CEO of Xbox. I just felt like it was this is very self aware. And it just felt like they want, you know, players to become co-creators in the game.

Play from 0:17:25 0:17:25 Right. Like R Rourke talked about this on our episode a couple, you know, a couple weeks ago, right? Like brands need to bring the audience and the fans in to like build better experiences. So, you know, again, prediction from Rourke. There you go. This is one of them. We I think we all talked about this and wondered why this wasn't happening to begin with, but just something that I think worth noting, major player in the software space, in the gaming space, saying ~ let's let's bring everyone along for the ride.

Play from 0:17:55 0:17:55 We wanna we wanna ~ continue. Yeah. I feel like there's a lot of people in their shoes that are not willing to say that. They're not willing to admit it and they're not willing to almost just embrace it and then be proactive in weaponizing it, like pivoting. Right. that's gonna be the downfall of so many of these companies. And we were talking about this the other day, right? with the the Opus clip, the pod workflow stuff, these companies cannot move fast enough.

Play from 0:18:24 0:18:24 They can't move fast enough. They just there are too many people involved in these decisions, too many subjective things going on, too much at stake, too much you like there, whatever you want to throw in there, those things don't exist. the for the for the fewer headcount and the for the fewer red lines and red tape that exist right within a in a company or brand or team environment. And so These smaller companies, because you just don't have those problems and you can move faster.

Play from 0:18:51 0:18:51 And same with maybe even investors in that sense, because investors can drive so many of those decision making. And if they're not aligned, you're spinning your wheels and you're doing something for somebody else's sake. So the ability to be agile and fluid and move and adapt and pivot is really an advantage right now. There was also this like really good post I saw too, which was about like what are actually moats now. So Lenny Lenny, I think he actually he might follow the show.

Play from 0:19:19 0:19:19 Actually, I don't know. He might follow the show. but yeah, let us know. There's there's twice. Blink twice. There's he he called out what's actually not a moat, which I think could could surprise some of the people, the things that you see here. So distribution. That's that's interesting. I feel like distribution still could ~ be a moat. ~ it is dependent on algorithms in a lot of ways.

Play from 0:19:46 0:19:46 But if you own, if it if it's not rented, right, in terms of social media channels are rented. Versus owned would be something like a newsletter or something where you've got your own data and information. I think distribution could still have a little bit of layer as a moat, but execution, he talks about brand. Brand could be, but he mentions a lot of times it takes too long to build up for it to effectively be a moat, right? Yep. Okay. ~ understand that. And so he says, what are actually the compounding advantages here?

Play from 0:20:14 0:20:14 So switching costs, okay, right? So the stickiness of something. ~ network effects, which I think I've been hearing a lot more about in the last couple of weeks. This was something that used to be talked about a lot in tech startups, building network effects. And what are those things, right? It is like I I would even say I think switching costs like the sticky might actually be a network effect. but network effects is like the ability to, it's like the more people that you have on this platform or your tool or whatever, the more powerful it becomes, the less the decreasing, right?

0:20:32The competitive advantages that still matter

Play from 0:20:47 0:20:47 Like the the more it's in trouble, kind of thing, right? And there are several different types of network effects within a network effect. So it's not just that, but a marketplace could be a network effect. ~ so a dual-sided market, like a marketplace. I used to work at a company called Lasso. I was for event production software. So you're not only appealing for the marketplace of like, hey, we've got all these companies that are looking for hands for future events.

Play from 0:21:17 0:21:17 And then also we have a marketplace with all those people that are looking for actual jobs for these events, right? So that there that's a network effect. Like that marketplace aspect is a network effect. Scale economics. So the more volume a company has, the larger, more efficient its operations can be, this lowers unit costs. And then data effects, which is also another something that we're hearing a lot more about, I would say, in the last couple of weeks here, which is just like data, data, data.

Play from 0:21:43 0:21:43 You know, a lot of these. SaaS companies and software companies do have a lot of data. Maybe that's actually an advantage that they're not paying enough attention to. So the more usage a product has, the more data it generates, blah, blah, blah, blah, blah. But I thought ~ hadn't really seen a post on that yet. Definitely bookmarked, definitely interesting to talk through because I think ~ this is evolving in real time, We're we're sitting through the biggest shift in the modern world, ~ in from an economic standpoint that we've ever seen.

Play from 0:22:10 0:22:10 ~ it's it's happening. we've talked about it, we've predicted it for years, and now it's happening. Like right now. I think we saw it with four opus four point six. I think a lot of people were like, wow, this is different. Five point five and what we've seen over the last couple weeks with all of the LLMs, where it's like you go on your feed and you see the most ridiculous stuff you've ever seen. And it happens every day now. It's not like every month.

Play from 0:22:35 0:22:35 It's every day. It's like, what? How? And then the ideation around that. So you're you're seeing this explosion of creativity because the capability is now there and it's no longer what ~ what type of prompt do I have to write to get there? it's like, you know, if you have the right context and you have the right vision, you can now create whatever the hell you want. And it's true. I don't want like I honestly I don't want to hear it anymore of that AI is stupid and it's a glorified search engine.

Play from 0:23:03 0:23:03 ~ we've built an entire team for this podcast with AI agents and it operates and it works, right? ~ that's real. We've six months ago we talked about like what we need to hire to scale this podcast. Six months later, we don't need to really hire anyone because we've built the tools that we need to service this thing. So it's ~ I'm not saying it's a good thing. We wanted to hire people. But also we don't make enough money from this podcast to hire people.

0:23:11We built an entire creative team with AI agents

Play from 0:23:31 0:23:31 So there you go. ~ either we wouldn't hire people anyway 'cause we don't the money, or we just wouldn't do what we're doing now. So like there's there's two sides of that that coin and we're now this way. Yeah. ignorant and naive if you say that at this point. I think is is kind of what you're saying. And like that's what I'm saying when we talk about like not having energy to go back. It's just like I I can't like if you can't see that, it's not my job or any of us's job to educate somebody that I don't know and get them up to speed.

Play from 0:24:03 0:24:03 By the way, they may not even want to be convinced, right? ~ not not doing that. So, anyway, that's a great point. And and so I don't know if that's maybe even a segue into the idea in general, which we were talking about before the show, which is ~ creative explosion. You you coined that, which I think is like the perfect two-word categorization of what we're seeing here, we talked about the historical sort of timeline, images, video, 3D.

Play from 0:24:31 0:24:31 motion, And then now, like even with I don't know, like the meme explosion that we're seeing too with the I mean these videos now, bro. like we saw it with with the Migos thing, right? Like that, that everybody's seen the Migos thing by now, right? Swapping in characters. And then you had, you know, the the tip toe in my Jordan's. You know, like my wife's tired of hearing about it.

Play from 0:24:58 0:24:58 You know what I mean? with the Rumpel Stillskin movie from eighty-seven with this creepy little guy, And then, you get into where was where's the other one? the the Drake one, right? Where it's just like we're we're hitting this new ~ man, am I gonna be able to find this thing now? We're hitting this new era though. I know. I know. bookmarked more in the last three weeks than I've bookmarked in six months.

Play from 0:25:23 0:25:23 Here we go. ~ I love the Ted Lasso one. It's great. Right. But it's just like, dude, you're just like, this is a whole new layer of memeing, of whatever you want to call it. I don't know, what were you saying? Slop as a medium? Slop as a medium. Slop as a medium. as a medium. That's I I I feel like look, it's entertainment. What is short form content? It's entertainment. No one's on here. ~ outside of X maybe to learn something.

0:25:33The AI creative explosion

Play from 0:25:50 0:25:50 ~ in LinkedIn maybe, ~ Instagram, TikTok, ~ slop as a medium, it's entertaining. If you can make something funny or remix current people into into funny situations, ~ it like people love it. We're trying to be a little less serious about it. everyone will get in there and be like, this is like the You you're like the IP infringement, whatever. ~ it's fun. who's getting paid from this stuff? ~ it's just ~ it's fun.

Play from 0:26:15 0:26:15 I understand there's a lot of different ways you can think about it. But in its current iteration, it's not that harmful. I I I ~ honestly I think everyone now looks at things. I think we have a whole shared reality where we look at things and ha the first question now is that AI? Right? That maybe a year ago that wasn't the case. It wasn't like yeah. Now it's like my my initial assumption is that is AI.

Play from 0:26:41 0:26:41 ~ and then like I have to like go backfill to find if it's actually not, right? So it's like it's a general assumption now. But yeah, I think slop is a you know, I think slop is a really like interesting business model. I think there's probably a lot to be made off slop. I think there's probably a lot of opportunity in slop. ~ but who knows? Like who knows the may I'm I'm sure somebody's doing this for the Giants, but somebody's always doing this for McVay.

Play from 0:27:08 0:27:08 Every McVeigh press conference has got this unfiltered version that you see on social media the following day, Just like with the what he really wanted to say aspect thing, and it's just you you see this all over the place. Bring up that one example you had too of an Instagram account that was like really here it is. well as well, which is just like great. And like, you know, just man, just I love this stuff.

Play from 0:27:34 0:27:34 this one this one of stifler from American Pie. there's this VHS style, grainy footage. But you can tell, this is done in Higgsfield as it's it's, ~ i it's hashtag that, but like all this is definitely just one workflow that's run over and over again. Generate a bunch of images, it's running the same type of camera movement over and over, right? Like this is this is agentized now, like, but this format works. Like this is stupid. You know, like this is silly.

Play from 0:28:00 0:28:00 But then it's like, you know, you go to some of the other ones and they're great. ~ these these little ~ babies that are running running the show over here kids this is hilarious it's just ~ chaos but this is this is pure fun right this is totally different you have some of the other stuff here I was loving the I was loving ~ the the mask you know it's so silly and like who is anyone upset by this?

Play from 0:28:30 0:28:30 Right? And you know this is where we're at now. And it's pretty impressive that we've gotten to this point. And ~ the creativity side, like I said, it's getting this is this is fun content, but then there's also people that are, like this guy. If anyone's not following this guy on on X, ~ Japanese creator and Japanese the shout out to A I C R E A T A R O A I Creator O.

Play from 0:29:00 0:29:00 And ~ Jap If anyone's not on Japanese X, like they have some of the coolest stuff that like it's just a different aesthetic and they're very open with what they're doing. There's a lot of them are sharing. Like this guy is doing, you know, things between Blender, Excel, motion graphics, ~ and you know, using this all via Opus. But like the fact that some of this stuff is built in Excel, right? And then bringing it into Blender and then using it as like this whole 3D sort of motion graphic side of things, it's really cool.

Play from 0:29:33 0:29:33 And again, who's who is who is thinking about doing, Excel motion graphics mixed with Blender, I don't know, three months ago? One month ago? Right? So I think this is like the this is where everything is going. Now everything can be used as a creative tool. Right? This guy, like I said, has had some some of the things that have basically I've looked at and been like, what the Hell is this?

Play from 0:29:58 0:29:58 And it looks really cool because it's different. this is something I've never seen before, right? Just like the way that this sort of functions. I don't know if my sound is on or not, but it's just a really cool visual. And they're controlling it all, via Blender, the camera. You can see the camera movements and what's going on here. Some of the other ones that I saw here, there's one other one that I just want to show, and then I'll back off this this one. It was like just looked so unique, right?

0:30:26Combining AI, Blender, Excel, and code

Play from 0:30:26 0:30:26 I'm not used to this. I'm not seeing this medium. And then even seeing it as like a in the blender space, I'm like, it looks even cooler than it does flat. I'd almost rather watch it this way. But again, people are just thinking of unique things. All of your talents can now be sort of amplified. Your craziest ideas can now be amplified. Use any tool, you know, everything with code. I guess, you know, Claude really had it, or Anthropic really had the right idea when they were like, We're just gonna focus solely on code and like everything will eventually just be code.

Play from 0:30:58 0:30:58 And, you know, here we are. Right. So this is crazy to me. the other week? References and domain expertise. What else do you need? Dip context. Context, domain expertise, and vision. That's if you got it, I mean which kinda sounds like ~ that's a new formula, right? And it sounds like everything you've ever done in business, right?

Play from 0:31:24 0:31:24 Is that how it works when people get successful? Context, domain expertise, and a little bit of vision? Mm. Mm. Look at you finding a thread. I like it. I like it, man. very strange. But like we said, things are moving fast. I I'm personally just amped about this stuff because I cannot I wish I had like 48 hours in a day to just create because I I I can't let I can't like my brain the neurons can't stop firing.

Play from 0:31:56 0:31:56 I just want to go I want to try the Excel thing. I want to go to Blender. I want to go to, you know, I want to do the motion graphics. I want to go try all the new tools. I want to mix and mash them all together and you know create demo videos and can I make this a 3D world and then live in it and then build it into a website that's whatever you used to think was like ~ can't do this. Now it's just like whatever. Like the other thing I saw the other day, sorry to like keep going here. Friend Black Light, friend of the show, right, got his video game off the ground.

0:32:25The new formula for creating with AI

Play from 0:32:26 0:32:26 He built this whole universe, right? Like we talked about this and like the the the couple universes that he has converging in his IP and now there's a video game. Right. So now you can live it through the content, you can live it in the video game, you can live it through the characters and their social pages, you can live it through the short form content. ~ it feels more and more like like ready player one and ~ you know, living in these these little worlds that anyone can build. And it's it can all be connected and it can all be on brand.

Play from 0:32:54 0:32:54 It can all be, you know, similar messaging. Like it's crazy right now. This go back and listen to our episodes from two years ago. It does not sound like what we're talking about right now. just go back to the store episode. I still I still want the like the player suit from from Ready Player One. ~ Like the R yeah, like the armor that he that he has. Yeah, it's nasty, bro. I mean like The the whole thing is it's just ~ it's crazy. I created a video game the other day.

Play from 0:33:19 0:33:19 Well, kind of a video game, like a computer, like a I mean it's just interesting how it's probably more interesting how it came to be, really, which is just like you know, my dad's retired, 70 years old, 70, 71, 72, 71. And you know, it's not doing a whole lot, but He's got his backyard and ~ squirrel. He sees squirrels back there. They're like territorial on his fence and this and that.

Play from 0:33:46 0:33:46 He was over at my house the other day. And he's just like, you know, like this is going on. And I was no, it's cool. He's like, I was like, we should just create, we should create a game out of it. He's what are you talking about? I'm like, with AI, like, let's do it. Because my dad's never really like I've never sat side by side with him and kind of like showed him AI stuff. It's just like whatever he's picking up on. ~ and so I was like, let's just build it together. I just kind of took it as an opportunity.

Play from 0:34:12 0:34:12 I was like, let's just we're just gonna talk to Claude and we're gonna tell him what we want and like what's in our head, and like let's just build it. So like it started with this simple idea that okay, we're just gonna have like two squirrels like fighting over, you know, on a fence, you know, like grabbing acorns, whatever whatever it may be. Maybe there's a hawk. Right, that's tried to swoop down and grab one of them. Or maybe there's a sprinkler system that can knock them off, or a dog that chases them, or whatever it may be.

0:34:37Building a video game with Claude

Play from 0:34:40 0:34:40 ~ let's just take this to the next level. And so we did. And I spent way too much time on this. Way, way, way too much time on this for absolutely zero reason, including this page. So I've got this whole page. I've got my own domain now. I I don't even know what I'm doing. But I created, I've got two games. And I've got a third one I'm working on, which is more of a Mario Kart like Grand Prix ~ thing.

Play from 0:35:07 0:35:07 but this is the one that that I think I share with you, which is like this this stash game, which is it's dope. Like it's basically like I created all these different characters, right? ~ so create character sheets, built out all the the different movements, this, that, and the other. And I don't know how this is gonna work if I'm sharing my screen on this. I don't know if this is gonna load right or or whatever. But ~ really just fun, ~ putting this together, thinking through the characters, like tree house.

Play from 0:35:36 0:35:36 You we're on top of a tree house. We've got this guy who's, who's power lines. You got a flock camera here. just really just had a fucking blast with this, right? And it's just like it's fun. And I and I shared it with my dad and did this whole thing. Let me see if I can play this while we're on here, if it'll let me. Right. So basically the idea here is I'm just grabbing as many. of these things as possible, taking them back to my stash. Yep.

Play from 0:36:01 0:36:01 I can toss ~ I can do like I can knock this dude off. Boom, boom, I can climb down here, go down here. Yeah. No, I w I I worked the hell out of this thing. I I think I went through four usage limits trying to get this fully here. ~ I was like, no, we need to make this looser gameplay. Like they need to move faster, right? Like there was a lot of iteration that that happened here. But it was just really fun. It was really fun to do.

Play from 0:36:29 0:36:29 And I'm glad I did it. And I had never created, I had never gone through and done the the whole video game thing before. but it was fun. Like it was kind of I didn't know it, you know, what the right tools were. I I but I figured it out. Right. And basically it's just through creative expression of like, no, I really want to push the realism on this. I want to, I wanna incorporate like a water sprinkler coming up. I want to have like this where I can change the season, Where I can actually change the time of the year, but this sinks to the calendar and so it's got seasonality to it.

Play from 0:36:57 0:36:57 So if I change this to December, it's got snow and there's no leaves on the tree. Right. And if I change this to fourth of July, like it feels like there's a backyard thing going on with fireworks. I'm like, whatever whatever it may be. And so you could just you get lost in the sauce and this stuff. And this is like this is the creative expression that we have now. And then you were talking to me about like, okay, like. It feels a little Street Fighter. So I was like, dude, I need I was thinking about doing a more of a battle.

Play from 0:37:25 0:37:25 Right. So then I created a whole, battle version of this game. Right. So, so, so there's knockouts and everything, right? there's special moves that you can do. What Nice. the I think there's finishing moves, ~ whatever it may be, right? Falls off. Okay. You know, like It's incredible. It's just, you know, I was just thinking, like, what what maybe I just build like maybe I make this an app like and I can play it on my phone or whatever it is.

Play from 0:37:54 0:37:54 There's a daily scoreboard, right? Like you share it. Like I don't know. It's just, it's just it's just crazy the kind of shit that that you can do now, right? And ~ yeah, this is just like one of those like fun little fun little projects. the Grand Prix, the Squirrel Pre one is not quite ready yet. So I'm not gonna share that. But It's like the Mario Kart version, right? You shoot an acorn, you can knock somebody off the path.

Play from 0:38:20 0:38:20 They're running through the fence, through a park, under a bench. There's cars coming by. You're running under the cars. Cars are honking. ~ it's incredible, bro. It's incredible. It's a whole little wor again, whole little world. Speaking of the the flock camera, did you see the did you see it sounds like a it sounds like the plot of a sci fi movie, but did you see the the guy that was torturing the flock camera? Okay, so he was like figured out a way to torture the flock camera by hacking into it and showing it twenty seven thousand images per second or something like that to ~ scramble its brain.

Play from 0:38:53 0:38:53 And it was like this it was like this sick ~ inverted torture mechanism, ~ the humans torturing the AIs. ha it's like the weirdest th again, right? Yeah. concept that we had never seen before, really, that exists as the underlying thread of that. That's why that to me is the ~ why that story is really interesting, right? Cause we're getting into all these new type of moments that maybe you probably either dreamed or saw through Black Mirror, essentially, right?

Play from 0:39:27 0:39:27 I it's here and it's it's not going away and it's only gonna get weirder. the flock camera. It was I was just I saw that and I was like, this is this is wild. But the speaking of that stuff, not to go down that rabbit hole, but there was other I saw once I clicked on that, then I my out my X algorithm started shoot me showing me other stuff. And there was people that were building little torture chambers for for AI. It's like really sick stuff, you know. Yeah, like building like little like training environments and like locking them in there and just like peppering them until they just go nuts.

Play from 0:39:57 0:39:57 It's we it's weird, man. it literally sounds like the next Netflix like hit series, like when the humans fight back or something like that. ~ you know, and that's like a very sick and twisted, like it's still a sick mind that thinks of that. You know what I mean? Like someone's got a someone's like thinking about that unless they're having AI do it themselves, which is even weirder. Like you know, AI on AI crime. ~ like where it's like, how could I torture a flock camera?

Play from 0:40:25 0:40:25 You know, Crocs like it's like, hey, how about we just show it twenty seven thousand images, it'll scramble its, you know, neural network and it's like, you have to push send, let's go. Like how else would you program that? I mean, I d I think you unless you're like a legit I and again, I'm stupid guys. I don't know the answer. If if if you could code something like that, you probably could. But to me, this is just like that though, it it's just like even if you don't know, you figure it out now.

Play from 0:40:53 0:40:53 You can figure it out. Anything now you can figure out. I mean, at least to a degree. I built this other thing. I did I did this other thing. I was listening to the podcast. I was listening to Do you listen to Modern Wisdom with Chris Williamson? I I do was I watch on YouTube. Or at least I put it on in the background and listen to the conversations that I'm interested in. Okay. Okay. He had ~ I forget, I can't, I can't remember how he says his last name. this guy that's a CIA, he was a former CIA guy.

Play from 0:41:18 0:41:18 He you've you've probably seen him around. ~ but he was yes, John Kiriaku. There you go. He he was telling some awesome stories. Awesome stories. He was he was telling something about how they were looking for some guy for a long time and ~ Basically, they sent this guy that he knew to this one country and they were looking for him, but they, they they weren't sure he was actually there.

Play from 0:41:45 0:41:45 And so after a while, they were like going to the market. They were going to this local market and they and the one guy thought he was with some other guy that I guess he was living with. And they s they he was like, dude, I think that's this dude. And the dude's like, No way, no way. Are you sure? He's Yeah, I'm pretty sure. And so they went and told, ~ I guess his his boss at the embassy or wherever they're at. And they're like, Hey, I'm pretty sure I saw so and so. They're like, Well, I yeah, but we can't really do anything about without proof.

Play from 0:42:13 0:42:13 we got to get proof. So, what I want you to do is ~ next time you see him, try to get a picture of him, right? And so it's like he kept going back to that same market for, I don't know, like a month, month and a half. Finally saw him again. They paid off. Some ~ locals to basically get into a fight, pretend like they're getting into a fight because they knew culturally fights at the in this country at the I can't remember if it's like Yemen or something. It was just everybody would stop what they're doing and basically make a circle and pay attention.

Play from 0:42:42 0:42:42 And so they saw him basically weave his way through the crowd and start to look, And while he's looking, they took some pictures, they figured out where he lived and all this stuff. And then basically they crushed up a bunch of like. glass really fine, get it got it into his toothpaste. And he was brushing his teeth, right? And so starting to get this. And so they had already talked to his dentist who he would go in to see, for this.

Play from 0:43:08 0:43:08 And then they long story short, they put him under and then extradited him, you know, while he was under. And he woke up. And I was just like, yo, this is crazy. But but he was throwing around like stories like that. And I was like, man, like I always find the CIA stuff so fascinating. So I was like, Yeah. Man, can you put together a daily thing for me to ~ teach me how to be like a CIA, ~ employee operative, right? ~ yeah, okay, yeah, that guy too.

0:43:13Turning curiosity into an AI learning program

Play from 0:43:34 0:43:34 And and so ~ okay. So so they put together, ~ I had to put together a whole, I don't even know. Man, do I have it up? I might have it up. But I put together a whole It's not a skill. It's like a it's basically like a program I'm going through. Right now. Hold on, let me see if I got this. It's not pretty or anything.

Play from 0:43:59 0:43:59 But basically it's got, each thing's got a learn, it's a live drill. ~ there's a quiz element, involved. Blah, blah, blah, blah, blah. so I'm like, just tr ~ what I don't know. What do you want to do? ~ Universe and CIA University, you'd you've been you've been having a good week. just all over the place. you see where my brain's at. I'm just it's just like that's where we're at.

Play from 0:44:25 0:44:25 It's just like you you wanna do it? Do it. Figure it out. Yeah. Well, that was a funny thing for the website, right? Like I I saw your idea last week for your website where it was like, you know, give me the ten wow factor ~ you know, ideas. And I was like, That's a cool idea. Never like really thought about, you know, doing that. But if I go into the ~ let me see if I go in here I could find the ~ I went nut there's something that I went nuts on and I probably shouldn't have gone nuts on, but I did.

Play from 0:44:56 0:44:56 And then what we're doing here. I feel like we're getting lost in the sauce on all this stuff. And that's where the ADHD brain ~ goes. But you know, like going back to this, I don't know if you do this too, but like I've done this enough times now, like with that squirrel game or these other little things, even the website, where it I go back and forth between both spectrums. I want wow factor ideas and I want you to red team the fuck out of this.

Play from 0:45:24 0:45:24 So I'm I'm looking for all the weaknesses, blind spots, voids, bugs, airs, contradictions, whatever it may be on one side. And the other side, I'm looking for these novel, imaginative, playful, wow factor, like insane take this to the next level type of shit. And I'm just going back and forth. You know what I mean? Like it's just it's it's sort of like this weird process that I'm I'm running with a lot of the creative sort of applications ~ going.

Play from 0:45:54 0:45:54 to both sides of the spectrum. I don't know if you're finding yourself doing that as well. Hundred ~ I I don't even know. I have to stop myself. Sometimes I'm like, just just don't do this. ~ I get to I get halfway through and I'm like, just don't, just don't do it. Because you're gonna spend the rest of your day and you have other things to get done. You know, and like you're gonna get obsessed, right? I know that's the problem. I've I have to self I have to self-regulate or else I won't get the million things done that I have to, which I I've over I've overburdened myself with work because I can now and that's the problem.

Play from 0:46:23 0:46:23 Right? It's like that's where we're at. I think probably a lot of people feel that where it's like, well, I could do that because I can run it through, my pipelines or my agents. And it's like, do I take this on? Yeah, maybe I do. And then you get into it and you're like, Why did I do that? And then you're like, now now I don't have any time to play, which is kind of what I want to do. But this was a like going back to the website, like what you said, like the you know, give me the ten wow factor ideas. So I basically started this chat like on our on our website project.

0:46:44A new approach to creative ideation

Play from 0:46:51 0:46:51 And I was like, you know what? You know, Opus, how about you just You just come up with some I wanna come up with some crazy ideas for this. But, you're you're basically gonna act as ~ my web design agency, but I want you to but put together a team. But also ~ make sure the team is made up of I wanted to test the different models. So, make sure you got Fable, Opus, Sonnet. Yeah. because I I I I this is something I hadn't really thought to do, but that's definitely something I want to try in the future.

Play from 0:47:19 0:47:19 ~ in terms of like almost assigning it based on the model about what's possible, what what would be the I mean, the idea I guess right model for the job, but this is sort of like a creative application of that. Yes. Yes. And then also it's like, there's some some models like you don't need all the thinking power. Some of the simplest ideas might come from the simplest models. Right. So maybe haiku. ~ doesn't like it just it doesn't go too extreme, but goes right in the sweet spot.

Play from 0:47:45 0:47:45 But I thought this was was interesting where I was like, basically ~ like the I want you to too? Yeah, let's let's do that. Let's I was like I want you to like basically assemble a team. But I want this to be cracked web designers and vibe coding nerds and terminally online designers, again, utilizing different personas, ~ creative ideation coming from multiple models. It's like structure as many agents or the operation however you'd want.

Play from 0:48:12 0:48:12 and like cause the first couple things were all just like based on our transcripts, like what we could do with our text. And like that was it. So it, you know, it came up with this whole team. You know, it has like you know, shader wizard for Fable, X Arcade, you know, game UI designer, Fable. ~ you know. This one's good in 3D renders, fluid effects, site hardware. We get a studio creative, ~ you know, motion language. We have one that has art exp you know, art director experience.

Play from 0:48:39 0:48:39 Sonnet is a you know audio reactive visuals, UI sound design. We have Sonnet, who's great at feet physics, gestures, phone first. Pixel Gremlin, haiku. I like that one. That's terminally online. You know, Easter eggs, cursor toys, secret modes. ~ like this was fun, right? Like if Haiku wasn't there, it wouldn't have gotten, you know, to be the persona of Pixel Gremlin, 20, 20 years old, terminally online.

0:48:41Building a team of specialized AI agents

Play from 0:49:06 0:49:06 And then, you know, vibe coder, right? So like what it did was it just sort of like ran its own experiment. Every sub agent sort of brought their ideas, and then, you know, as Opus here sort of took all of the ideas and then they passed through like refinement, and then it provided me, you know, with ideas, right? And it remixed it three times and then it Verdict, and then it built out this entire artifact of different things that we could do, like for the website.

Play from 0:49:34 0:49:34 Like one of the cool things, and I'll show it because hold on, let me pull it up now. So I sent this to Drew is like I, you know, I I don't know anything about AEO, SEO, really. ~ but like here was the let me pull this up. I don't know if this I this is first off, this is like the I had it ~ structure this whole artifact of ~ all right, here's a three hundred and I've ran this a bunch of different times with a bunch of different teams and it came up with all these different ideas.

Play from 0:50:05 0:50:05 Right. So we have all of the ideas here that it came up with cataloged and it's all saved. we have you can filter it by our Easter eggs. There's some fun ones like, you like snoozing the website, like an alarm clock, and then what would happen with that one. ~ you know, like there's some really fun ideas when you go in through this and there's give Fast Hours MCP, something we might consider here. You guys can tap into any anything you ever want, anything we ever set in here. Which is like there's some cool stuff that it came up with, regardless.

Play from 0:50:34 0:50:34 One thing that it did come up with, which is more business focused. So let me see if I can find this without having to scroll through everyone's screen or have everyone watch my my screen scroll. This is the I had it ~ because we're our domain is hosted on Cloudflare and Cloudflare is just really good with the MCP. ~ and where did I just put this now? Cloudflare is really good with the you know, with the MCP with Cloud.

Play from 0:51:02 0:51:02 And then I had it sort of like track. And this is funny, not funny, this is like interesting, right? Because we're we're tracking sort of like how our site gets hit by AI. not just the search engine optimization, but like how AI is hitting us. And this was run the site was live for one week, right? Like one week. It's not like we've been up for a long time. But someone asked an AI 54 times about us. And then also you have people who were coming through OpenAI and EXA.

Play from 0:51:32 0:51:32 I don't even know what EXA is. Apparently it's another LLM. But the interesting thing is not just from AI search, the training, how much training is getting done on our website. Now, the big outlier here, Meta. Meta had 20,000 visits in a week to our site to train. Right. So ~ anyone who's digging it's like these these training runs, they're just out there spamming the internet.

Play from 0:51:58 0:51:58 Now, what else, not to not to mention the last week or in the last day or two, what did Meta just release? Like they might have been doing a big training run, scraping everything on the internet. Right. So this is this is interesting. And you see how much your site gets hit for training, where they're coming in, what they're looking at, how many times they're rereading stuff. Right. So it's interesting. ~ you know, just to build like have to be able to build this type of stuff for us on the back end to understand that we can tailor our whole website strategy now to get more of this.

Play from 0:52:32 0:52:32 So when people ask on the AI, we're gonna be quoted, people are gonna come to us, we'll get more listeners, we can do more stuff, maybe we'll make some money from this podcast, guys, so we do more fun stuff for you. But like this was again, this all started with that random, just like. Drew said, give me wow factors. I said, All right, build a weird team to give me weird wow factors. And then I just implemented one. I was like, what's the easiest one you think has the most ~ cool application, real quick?

0:52:39Tracking how AI crawlers interact with websites

Play from 0:52:58 0:52:58 And it was this one done in five minutes. Now we know. Right. So again, ~ I love that. There's still some weird things that you can ask for out there. ~ I'm I I don't want to sit here and say ~ prompting is this ~ super important thing. I think it's more about the the frameworks and how you're thinking about it. But there's another one that I've been running too that so ~ this idea of a skeptic, right? So here's what I was like doing even with my site and and the even the game where it's like be a skeptic who comes to my site.

Play from 0:53:28 0:53:28 You're someone that hates me. You look for flaws anywhere. How would you revise, tweak, add, remove? What we've come up with here in order to leave you, the skeptic, as someone who's blown away with the light, completely changes and reverses your perspective about me and my skill level, and has no other choice but to proactively reach out to me and praise me for this. Like, yo, and it just w it went so hard, you know, because like then it's like it gets into not only like the red teaming aspect, but it's like, How's somebody gonna critique me?

Play from 0:54:02 0:54:02 you you're almost like you're getting ahead of that. And it's just like I before maybe even last week or a week or two ago, I I hadn't really even thought to to do go about it that way, you know? And it's just like you, you know, just like these different pathways that still exist and sort of like the creative application of how you execute this stuff. I mean, ~ I just thought like even that gave me different results than than a red team, or you know, like it's just so many cool things that you can do.

Play from 0:54:29 0:54:29 And ~ ~ we were going through man, I like I don't even know that we've really given enough love to our entire process now with the pod and how we've done that, right? Like we can't we came along like we there was that point in time too where I I think we built like that Jeep we built like a GPT for the pod that had all the transcript information in there, right? This was like I don't know, year and a half ago, two years ago.

0:54:51Using AI skeptics to improve creative work

Play from 0:54:56 0:54:56 Where it's like you could tap into anything that we said, how we felt about it, this, that, and the other. Now it's just like, okay, we're gonna record this podcast. All right. now what we have, and we've touched on certain parts of this in previous podcasts, but this flow is insane. And we'll I think what we'll do in the future is try to we'll provide a visual for this, which will make more sense. But here's what we got now. We're recording the podcast, okay? We've got Riverside MCP.

Play from 0:55:21 0:55:21 ~ I could just provide, ~ it's already got the episode number in Riverside. So all I'm doing is I'm running an entire ~ skill. system called YouTube that's going to give me the title suggestions, episode descriptions, chapters, tags, hashtags, targeting, ~ Yeah. a few other a few other components in there. Okay, got that. I also got another thing that I run.

Play from 0:55:47 0:55:47 Yeah, chap yeah, chapters. And then ~ I've got another thing that basically takes that same episode and breaks out Here are, ~ 10 LinkedIn posts, here are 10 ex posts. Here's, on certain topics that we've hit. And here's a newsletter post. And this, whatever. I don't even use them. I don't even have the time to use them yet. And then you've got, so then you've got that. Then you've now got this system that we've talked about last few episodes, right?

Play from 0:56:14 0:56:14 Where we've now incorporated the ability to just habit pull clips, better clips than we get elevated from, even inside of Riverside. Right. That ~ Riverside had magic clips, right? And half of them would be good, half of them be, I wouldn't, ~ no, and they grade them, right? Opus clip would grade them and this okay, but they're they're not reliable and consistent enough. Now we've got this that clips them, adds in the motion graphics from Opus.

Play from 0:56:43 0:56:43 Even last week, then I took that, right? And it's like, we've got all these great clips, but now like there's a branding opportunity from a thumbnail standpoint. I'm just like, can we do something with that? So now I create a whole system that layers on top of you give me the clips. Now I feed those in. It's gonna, it, it basically locks in our characters and goes and put those together. And I ran that system with with Phil last week and we didn't even have a character sheet for Phil, right?

0:56:47Our AI-powered podcast production system

Play from 0:57:08 0:57:08 Yeah. So it's like we needed somebody something new for that. Put that together. And those were like straight hitters, like one-shotted. So now we've got the thumbnails for the 916 YouTube short Instagram reels. Then we've got the website component. That goes automatically onto the website. You've got all the data, the information that comes from that, or everything that we've said, everything you could talk about, ask about, AEO, SEO value that's going in there.

Play from 0:57:36 0:57:36 And then we've got ~ there was an artifact to component that I put together for this that breaks out the episode in a story format, includes the clips, embedded clips. ~ this is an insane workflow, right? That just keeps building on itself. Most of which was not possible six months ago. You know what mean? ~ wasn't possible. You just talked about it. We were talking about how are we gonna do this?

Play from 0:58:01 0:58:01 we probably need to get sponsored to to fund to get somebody in here that can help us with the clips, could do this, could do that, right? And even then we got to train them on what we do, our boy, like how we do all these things. It's not like we could just hire somebody and get them get them to work right away, right? We would need to. Hey, this is how we go about doing it. Here's our process. We'd have to give them access to things, this, that, and the other. It's just like, dude, we solved it. Fucking solved it. And maybe if we have a employee now, they can just focus on making the show better and not like having to do the ancillary stuff.

Play from 0:58:32 0:58:32 Right? if we were gonna have ~ a ~ a a junior employee that was gonna do a lot of stuff, right? They were gonna have to do a lot of stuff and you'd have to teach them that and that would be, I'm not saying this is bad, but maybe that junior employee can just focus on making us more viral on YouTube shorts, right? or something like that. ~ that could be their whole role. I don't know. But the the idea that it started with a few things, we started getting the the thumbnails automated for the episodes. That was that was step one in terms of ~ automation on the creative side.

Play from 0:59:01 0:59:01 You had the you had the episode logistics run through. Yeah. Hours. I was doing the I was doing the thumbnails for the the regular podcast episodes. And each time I was doing that, right? And I'm a I'm a perfectionist when it comes to putting things together. It would just take me way too long. I I'd have to, okay, what image is going in the background? What's the thumbnail text gonna be? and I'm already doing the the podcast export and trying to edit.

Play from 0:59:29 0:59:29 And it's just like it became too much. And when you started doing it, You were ~ doing the same thing, but then you you found the consistent you you found the characters. And then you started being able to codify it. Right. And then we j it it it just evolved and evolved and evolved and became basically, not not completely automated, but ~ that went from I don't know, six hours down to thirty minutes for us now.

Play from 0:59:56 0:59:56 ~ it it's just that and that's one part of it. And then now AI's unlocked us to do all these other things which we just Couldn't even imagine being able to do ourselves before. is just ~ ludicrous in terms of time and cost. Like what that takes, it runs in about two hours. It's just nuts. two months ago, even a month and a half ago, I was just starting to get enough time back to start doing the clips. But going back to the magic clips, it's like, well, they're not all great.

Play from 1:00:23 1:00:23 So then I had in, those YouTube flows, it would pick out timestamps of things that would be better clips, but I'd still have to go and review them. I'd have to find them. I'd have to go in and edit them, right? And that would take me hours just to get a few clips out. Now that's gone. That's completely gone. ~ I I I'm just blown away that this has all just taken place in the last ~ month, And the other thing that's wild, right?

1:00:51Cutting thumbnail production from 6 hours to 30 minutes

Play from 1:00:52 1:00:52 Like is we have not only because like we've we've catalogued eighty-six episodes into transcripts and into pages and blog posts on the website, which Claude did in an hour and a half. Now it has the, you know, it's run the it can it built its own routine and skill to go do this every time the episode drops. So like we have it run on Sundays when the episode drops. It built its own little agents in Cloudflare that can just talk to it.

Play from 1:01:18 1:01:18 And it just it pulls the transcripts, it follows the formula, it gets the page up, the page is live. We've we push it, I it I just say push live, bang, new pages up. ~ I can't imagine having to do this f all the time. this would it would just it just wouldn't get done. Again, it's things that wouldn't get done that are now getting done. And it it's I'm I'm the closest I've ever been to hitting my max plant, my my Claude Max subscription like percentage.

Play from 1:01:49 1:01:49 This week I'm at ~ eighty-nine percent by Friday, which is ~ I still got a lot of things I gotta run. But the amount of stuff that's already been run on it this week is l it just again, for two hundred dollars a month, this is it's a joke, like of how much stuff has been produced this week. Not to mention, you know, 3D software that's running in the background right now on four separate, you know, four separate sub agents building four separate things in three D.

Play from 1:02:16 1:02:16 This, the clipper, the website. ~ you know, my other three projects that I'm running simultaneously, some of the other things I'm doing for my contract roles that are now like I'm checking in on this stuff to make sure it's getting done. But the weird thing is like the code, like the code or the context switching here, where it's like I've never been more like bouncy back and forth between things where I'm just like bang, bang, bang, bang, bang.

1:02:37Automating podcast publishing with AI agents

Play from 1:02:42 1:02:42 And it's like I get lost some days. I'm like, what did I actually do? I don't know. I even know what got done and then I go back and look at it and I'm like, Holy crap, I got two weeks worth of work done today. ~ I don't even know. I don't even know, it's like you black out. You're like, my god, this is getting done. Okay, good. Let me go do this. ~ a even from a contract client perspective, right? It's like, okay, I I've got a I've got one client that is marketing related work.

Play from 1:03:11 1:03:11 I'm doing, you know, a lot of like marketing skills, agents, that kind of thing. But other marketing teams would probably have those same needs for those things, right? It's like, so once you go through and do it once, again, it's like the skill thing. Like, What if I have another client that is a look alike client to that? I don't have to spend the time that I did with that on the next one. Right. So I think it's ~ two thinking about it in terms of ~ having finding those things that you you maybe want to lean into and maybe pursue.

Play from 1:03:45 1:03:45 And you're you're probably gonna experience those same things again. And it's that same mentality that we apply for the personal stuff, but also on the business side. it's like if I'm creating an agent to, you know, go through, I don't know what it may be, but like there was one like with ~ like a content auditor, for example, where it goes through every page, audits the content against like positioning that they have, messaging, understands like what that page is supposed to be for, audits, you know, does it meet the expectations of what this page would be for?

Play from 1:04:17 1:04:17 What are the things that we care about here? Right. It's like solve that once, go to another client, have that same thing. I have a white label version that I could just apply. Right. So if I spent, I don't know, five, six hours going back and forth to refine it and finally get it to a place. And even then, right? Five to six hours, Rory, is like, how much of that time is actually us time that we're doing that? Maybe 30 minutes going back and forth, right? ~ yeah. So, it's like working smarter, not harder, too, in that sense.

1:04:45Building reusable AI workflows for clients

Play from 1:04:47 1:04:47 It's like, okay, if I'm gonna do this, man, maybe I should probably pursue something like this again, just so I can. bring more time back into my space because spate time is is the bill and it is the the competition of actually what I can I can do and take on, because I'm tak I'm trading time, for for for service and money, right, in a lot of ways. And so if you can remove yourself from that, right, it's like a, it's another way to to think about So the this one this one project I'm running right now, I'm doing it almost as a test because it's something that I've done before many times.

Play from 1:05:22 1:05:22 But I've the I'm running the whole thing through the entire pipeline that I've built for my whole business and seeing if I can ~ how much I'm tracking how much like manual time I put into it. And as of right now, the only manual time is me writing the actual emails to the clients because of like what I'm seeing, and then like Finding errors and like that's it. So like everything from initial like scope proposal to the initial proposal email to the contract negotiations, all that.

Play from 1:05:51 1:05:51 It's all been run through certain pipelines. Project management now has been like we talked about with Phil last week, that's all been structured in a way that like I have my tasks, I check things off. We you know, the agents sort of work off that. And then, you know, it's building workflows and you know, what it's doing is I have testing protocols for what it needs to do and what it needs to look for and different, you know, agents that are looking for errors and how they're looking for the errors and then have finding edge cases and all these different things. So it's running to, ~ for fast creation and iteration, it's running to magnific MCP for the webs for the flows that it's building, it's going to weave.

Play from 1:06:23 1:06:23 And then whatever results it's pulling, it's building in its own Figma page so that I can share that and say like here's what's going on. Right. So it's it's running itself. And there's certain times where I find it where it's, you know a problem in weave might be, for example, like, the inputs, the inputs are, you know, like the the products are scaled differently so that when they come out in the outputs, they they're scaled, you know, they're they should be because it depends on like what you put in on the input.

Play from 1:06:53 1:06:53 If it's a you know, if the the product should be really small, but it takes up majority of the frame, it's gonna make it look bigger scale wise in the output image. And it was like, you know, if we're trying to make this completely hands off, like I would have to build a skill for myself so that I could edit the images. I'm like how else can we do this? It's like, I can build a custom node through your custom node ~ skill that I found in GitHub and we can, host it on replicate and we can just put it in there so they don't have to worry about it. And I was like, You're remembering things?

1:07:01Running entire creative projects with AI agents

Play from 1:07:21 1:07:21 I don't even remember that I had. and it's just like, okay, we'll just do that. Cool. ~ so you know, it's cool to see that, but then also like it'll f it'll do things where I'm like, you know, it's like, I built a, you I built in the Topaz upscaler nodes that we can make this like the correct specs that we needed to for the final output. I'm like, just use nano 4K. It's like, okay. It's like you don't need to do the upscale to get it to 3000 by 3000, right?

Play from 1:07:47 1:07:47 Because it was doing it at 2K because we were testing and it just like thought that's how it had to be, right? also it's just like that like going back to ev even a few weeks ago, ~ like and maybe it's just because I haven't done a lot of I hadn't done a lot of creative projects in a while. There was a gap. But when I was doing the website stuff and then the game stuff, you know, magnific MCP, you've got runway MCP, where I've got unli you know, unlimited credits.

Play from 1:08:17 1:08:17 And it's just like use this for this, use that for that. And it's doing them both simultaneously. I don't even have to worry about generating the images. I don't have to worry about the prompts. I'm just giving it the context to go and do it. And it comes back and it does it. Right. even with the the website stuff. ~ I, I don't know, I got to this component where I wanted to do like when last time I showed this, this was just like a screen and then like one s thing.

Play from 1:08:43 1:08:43 And I was like, I, you know, I kind of want to do like a flat lay. I was like, I think that would be clean because I could I could almost lay out items that feel relative to the project and branded. And then I was like, well, why not make it interactive? Right? why not make this whole thing interactive? So, ~ this this whole Roku remote changes what's on the TV, right? Based on graphics that I created. ~ I could take these pins and I can plop them in anywhere and it ~ pokes it in, right?

Play from 1:09:09 1:09:09 so each one of these things has them, but it created these images just by me doing this stuff. Right. so it went and generated all these images. Look at the bot like the bottle moving. this pen already, I already loaded this, but this thing rolls his eyes, right? Like popcorn popping on this thing. Like I just started adding all these fun elements, but it created all these images.

Play from 1:09:34 1:09:34 There were a few exceptions where I was like, you know what? I actually want to have a few items that were on my desk. Like I had ~ like this actual mug is in one of them. ~ but it's just like so fascinating because I didn't Prompt for any of this stuff, right? if we got these, this is like a waste management client. I got these flies that show up and interact with the mouse, great. right? Like a steaming cup of coffee, right? ~ all this stuff. It feels alive, right? let me pour, let me pour some whiskey.

1:09:49Creating interactive websites with AI

Play from 1:10:02 1:10:02 Let me pour some whiskey in here. Let me do let me do some shit, you know? It's like, bro, like and I Only took a few items on my desk and I prompted for none of these images. Right. It's just like it's just an interesting new way of doing it that I hadn't been had to or wanted to, even maybe to some degree before. And it's just like this was a project where I wouldn't want to spend a lot of time in a tool doing this myself when I just have the vision and I want it executed.

Play from 1:10:35 1:10:35 So ~ for the ~ specifically for this one, you know, a lot of times what I'm building is system prompts, right? Like this is just a how to get the whole thing to work exactly how we want it to every single time. And I because I've have extensive context on system prompt building and a lot of other examples and a lot of other workflows and things like that, you know, the the context of it is is great. And then I'm saying, you know, okay, here, like what here's the here's the problem solution, here's what we need out of it.

Play from 1:11:05 1:11:05 Right. And then basically I'm going, go run to one of the MCPs, whichever one it might be, and take XYZ product and we're gonna test all the different angles of the system prompt until we get exactly what we want with the different models. So instead of me sitting there and all right, let me test this system prompt with nano GPT, ~ ideogram, this, that, whatever, then it's like, okay, it'll just go run the the structured test, and then I have the, you know, I have like the the quality control person being like, okay.

Play from 1:11:34 1:11:34 these two one, let's let's, continue going down on this path. We might need to adjust the the text prompt this way. Do we need a different input now? Is it gonna be, you know, from from there on out, do we need you know, is this gonna work with every product on the site or is it only gonna work with these types of products? So then it'll go pull from the website, you know, scrape those like, is a is a tube different than a cylinder, right? Like is it gonna like work the same way?

Play from 1:12:00 1:12:00 Is it gonna be structured? And like you can cover all the angles because I've given it all the instruction. ahead of time how this needs to work and it can just go test. And then when it comes back, it's like basically I figured it out. Cool. Let's go run it and see the whole thing work, And then I'll find the bugs. then I'm coming up with the solutions. there's a there's a really cool solution I found for consistent lighting in ~ you know in ~ with nano or ChatGPT or whatever.

1:12:22Automated creative testing and quality control

Play from 1:12:27 1:12:27 And it's essentially ~ a 3D technique. It's just a light sphere. Right? Like the light sphere where you put the it's like a the just ~ a sphere and whatever direction the light comes in from it's just using it as can basically standardize the lighting that way and it'll translates perfectly. So it's like the same exact thing every single time. And I'm like this is it's like if you had a static camera. You know, if you just had a camera on a stand in a studio and it was just blasting from the same angle for every product.

Play from 1:12:57 1:12:57 Like you know, it took a little bit to get there. to figure out that exact thing to make it so perfect ~ so standardized and basically not screw upable. That screw upable Scrubbable. I was looking for one. Yeah, I was looking for one. We're getting c we were getting kinda deep into the podcast. I was getting a little worried. We need we needed so we needed some makeup word, now that you can we can show our intelligence now that we're telling us telling everyone that the machines do everything for us.

Play from 1:13:26 1:13:26 ~ but like for anyone who thinks this isn't work though, like the amount of engineering that goes into this stuff to get it there, like the amount of mess up, the amount of like the amount of workflows I've built and system prompts I've built to get to this point where I can just say go do this like without that and starting from zero. yeah, it's a little exhausting. It's been ~ it's been a climb. there was there was something the other day really interesting with a a texture, a leather texture I was working on in 3D.

Play from 1:13:56 1:13:56 And it was like it it could not get it in Blender, it wasn't understanding the texture and the bump map and everything. So it just opened up a side chat and was like, ~ just ~ write me this banger research prompt for figuring out, bump maps and like leather grain in Blender and how to make it come to life. And not make it so 2D flat and static.

Play from 1:14:21 1:14:21 It just sent me this thing and sent the whole thing off. it built its own new set of skills, did its own new set of research, came back and fixed it in one shot. And I was like, you know, except like I don't know how to describe it after a certain point with you know Blender. Like I'm here's the nodes we gotta use, like this is what we're looking for from like the lighting direction, and like we need the the pebble grain to be this side. There's only so many ways I can describe it for it not to work.

Play from 1:14:46 1:14:46 Yeah. expertise. So if you don't have it in that particular area, then it kind of comes down to the I don't know if it's a skill. It feels like a tr it feels like a trait and a skill, maybe, ~ of knowing how to what am I trying to say? Like Trying to contextualize it in in in a way that you can't ~ verbalize or you don't have the vocabulary or the domain expertise for, So you're in you're you've got the ingenuity enough to to go and ~ find ways to loop around to figure it out, thing, Yeah.

Play from 1:15:23 1:15:23 if you have if anyone's working with video, right? I think we talked about this maybe one time. if anyone's doing a lot of AI video and doing stuff like that, ~ I've always had a problem with some of my standard prompting tools that I've had that build the prompts and run it. That, the the the camera d doesn't feel alive. It doesn't feel real, right? it feels like very ~ structured and ~ nothing's nothing's there's there's no ~ exciting element to it.

Play from 1:15:50 1:15:50 So I got built, I just built a skill around a camera operator that talks to now my prompting engine and it gets way better, more dynamic results because it just ~ instead of thinking like a prompt, it's thinking like a camera operator. Right. Or you can do this in any sort of way to customize it. Right. Like I have I have ones that are more for like vintage style. I have ones that are more for 3D style now of what I want like a if I want and I know something's gonna be sixty FPS and I want it to look like that, you know, that that really clean animated Pixar style, right?

Play from 1:16:27 1:16:27 Like there's a way to describe it in a better way. There's a way to build scenes in a better way. The the one I did I don't know if I even showed it last week was the you know, the the Fast and the Furious thing that I did, which was just nuts. maybe I'll actually pull that up because that was that was fun. I'll show the difference between it. was like the I had Blender, I had Claude and Blender go break down like the final drag race scene from the Fast and Furious and like build it out in a three D model.

Play from 1:17:04 1:17:04 Right. So it built it out the three D model, like it has like that's what's happening down here in Blender. It's building the entire race, like who's ahead, when they jump the train, when the car gets hit by, you know, when the car hits the truck and then flips over, how to do the real physics and stuff. So then like I'm baking that in. with ~ did I put the references in here? I don't know if I did or not. Some of the did I put the reference in yes here.

1:17:10Better AI video with virtual camera operators

Play from 1:17:30 1:17:30 So like, you know, what we're putting in as reference is the actual blender footage, right? Like this is what's recorded and you know Blender renders the video. Then we're putting in the you know the car contact sheets and then I put an environmental sheet in here. I don't know if I showed this or not if I did. ~ but like I put images from the real ~ location in not just from the movie set like I had to go look for location shots from the from the actual set where like I did research on where it was filmed, what street it was on, what part of LA it was in.

Play from 1:18:06 1:18:06 ~ you know then it also did the little cinematography card. So it got all that down. And that was fed as a reference, right? Now that can go into into here and that's the animation that we get. And now I'm like recreating this scene. Is it perfect? No. Is it still wild? Yes, I had people commenting on, why there are two double yellow lines. I went and looked at the satellite photos. There's double yellow lines on that street. I can't believe it got that right. Kind of thing. and then cool thing about that is then you can now that you have that animation, I can shoot it from two different angles, right?

Play from 1:18:35 1:18:35 Same scene, different angle, just rendered slightly different, right? It it the timing is exactly the same. They'll hit the tracks at the same time, and then they'll, you know, they'll go under the the little piping here at the same time and then both hit at the same time. Right. So it's It's crazy that this is this this is where we're at. ~ I don't even know where I started with that, but the idea was are stunt men in trouble, bro? Because I mean, what why would you put somebody in a super risky situation in the future when you don't have to?

1:18:39Recreating Fast & Furious scenes with AI and Blender

Play from 1:19:06 1:19:06 Also, right? Like when the car's flipping or whatever it may be. Or I mean, maybe there's not a guy in there, but like even driving the vehicle to flip it, right? Like I'm sure Tesla can do it now. Put some Tesla technology in there and have it just do the the stunt. I don't know. Shout out shout out to those stunt men. You gotta subscribe on this one. I don't know. But comment on my LinkedIn though. They were like, they're like, does this scene mean anything? ~ if there's no stakes to it? I is like, ~ maybe I'm just being skeptical, but there's no stakes.

Play from 1:19:33 1:19:33 there's no one driving the car. ~ even when I'm watching the movie, I know there's a stunt man in there. It still feels like there's stakes. I'm like, it's a good point. ~ noted. I like that. ~ but at the same time, you know, the fact that we can recreate this with real live data from stuff, and is it 100% accurate? No. Is it wild that We can create it. I can just say go watch this scene, create a blender scene from it and ~ reshoot it from different angles. Yes. Wild.

Play from 1:19:59 1:19:59 ~ but it's just weird, man. Like I I don't know. And the battle now is between creating content and doing stuff for to build a business. Like I don't know if you're having that. Yeah. sure. yeah. Like, and also when you, when you're too busy, you don't get a chance to, well, one, play with new things, but also like tap into a lot of these, a lot of these things we we we get to a point where I'm not sure we've actually finished what we wanted to finish.

Play from 1:20:29 1:20:29 And it's sort of in a good enough state, right? Whereas like ideally you want a great state. And so there's like, there's the new stuff. Like, so you obviously get inspired by new stuff. You've also got the stuff that you've already worked on that you could improve, right? ~ finish. And then you've got the stuff that like you you need to fix and like maintain and and make sure it's not stale anymore. And like I there are days where it's like, okay, I don't have any client calls.

Play from 1:20:58 1:20:58 That feels really exciting to me because it's like I've got all this other like I could spend a whole day just in my ecos. I could spend a whole week in my ecosystem just like improving things, getting things tighter, eliminating things that don't matter anymore, right? Testing new ideas. Like I don't know. It's just, it's, it's a whole rabbit hole that never ends. It never ends, bro. I mean, it it really doesn't. Like it just there are so many things I have saved that I want to get to and I just haven't.

Play from 1:21:28 1:21:28 Like I just have not done it. ~ I do want to I do want to Show something because I did release this. I I second guessed myself whether I was going to. but I if you've stayed toward the end, like we always do valuable stuff. Always stuff. So I gave away, I think there are actually 12 things in here. Yeah, there are 12 skills in here that I've uploaded and I've shared.

1:21:51The problem with unlimited creative possibilities

Play from 1:21:57 1:21:57 so I did like a LinkedIn post yesterday on this, and I didn't want to just give it to anybody. Like, if I was gonna release it, I don't know. I I I probably shouldn't even even released it. But I was like, okay, look, if you if you retweet this, I'll send you the link for this. Boom. Okay. But I'm sharing it on this podcast because it's out there. And I feel like if you've stayed to this, ~ go get it. Go get it. There are 12 skills in here that I built that you have access to that you can just go do.

Play from 1:22:25 1:22:25 So this is GitHub slash DC Brook slash Claude dash skills. we'll put that in the episode description in case you want it. Maybe you want some of the stuff, maybe you don't. I think there's probably some value here for just about everybody that exists. So, ~ the blind spot, looking for ideas you haven't thought to ask for, domain jack, prove steal, prove an idea from a different industry, puts it to work on your problem. We talked about the I adopt thing. So I run this when I see new ideas and I'm like, how can I apply this to what I've like, do I already have something like this?

Play from 1:22:57 1:22:57 If so, like is there stuff that I can build on top of that I can steal from that, like whatever it is, right? Plenty of stuff in here, but That's available. So I just wanna shout that out before we before we wrap up too. That's I that's very generous of you. I would also like to maybe steal a few of those. ~ you know, just to see what there was I think there was one. It wasn't not not a red team one, but maybe something similar to that. It was I don't know if it was a skeptic or something like that.

Play from 1:23:23 1:23:23 I'm like, you know. a red line, it's a red line one that you could apply to red team. But having run red team experiments now without it, I don't know that that one would be the most valuable of this group. Just because I feel like these models know red team really well. ~ they get it. there's there's a whole like you I don't think you need a skill for it, to be honest. ~ but there's some other really, really good ones in here. Like the mine one is another one.

1:23:36Drew shares 12 free Claude skills

Play from 1:23:49 1:23:49 So it's like it takes any transcript. And so I was listening to that. John Kirick, what how do you say his last name? Kiriyaku. but I'll hear that and I'll be like, okay, boom, mine, and it'll just Yeah. it puts things into my ~ idea bank for maybe future posts, but it also strips ideas out and does the same thing with as adopt some ways where it's like, okay, I could apply this to this that I have going on with a client, I could do this, right?

Play from 1:24:18 1:24:18 negotiation stuff from a CIA perspective, maybe I could apply that to a client. It's got it finds the correlations between You know what I mean? ~ so it's just ~ cool way to connect dots. But yeah, man. It's there. speaking of the content creation, I'm I'll I'll do the I'll do the Grokbot situation here because I wanted to I just wanted to show ~ what 30 minutes of of chaos in a brain can look like to try and get like I'm like, I don't know what to do with this thing.

Play from 1:24:46 1:24:46 Everyone's on it, I want to try it. I don't know, I figured just test it out. So ~ what I did was ~ what I did was create like, no, they're all Lord of the Rings characters, but this is all like around X. there you go. Right. So everything here is I'm like, well, let's just optimize X. Right. So we have I was like, I need you to build like all these different agents, right? 'Cause then I want to have like the different environments like Rivendell and Minas Tirith, which you have our research hub and our planning hub where they can all collaborate.

Play from 1:25:14 1:25:14 But the ~ you know, it was like if you're gonna break down like individual tasks for social media, I was like, I have, you know, AI, you know, I have this one that has is focused on like new things in AI tools, right? Which is this one's Great. It's like pulling all the things that are recent, plus also ~ trending things on hugging face that are coming out. ~ we have tet you know, formats, right? Trending formats. So again, that's pulling like what's happening, like video formats, things that.

Play from 1:25:43 1:25:43 Legolas, who's scouting everything, he's filling out, you know, just like going through the entire, you know. We have Gollum who's analyzing viral posts. We have Sar we have Sarmon. Yeah. change the character pictures for these bots too. You can do custom custom images for these. So you could actually have images of Gollum or Legolas on here as well. So not to not to take you down an absolutely worthless rabbit hole, but you could do that.

Play from 1:26:14 1:26:14 You could do that. But like this is how it siphoned you can get. You have your algo expert, right? Like reading through the X algorithm. You have, you know, analysts on my own posts. I have the strategist for my own posts. Now we have Pippen, who's also I'm I'm looking have him looking for trending memes, like you know, like the Susan Dell and ~ like some of the other stuff that was going on. The and then you know Elrond here is the content expert who's like working with the the algo expert, my past per my past, content, my strategist.

1:26:18Rory's Lord of the Rings AI agent army

Play from 1:26:44 1:26:44 So they're all working together. And then ~ Minas Tirith just stands them all up and they all report to it. Right. It's like I'm like, all right, let's just see what we can do here. I don't know how this is. If anyone knows how you can build one of those I want to build one of those little interactive environments where they're all just talking to each other. And it's like just going. And I I just want that to be like the thing. Like a I've seen that happen where people build Like a like like the Sims, like for your for your agents. I'm gonna build an office for my agents.

Play from 1:27:11 1:27:11 And this is just for X. I'm like, I don't I don't know what this is gonna bring. I'm spending fifty dollars this month to see what it's gonna do. If it helps in any way, shape, or form. I'm sure there's a million things I could do with it. Because the other thing is a lot of people that I've heard that are using it, they're like, I'm doing it for ops. I'm like, isn't everything we build now ~ on Claude for ops? I'm I'm already set, I already have that done. it's already runs every day. I don't have to I don't need another platform for it.

Play from 1:27:37 1:27:37 But it's cool. And I think the ~ there's probably some things that I don't necessarily understand. have you gotten into the ~ the agents, agents that'll just do stuff for you at home, I need you to rebuy this for me, or can you put me on auto subscription on Amazon for ~ distilled water? ~ I haven't done that yet. I still don't want to give that up. It feels weird. Yeah. I don't think I'm ready for that yet.

Play from 1:28:03 1:28:03 I ~ yeah, that that that's just like a whole other rabbit hole, ~ And I don't know. Maybe maybe it's just because ops, like to your point, ops are business related stuff and creative stuff come before that in my in my priority list. And I think it's the same for you too. So it's like I haven't no, I haven't. Fun fact for anyone here, again, if you've stayed this long, Nuggets, if you've been looking for a flight anywhere and you know that dynamic pricing is a thing right now, right?

Play from 1:28:37 1:28:37 ~ you have you ever played the game where you have to go, you know, where you you're trying to book a flight and then you go back like the next day and it's like a hundred and fifty dollars more expensive, and you're like, come on, it's like the same seats are still selected. It's just like tracking you cookie wise and charging you more 'cause you're looking at it for a second time. If you go through the agents, if you go through Muse, if you go through ~ I I've only tried this with Muse, but Muse has direct link to the API. So you get API pricing and not the dynamic pricing track to your IP address.

Play from 1:29:05 1:29:05 So like they can go look and search for things like that and get direct pricing, not dynamic pricing. so that was like I saved probably six hundred bucks on, you know, a three ticket flight. Cause I had Muse do that. ~ prices right now are insane. I just paid eleven hundred for a trip to LA. For just you. Just me. Just you.

1:29:24Should AI agents handle your personal life?

Play from 1:29:31 1:29:31 ~ we used to fly we when my wife and I were first dating, we used to fly to Rome round trip for five hundred dollars, six hundred dollars. It was multiple times we did that. No shot. Not maybe maybe two and a half X that price. And it's from JFK, so there's no stopover. There's like a million flights that go there. Right? It's like a like I think we did that like four or five times in a row. never spending more than five, six hundred dollars.

Play from 1:29:58 1:29:58 Now it's if you get there under a thousand, that is a that is a real treat. But kidding, Mira. the it's you know, makes you want to travel less 'cause you're like, what the f even sometimes in the US it's like, why does it cost six hundred dollars for me to fly to to Florida? Like these flights used to be hundred fifty bucks, two hundred bucks. And now it's yeah. how you cheap it used to be to fly to Vegas? They just want to get you there.

Play from 1:30:23 1:30:23 I feel like those those prices have changed drastically too. I was just looking at the the other day because I was looking at maybe flying out to the the ~ Raiders Rams game ~ in Vegas. It's like eight hundred bucks. Just go to Vegas. It's like, what? To Vegas from Atlanta? Like, I don't know, man. It it's wild times out here with the with the with the flight stuff. then you gotta buy everything on the pl ~ the eight hundred bucks to get your body on the plane.

Play from 1:30:51 1:30:51 Then it's like, do you wanna have your reserved seat? Or do you wanna reserve your own seat? Do you wanna put bags or Yeah, with the they separated like selecting your seat is now a tier, right? It's like when did that happen? You know, like that's that's crazy. That is so crazy. Anything they can do. time. Yeah. It's it's really annoying, not to mention the gas prices. That's why it's all so expensive. So not gonna talk about anything of that sort.

Play from 1:31:17 1:31:17 ~ because we'll get we'll get dinged. But I think, hey, can we can we figure out, ~ certain ways to make things cheaper? AI. I saw someone great great idea for your little, personal agents to do, or saw someone doing this on GrockBot was like, go scour the internet for like my my personal data and then have me unsubscribe from all of these ~ call services and data opt ins where it's like you get your your phone ringing thirty five times a day, that's a good one.

Play from 1:31:46 1:31:46 ~ Yeah. Yep. ~ that kind of stuff. I would love I'm love to have that that thing just on autopilot all the time. there's certain w there's certain little things, quality of life stuff like that. Yeah. ~ I would never want to experience a a another unknown number calling me 17 times a day, which I don't know about you. I made the mistake early on. ~ I had my phone number on one presentation that was on ~ that was on YouTube that went like somewhat viral 'cause it was in my like a like about me section.

Play from 1:32:20 1:32:20 ~ I was on stage and they recorded it and then they put it on YouTube. And then it's like right after that, dude, I get fifty it it started right after I went live and then it was like my my phone for ever. Wow. I like won't stop now. That that that piece of information's been sold a million times. Really annoying. still getting the I'm still getting the one every week, the the text that we're in your neighborhood fixing roofs. You can we come to your roof?

Play from 1:32:47 1:32:47 ~ that that literally happens once or twice every week. I mean, it's it's insane. It's insane, bro. We're in we're in your neighborhood. I'm sure you are. I'm sure you're in my neighborhood. Also, I fixed my roof like literally a couple months ago. And I didn't even need to fix it. So leave me the leave me alone. what they're doing. Like I had I had Muse. I'm like, tell me the most ridiculous the most ridiculous affiliate scheme you could come up with.

Play from 1:33:16 1:33:16 And it was just like, you know, this probably not that ridiculous, but it was, you know, we're gonna we're gonna monitor weather data. And what we're gonna do is we're gonna ~ spin up, you know, a bunch of different content streams to for like disaster relief and not like disaster. It's like basically like roof and like damage removal or like trash removal. And all we're gonna do is spam the area once the once the storm hits and then sell pay per call ~ to like the roofing the actual roofing services.

Play from 1:33:43 1:33:43 So it's like it's it's not dirty, it's just smart. It's like they're what I'm saying, but it's ~ yeah, it's I mean smart people figure this out. Yeah. and the people who need the people who need their roofs fixed get the roofing companies. It's just like you're playing connector faster than anyone else can play connector. That's the the affiliate mindset there. And paper paper call paper lead is sometimes can be really good.

Play from 1:34:09 1:34:09 ~ two hundred dollars a call. If you're doing that at scale, then it's not bad. Mm-hmm. a couple quick things before we wrap too. ~ probably call this out. Midjourney talking about an MCP limited community. There's an application for it. Gotta fill it out. I did. I did. I think there was something where it's like, give us a com give us your most compelling reason why or how you would use this.

Play from 1:34:37 1:34:37 And I was like, yeah, I'm not gonna think too hard about this. Yeah. came up with some some bullshit, but I was just like, Yeah, be nice. Be nice. ~ but that that's crazy because I mean this feels like a better route than the API. You know, like this feels like, okay, cool. Well now we've we're actually maybe gonna give you some way to use AI within the tool. ~ which is smart.

Play from 1:35:02 1:35:02 And then I think there was one other what what was the other thing? there was like the big thing too, I think. it's probably warrants a bigger discussion and maybe we save some of it. But if you guys didn't see this piece too, was super interesting, which is Apple and Meta long rivals teaming up here ~ to make the Muse AI model local on Mac computers in the future.

Play from 1:35:28 1:35:28 Which, I like it. you know, when you read it, you're like, okay, that's interesting, but it's it's it is a big deal. this is a really big deal because right now with the subscription models, right, we have to pay Chat GP or pay OpenAI or Anthropic for these models. And if you've got something that's already built in and already there and has the same capabilities and the stickiness doesn't exist, the network effect of it isn't a factor, that lunch could get eaten.

Play from 1:36:00 1:36:00 That'll be interesting to see. I I I don't know. ~ that to me is super, super interesting. And I wonder how that's going to that could even affect, ~ Grok. type in dollar sign type in dollar sign meta on X I just wanna see if as well let's see if the stock went up with that. Ticker, yeah. ~ it went down. ~ it's this I don't understand the stock market.

1:36:03Midjourney explores MCP integration

Play from 1:36:25 1:36:25 It just that's today though. Let's see. Over the week. it is down over the week. ~ that to me is such a big piece of news. It's like like that just every Apple user. That's what I mean. Yeah. the subjective nature of the stock market in general. It's like cause you ~ live off big news and hype, to see these bumps. And you would if you if you're invested here, you would want to see a big, a big jump from that.

Play from 1:36:52 1:36:52 But that's an industry winner for me because when you already have all of Instagram, Facebook, WhatsApp, data, training your models, then you get all of that, then you get all of Apple stuff. You have the biggest pool on earth, and then there's no way. ~ how do you compete with that? ~ and they have the compute to do it. they have all of it. ~ Meta's really set up. They're set up very well right now. in the last six to twelve months, haven't they?

1:36:55Apple, Meta, and the future of local AI

Play from 1:37:20 1:37:20 They were basically they were basically a non factor and super quiet there for the first couple of years of this whole thing. They they c they felt like the last kid that, like a musical chair, you're just trying to to wiggle your way in there, you're just looking for a window of opportunity, ~ to jump in with the rest of the crowd and find your spot. it they felt like that for a while. Or they or just really patient, ~ patient about things.

Play from 1:37:47 1:37:47 they were paying really jump on stuff. ~ you know, they they kind of didn't they didn't fail at a lot of stuff I think Llama kind of failed, but like they didn't really push it too hard. It wasn't like a big risk like exposure that they had. And now it's like they they hit Muse. Muse is good. They just open sourced or you know, s released their open weight coding agent that you can now download it that and use it.

Play from 1:38:14 1:38:14 if you'd like. It's apparently it's beats ~ apparently it beats anthropic and ~ Astra in certain things. you know I don't how those benchmarks work. I never understand them because it they don't make it feels very arbitrary. But that that that right there from from Zuck. It says Muse Spark open. You know, weights releasing today. Runs on Mac Studio. Thanks to his I I asked Muse about that. It said it's needs about one point five terabytes to run.

Play from 1:38:43 1:38:43 So Yeah, go get yourself an eighteen thousand dollar Mac Studio to run it on. you know, if you want it. But it's Gotta get that write off going. Might might need that for next year. I was I was ~ what Apple did recently, which I think is actually very smart. I other companies might have it. I don't purchase hardware that much because, ~ it's just something you don't buy that much. They have the leasing option now for computers and hardware like that.

Play from 1:39:08 1:39:08 I looked that one up full, fully loaded. It's three years and it's like three hundred dollars a month. But ~ for eighteen thousand dollars, that's not terrible. And you can also just it's good because of the obsolescence, ~ three years is gonna be worthless. You know what mean? It's like i they won't be any the way that we're developing, the way that, you know, difference between the one and the four chip like was like, you know, two or three years but like significant.

Play from 1:39:34 1:39:34 Now five here, I'm sure we'll get six and whatever sooner rather than later. But ~ Yeah, hardware, hey. I was I was looking for more power. I'm like, how do I get more power? 'Cause I I'm just just this is like Back in the nineties when I w when you're when you're using lime wire and just ~ destroying your computer. like you you had to be a real patient.

Play from 1:40:00 1:40:00 Fucker back then. because everything was just so slow. You but like it was to the point, Rory. I don't know if you were like this, but like I've been using computers from from like I don't know, like a young age, and you got so used to how things would hang and lag that you almost knew how that particular lag would solve itself. So like you knew if you clicked and it sort of did this thing where maybe it It's good for a second and then it lags and then you're just like, okay.

Play from 1:40:33 1:40:33 I know this is gonna hang for, I don't know, probably like 10, 20 seconds. Then it's gonna come back and I can kind of do this and I can do that. Then you like I used to start to get like superstitious with like, if I could click my mouse over here and do this, like it would alleviate the lag faster. You like you just played these fucking mind games because of just the era that we grew up in. You just like had all these like, I don't know, it was that just me. I just had all these superstitions too that would work themselves in where I'm like, if I open If I open this window, if I close this, I like it's gonna resolve this.

Play from 1:41:01 1:41:01 I click here, it's gonna alleviate the stress of this, right? ~ I definitely had that. I definitely had if I restarted twice, it would make my computer faster kind of thing. Yeah, you know, like that kind of stuff. Like if I put it to sleep twice, it'll come back faster. You know, the lime wire stuff. That was still remember I I still remember that moment as being the coolest thing that had ever happened to me as a kid. I was like, wait.

Play from 1:41:27 1:41:27 Like I can just I don't have to go to a store to listen to music to like buy a C D. I was like, this is sick. I still liked having a C D or a cassette tape. To put in my analog, ~ speaker system. That was sick at the time, because I had ~ spent my working money on that kind of stuff when you're living at home. Look what else are you gonna spend it on when you're 12, 13? and then your trunk? Yeah, no, not my trunk.

1:41:38Remembering Napster, LimeWire, and early computing

Play from 1:41:52 1:41:52 In my room, yes. In my room I had these these old, I think they were, I don't remember if they were Bose or Harman Cardin, these wooden wooden box speakers that were God, it was such clean my parents probably hated me, have a subwoofer and that going on all day, ~ I just in my room listening to music all day. But the just that feeling of like this feels illegal.

Play from 1:42:17 1:42:17 It was illegal, and it was just like everyone was doing it. It was like I can't believe this is can't believe this is a thing. You know, Napster, of course, like it was on that right away. It was like what I don't even know how we found out about this stuff. We didn't have the internet. Like, how did we find out about Napster and LimeWire? Do we have aim? but we but but yeah, yeah, like yeah, I guess what what what would have been the method of travel for information? I guess it would have been like you would have had to log in computers, and I guess like Yahoo was still the search engine at the time, right?

Play from 1:42:48 1:42:48 Where it was like Yahoo AOL. Stuff like that. And I guess that would be on the front page of news. And so you would land on these sites as almost like your news aggregator at the time, right? ~ It was like no one even knew what Napster was until it was, ~ the the adult world did not know what Napster was. They couldn't even talking about when when was that nap Napster like what? ~ was that mid nineties or was that ~ late nineties? Napster release date. Let's see.

Play from 1:43:14 1:43:14 This was June one, nineteen ninety nine. So I was I was ten. And I remember I remember being on there. And then it was gone and then Napster. is still a company right now? Yes, I know that 'cause I forgot what they did recently. AI focus according according to Google is AI agents, business and hardware solutions. How about this one?

Play from 1:43:40 1:43:40 Shut down July 11th, 2001. Napster, the original, the music sharing platforms. That didn't even last two years, right? Or it did last two years. But that was quick, and then it was Limewire. Right? I feel like it was lime. Date. May third, two thousand. So that was right after and then yeah. So they had Napster as the sacrificial lamb. They were the first Yeah. and they got smoked.

Play from 1:44:07 1:44:07 And then once the model is out, everyone's like, we can just do this file sharing thing. free yeah, peer to peer Foster and client. I don't like that was like when then then torrents. I don't know if you remember Torrents, downloading Torrent files to get movies and stuff. So you didn't have to buy ~ the the bootlegged film, VCR recorded version of of the movie some dude's trunk yeah. There was a site for a while where man that you had all the newer movies on there too. It was great.

Play from 1:44:33 1:44:33 Yeah. It was ~ but like you download a movie, a movie would take two days to download. you you already know you're tying up your computer for for that amount of time and then when you had dial up it was like, Okay, so no one can call me either. So ~ mom, dad, sorry. Like can't use the phone right now. dad. What are you doing up there? Don't ask any questions. You don't know. You don't get it. You never will.

Play from 1:44:59 1:44:59 Hey, it's it seems simpler. Now we got flying cars. We finally finally got so shout out. Who is shouting out? Well, we can shout out Luna, who's also shouting out in the back over here. I need you to take it easy. ~ who do we shout out? I don't know. Like the Okay, Music Pirates. Yeah. Then about somebody else in the moment too? Can't remember.

Play from 1:45:25 1:45:25 Well's you know, what's what's first if you've been here for an hour and forty eight minutes, like you gotta do one thing. Lots of rapid fire. Rapid fire. But if you've been here, you gotta do one thing. it does you don't have to like it, you don't have to subscribe, but you gotta do one thing. If you're gonna like or subscribe, leave a comment, but you gotta tell somebody. So if you're not familiar with this, you gotta tell somebody maybe a little bit unex unexpected.

Play from 1:45:54 1:45:54 You know, maybe ~ who can we come up with today? This one I'm I'm actually like at a loss right now. I don't need I think if we've gone through so many. is fried after that 'cause I feel like we just speed ran everything that we wanted to talk about and just try to get it in under sub too. But How about how about this one for nostalgia? Your local radio shack or like sharper image employee? How about those? There we go. Circuit City.

Play from 1:46:20 1:46:20 Wow, circuit city, good one. It's a good throwback. I've I haven't thought about those. a Radio Shack had a real vibe to it. They did. So does Sharper Image. And then what was it? What was Beeper World? Do was there a beeper world or something like that? Remember Beepers? Do you remember you don't remember that whole era of beepers? know about no beeper world. I don't know about no Sharp, whatever the hell you said. Sharp? Image? No.

1:46:43Final thoughts and shout-outs

Play from 1:46:45 1:46:45 Is that every mall in the nineties and two thousand? It was like with ~ i all it had was weird tech. But it was it'd have ~ it was like the first place you ever saw that had a massage chair, ~ had like really cool stuff. Yeah. Like it you you'd go to sharper image, then you to Hot Topic after that. And then you go to you know, then you'd walk go into hot topic. I wasn't I wasn't that that wasn't me.

Play from 1:47:13 1:47:13 Yeah. Yeah. Spencer's. Spencer's gift, then like, you know, swing around to models or sports authority. ~ you know, not buy anything, just look at everything and just loiter because that's what we did. Yeah. I'm just remembering my my giant ball from the nineties, early two thousands. any of those companies. How about that? Yeah. And like Panda Express, who just dealt with us buying, you know, stuff when we were kids and just being like, Can I like six orders of orange chicken?

Play from 1:47:43 1:47:43 And no rice and just give me the chicken. God dang we were ten. You know, the peop we were allowed to go to the mall when we were ten, right? Like this is that that definitely does not happen now. Definitely do allow ten year olds to go to the mall by themselves. But era. this different era. But hey, this was ~ I think we covered a lot today. We've rapid fire. did cover a lot. If you if you guys are curious about anything diving further into any of what we talked about, because we did speed run it, let us know.

Play from 1:48:11 1:48:11 Also, I think we are scheduled, at least right in this moment, for guests the next two weeks. two totally different guests, two totally different conversations that I think are gonna be awesome. And then ~ man, we're already out of out of October after that. So we're Really looking at the end of the year. We've got other guests that we're lining up right now through the end of the year. ~ I I wonder if we're are we gonna be at a hundred episodes before end of year?

Play from 1:48:36 1:48:36 I don't think so, are we? It's gonna be close. Cause we're on 87 right now. Where I think we're about 13 weeks out from the end of the year. thirteen weeks or no. We've Friday, January fifteenth, if we hit every single week. So probably looking at the new year on that one. But ~ yeah, lots to take in here. ~ we'll share some of the stuff that we talked about in the episode description. Let us know if you guys have another also let us know if you have a guest that you want to see on here.

Play from 1:49:04 1:49:04 I think that's also another good one. Drop it in the bucket and ~ yeah, everybody have a great weekend. We'll see you guys next week. See ya.