AI Filmmaking Tools, Robot Liability, and GLP-1 Ripple Effects

Andrew Mayne: Hello, and welcome to the Weird Things Podcast. I'm Andrew Mayne, joined by Brian Brushwood. Hello, hello. And Mr. Justin Robert Young. Hello, friends. I've got some science stories and things to talk about, but I do want to do an AI and a tech story, but I actually want to talk about, there's been a lot of stuff about really fast model advancements and new AI's that are now recursively self-training to be scientists and whatnot, and there's a whole open AI solving an open Erdos conjecture. The kind of thing that still sits to me right now is Google Omni. Okay. Brian, have you seen this yet? Oh, man. I feel like Google Omni is their image, their video model, isn't it? Right. And what they did now is you can do video with, meaning that we could go shoot some movie or something like us sitting around your table us talking about stuff and then say, okay, make it a Wild West bar. Give us hats. Do this. Do that. And I think that, I thought that, you know, what OpenAI announced when they had, remember Sora, guys? When they had cameos, when you could put, we could put, yeah, when we could put ourselves in, I'm like, this is a great thing because this is a great step towards a bridge between people and AI. Now Google Omni, where you can actually capture real performances and then alter them, do all kinds of crazy VFX stuff, whatever. It doesn't have to be like aliens and robots. It could literally be make the time period or just these things. I was just so excited about this because as a person who's obviously very much a technologist, but also a person involved in the arts, I like steerability and control. And that's one of the biggest criticisms for a lot of stuff like video models and stuff is that they don't have enough steerability for artists. And now I think this is a really great direction this is heading. Yeah. Well, so number one, I'm a little bit amazed. Just now I was searching for Google Omni and boy, oh boy, the SEO, the ad page keywords, I couldn't find it in traditional search. I had to go to YouTube and find a video to explain what Google Omni is. How do they position it without saying the Sora word when they launched it? Well, I mean, they've been building their own models, their VOs and whatnot. And yeah, I think that the Cameo thing was certainly the first place that I saw it. I think when anybody saw it was in Sora. But again, I think that what came to the table here was the fact that you're using original footage and able to like regenerate it without not changing. It's not like changing the dynamics, the performer or the art, you know, alter the video like you'd Photoshop, you know, an image or ask for AI alterations on an image. So, yeah, I mean, if you're thinking about it on the minor level, how many times, Brian, have you shot a thing, but there was some little element that was left out or it would have been great if you would have had a blank for a blank, right? Like you are somebody who actually here's here's a real a real example. Brian, how much do you care about shooting things at golden hour? Quite a bit. Quite a bit. A lot. Right. Yeah. Like because it matters because it you know, everything feels good. It looks good. The idea of wetting down the the surfaces behind you so it reflects and gives a little bit of that kind of a shimmer. Those are things that as this technology gets better, but but the concept of shooting something and then having that be added is the promise here is like beyond some of the more wacky like turn me into a chrome man or like add a bunch of hearts around my head. The real world promise is or even like, you know, putting you in a foreign world, like little on the edge things that can just that that, you know, you you will take time out of your day to make sure that you are shooting in this 45 minute window because, you know, it makes all the difference. Now that kind of stuff can be at a at a at a at a at a at a buttons press. That's so interesting because there's there are two very good reasons to do it. One, of course, Golden Hour is is good for all the Golden Hour reasons, you know, faces fill in nicely and and and and and. But the second layer is that it is inherently challenging and there's a tight window and it's a bit like doing a one or we were talking about that in the Great Night Bones podcast where it's like, well, if it's not a one or then does that subvert the stunt? And that'll be interesting in a world where a where where you can press a button and all of a sudden what was a hallmark of very difficult to do, kind of like five years ago, having lots of targeted emojis in your email, everything perfectly formatted was an indicator of a lot of care and now it is not. That's interesting. That's even the concept of a one or right. The concept of a one or took a gigantic hit in the era of digital filmmaking. Well, yeah. Along with the crazy Unreal Engine sets that they had. Well, I mean, yeah. Beyond that, I mean, I'm just talking about, you know, DV and digital editing and stuff like that. Like the idea of a one or in a film world was a was a real threading the stone needle or else you were totally screwed, especially when it costs that much to shoot. You you had you know, it was it was some real daredevil stuff in a in a digital world. It's like, you know, it takes everybody's time, but you can do it a million times if you want to, because the cost of the film doesn't matter as much. And then digital editing came around where you could be like, OK, well, I can actually add tricks here that were not available to me before. And you can be very, very fine in how you're editing stuff. So it's like it's always changing. And, you know, to the Omni thing, you know what Golden Hour does. Like you mentioned, it fills in people's faces more. What if everything you did and you shot just had that? How many times how much time is spent adding the right tint and maybe even the right like slight film grain to something in a post work, like to make sure that it that it's good. Now you can just kind of have that have that have that put on there. I I'm going to go a step further is, you know, when you guys do your world's greatest con. And there's there's a world where you spend a week capturing stuff to do, you know, reenactments, you know, and a check down at the table and say, great, this is going to be the lab. And, you know, you know, this is the place and just say, hey, 1970s lab here. So because you can just throw in photos and say, put this person in this environment. Yeah. Speaker 01: Yeah. Andrew Mayne: Yeah. I mean, there's I think what they really cooked with it. I have not played with it. I'll at some point I'm in buried with other stuff, but at some point I'll sign up again for the MKUltra setting for Google and play around with it. I don't know what at what tier. I think you need to join Sea Org to be able to to play with it. But I do I do want to do it because obviously video models are something really, really fascinating. And the one thing that I think this has gotten as good as anything since Sora is physics. Like, Sora was exceptional at physics. Sora was good enough at physics that it could do slapstick comedy. That is even for all of the the videos that are coming out now because politics is saturated with it that have their own reasoning as to why they're important or funny with like Spencer Pratt and stuff like that. But there's no slapstick in the way that Sora could where Sora could have, you know, how many of the viral videos from Sora were like a cat playing a fiddle too fast and somebody slamming open a door and saying, like, knock it off. It's like that's the kind of thing that you only when the physics match. And it wasn't just it wasn't just the physics. It was also the comedy timing, the punchy. Yeah. Come on. See? Yeah. Hey, buddy. Like never a dead moment. I mean, you could I don't know. It had timing chops in a way that I haven't seen replicated yet. The one. Yeah, that's that's the one thing, though, is that with with Omni is that like it does feel like it understands physics a lot more than Vio did, which obviously Vio is text to video. So that is creating a whole world, but it's where, you know, Google has been deep minded within Google has this idea that like the LLM revolution that we are in the middle of right now is not the path that there is a different way to apply to say I and it's an AI that understands the entire world and is not just kind of next varying more complex versions of a next token prediction. Well, I'll tell you what it makes me wonder, because what I've seen from a totally different angle is there's a few blogs and newsletters that talk about the YouTube tactics stuff. And then one one of them was about a buried setting about like check whether or not there's any synthetic manipulations in your video. And but then as a parenthetical, they mention, of course, if you use any of the Google tools, this setting doesn't matter because everything is synth ID and it's baked into the DNA that everybody knows you did it. And and so now what I'm thinking of, Justin, is that interesting question is, let's say I do have a 90 percent their video and I could press the 10 percent sweetener button to make everything perfectly golden hour. Now there's an interesting artistic calculus to do, because if I don't press it, then whatever I got, I get artistic points for having gotten it. But then if I if I do hit it, then the synth ID says, well, there's some amount of synthetic in there. And that's just the art question. And then I think about the in the hip hop remix culture, it makes me wonder what kind of things could never have gotten close to being done without this. And it makes me want to artistically lean into all of that to where like basically there are magic tricks where if you know that there's a marked deck of cards, they are less interesting than if you know for a fact that they are not marked deck of cards. Right. We're we're continuing the everybody becomes a magician game where where there's also kinds of tricks where it does not matter at all to their impressiveness whether or not those cards are marked or not. And then that that elevates the meta, the artistic meta. That wasn't really a question. I, I think that what I was getting at, what I like about this is that a lot of people think generative AI is I just wrote a prompt and I got the thing. But if you're out there shooting, you decide, Brian says, I'm going to make a Western, I'm actually going to go use some real locations and stuff. But hey, I can't afford sets. I can't afford lighting in the way that I'd like. I can't afford these things. But I have ideas and I know what I want to do. And you go shoot with real actors and you bring in some people to creatively collaborate. You know, you have a friend that says, hey, I'm a costume designer. We don't have the budget to make the costumes, but I can design them. And a small team, like much like in video game development, working in a game engine treats it that way. You know, you know, game engines didn't destroy the video game industry. The people used to have to hand code everything. Then somebody said, let's use a reusable engine to make this easier to do it. And it wasn't nobody's really paying attention to go like, ah, the artisanal things of having to write an assembly to get this thing to work. And I do think we lost some things, by the way, with game engines. I think the gains were greater than with the losses. But you look at some of the super innovative things that were being done in gameplay and whatnot. That being said, I think I think that this is a transition point to the idea that you can go out there as a filmmaker, bring in some creative people, and you can point to every choice and say, yes, a human made this choice, a human made that choice. But no, we didn't have the budget to pay for horses. So we're sitting on saw horses. No, we couldn't go shoot this at, you know, like a Hollywood studio in the middle of Arizona. We had to shoot this in Michigan in winter. But we did that. And that's that's what I think that I think what will happen is you'll get some stuff where they can be amazing and people will do a lot of effort to show you the behind the scenes to show you just like using blue screen, just like using other stuff to go. Yeah, we use these as tools. But look at these people making choices. You know what? It makes me think of like, let's reduce every decision that happens in, let's say, a short film or a movie to either a human decided it or performed it or did it or we rolled some dice. And really, the game becomes you only get embarrassed and get in trouble if the number one thing everybody says is their favorite thing about your piece of work is the thing that you knew you pressed a button or rolled dice on. And so it's like, great, well, then now the name of the game is just make sure that's not the best, most interesting thing about whatever it is you're creating. Yeah. Well, I think that, yeah. Oh, it's interesting the various ways that technology polishes over some things, but it's fun to watch ways in which we're increasingly able to see the human. For example, in what will be my pick later, there are some de-aged Nicolas Cage moments in Spider Noir where, but I have no doubt in my mind that the currently aged Nicolas Cage performed them because when he swings his gun around, he doesn't swing it around like a late 20s doughboy. He does it like a mid 60s actor. And when he's running, he's running at an appropriate age. I mean, I think that's fascinating. That was the same thing with the Irishman. Oh, yeah? Yeah. It was Robert De Niro was de-aged and it's like, that is definitely still old Robert De Niro. Yeah, it was weird because at one point he calls him kid and I'm like, what? I'm like, how am I supposed to pretend he is? Or I think the greatest sin is Sigourney Weaver in the Avatar movies playing the two. A teenager? With the mature woman voice. Speaker 00: Yeah. Andrew Mayne: Okay. But yeah, I am just excited in general about the ability for artists to have just this chance to just be more steerable. You know, we have what, Toy Story 5 coming out and we see images of Tom Hanks and Tim Allen and why do we care? Because one, we like those guys, we like their performances and we know that actors provide their own steerability into something. And I think that that's, you know, Brian, you brought up before, a lot of people hate the idea that somebody can press a button and have it look as good as something else they work hard on. But I think that when you say, hey, look at this control panel, there are 9,000 buttons and the people who are making great choices are the ones who know how to press them. I saw this back with Dali, the first one I did, the Dali launch at OpenAI. I'm Mr. Prompt Engineer. Like, I'm the guy that thinks that I know how to talk to AI. We put this in the hands of artists who understood artistic terms and photography and painting, and they made much better images than I did. And that was a wake-up call to me of just how much that terminology gives you an insight into steerability. And you can spend, you know, somebody could spend forever and forever trying to make their own Spider-Man noir, and not understand what a Dutch angle is, noir angles, and the only times you get it is when the AI happens upon it, versus somebody who understands those things and deliberately does them. Yeah. I think a lot about, like, the ability to just know who to summon from this magical crowd, where it's like, you know, you've got an audience. I think I made this parallel a few weeks ago. I'm doing a magic trick on stage. Somebody has a heart attack. If I turn to the crowd and say, what do I do? The crowd's got one answer. UNKNOWN: Woo! Andrew Mayne: But if I say, is there a cardiologist in the house, then I get a very different set of answers. And likewise, let's say you want to make your own Spider-Noir. It's like, I need a cinematographer. Explain to me what uniquely made for this type of things during the movie. What are the different types of equipment? Like, great, what might it look like with exactly this lens and this type of film stock or whatever? Great. Of these three choices, now you've summoned somebody from the crowd to lay out a menu from you and behave as an expert. And now we're back to, in Brian's game of find the human, you found the human. Yeah. Vincent Velkap is Dutch. He says all his angles are Dutch angles. Speaker 01: Nice. Andrew Mayne: He'll split them. Hey, gentlemen, I've got an update. You know, there are moments you never forget. Where were you during this moment or that moment? And they're going to ask, and you're going to have a memory. Do I need my lawyer for this, Brian? Not yet. I don't think so. Uh, let me, let me see if I can find, oh, no. Uh, oh, dang it. Okay, we'll, we'll do this one. That's fine. Uh, where were you the first time you saw an unprovoked robot attack? And the answer is going to be, this, this is the one. Get ready. I mean, you're talking about like the Kubota manufacturing line, Mr. Air would kill the human in the 1960s. Or just remotely controlled robot, by the way, that freaked out. Uh, I'm talking about this awesome moment right here. So, basically, uh, uh, there's a robot accident. Uh, robots do it a little, uh, looks like a unitry robot is doing a demo of some moves. Uh, but a kid wanders too far into the line. Um, but it, it's funny because in this case, very clearly there, uh, you know, the stakes were low enough that we'll walk past this moment, but imagining, uh, somebody gets a fatality there. Like, who is at fault? Is it the event holder? The guy, Brian, can I tell you? The guy, the guy, the guy, the guy, the guy, those are remote controlled robots. Sure. They're not autonomous. So the guy with the holding the remote control in his hand, who's off camera is responsible. Uh, uh, uh, okay, but, uh, uh, or, or the person who didn't put up stanchions to keep the people away. I, I just, I just, what the reason I'd say is like, I see people will watch these, it's not to accuse Chinese robotics industry of a mass propaganda campaign, but every time you see, unless it's on stage in a box and it's doing a prerecorded thing, when you see it walking or running or navigating, sometimes the camera will shift and you'll see the guys holding the remote controls. And I'd say that's, that's the thing is there's a lot of perception among people. They think that they're watching autonomous robots. They're watching RV cars that walk, RC cars. Speaker 01: I like RV cars. Maybe, yeah, maybe they're registered voters too. Andrew Mayne: Uh, Brian, yeah, you were, uh, uh, uh, well, you think that kid got up? Do we know? Oh, uh, uh, uh, yeah, yeah, we can go wherever you are. I'm sorry. I didn't mean to destroy the illusion that it was an actually autonomous robot. It wasn't, it was a remote controlled thing. I just, I just, you know, I'm like, I'm sorry. I will, I will, I will. You didn't really like, like let the bid unfold though, man. You know, I guess I get, I know. I'm just saying like, we, we, Darth Vader's his father and that's like, they're like on hot. No, I mean, I guess the thing is, it's just, is a guy that is, is a, uh, very concerned about how we look at AI and robotics and stuff and things like this is I watched the media run away with this stuff or go look at all this stuff. Like, well, it's like, uh, no, that's not actually, it's like the, to me, it was like the psychic power stuff and ESP and stuff. It's like, well, no, that's not actually what that is. So, so the, the, the, the narrative of unrestrained, out of control machine intelligence and specifically as it applies to violence toward humans. Uh, there, there are, there's great examples of unitary robots just, again, my channel's flooded, it's like freaking out at restaurants. No, no, no, no, beyond, beyond the robots, but I like your, your larger point. I'd like to kind of pivot to the larger point. The, the idea of, uh, the narrative becoming out of control of like these intelligences are breaking out of the box and, and, uh, hurting specifically as a results to hurting. Well, I mean, part of it is the narrative that like China is like so far ahead in robotics is this, that's the thing that China has been spending a considerable amount. Even, even when you see a freak out, like you see a thing where robot hit a kid, like, yeah, but look, the robots are in parades and interacting and where are ours? It's like, and I think China is very, very good, very far along in mechanical stuff, but the autonomous stuff, they're not. And so I'm like, well, that's not actually the, it's the, it's the operator who got too close to the kid and did that. You know, it'd be like, if I saw a drone, we watched a drone, you know, human controlling a drone and it hit a kid like, oh, the drone went crazy. Like, no, the, the operator literally had his hand on the controls deciding where it went. Screwed up. Yeah. Is that like, I mean, I, I've seen a few people talk about how China is super far ahead. Like Ashley Vance, that's one of his favorite things to talk about is how far ahead China is. Mechanically, yes. A hundred percent mechanically China makes actuators, makes this. But like I said, every time you see this really cool demo of a thing like, you know, like walking or running, you, you'll very rarely when they, cause again, they're, they're company, but when you, and those things are impressive, they're amazing. Unitries builds great stuff. But when you pan the camera to the left or the right, you'll see these guys holding RC control things and controlling them. And that's the thing we're not, that's the part of the trick we're not supposed to see. Because when you see that, you go, oh, I, I get it. They're not, it's not this way. Cause like, oh, imagine a battalion of these just running into battle. And it's like, oh yeah, we're, you know, it's like, you know, Ukrainians using drones with their control. Yeah. Uh, so the, the question that this was simply the prompt to, to get to, uh, everything that I've heard so far, it sounds like a lot of present tense reality-based stuff. None of which was, was what I was interested in. Uh, what I'm interested in is, um, if this is a mild depiction of a real thing that, that I'm fairly certain will happen when you have semi-autonomous complex systems, uh, how do you measure liability? And the closest comp that comes to mind is, uh, uh, dog attacks or animals attack. Right. So, you know. I, I would say that autonomous vehicles on the road already are the thing we do for that. And, and that's been a lot of legislation time was spent figuring out like, if your Tesla is in full, full self-driving mode and it's somebody who's accountable, that's why right now you have to click, I'm accountable. You hop into a Waymo, Waymo's accountable. And that's the thing that the point at which whoever's engaging with it, there's an agreement between the owner of it and the person who's using it. So in a Tesla, if my Tesla decides to just run into, you know, somebody's house and go through the living room, Tesla says, well, Andrew, we, you signed the thing that said you're responsible. I get into a Waymo and does it. Waymo takes accountability. Uh, that, that is also a present tense story. That is interesting. That is not, uh, where, where I'm, I'm interested in. What's your hypothetical then, Brian? Cause you're saying, you're saying if a thing does a thing and I'm giving you examples of autonomous vehicles and things right now that are doing it and how that was decided, where's the, where's the bifurcation? I'll, I'll tell you what, Andrew, I'm going to give you the end line of the movie and then maybe we could go back and walk it together. But I was making a case that this moment in miniature is exactly what Michael Crichton spent his whole career predicting. This is a miniature Jurassic Park. And then it's fascinating to me that you have a creation, humans showing up to gawk at the creation, you have the plan and you have an unexpected complex system spinning out in a way where there is a consequence. And, uh, I'm fascinated by all the many, many interesting stories that, that, that we, that, that could be in the future. But here, Brian, if this was all I knew about robotics or the only things I knew about this interaction, I would agree. But I'd say that going back to the history of like the first industrial robots that went out of, you know, first industrial accident, 1960 robot killing somebody, following the whole autonomous driving debates back and forth and accountability there, I would say you'd have a different perspective on it. That's all I'm saying is you'd look at like, yeah, well, this is actually RC. I can go back 10 years earlier and show you other examples, case law, things like this to say that where this was determined. But I like, like, yeah, well, this is, this was a remote control thing. So I, I'll bring in, you know, I can wax about this from a imaginary or, you know, fiction sort of point of view, or I can say, well, this is where the data's been and what's been going on. Well, my question was in the future, what, how will this go and what precedents do we have today? So, but meanwhile, it sounds like, uh, yeah, you, you, you certainly let somebody have it. I don't know who they are or what their, their, their situation was, but I would love to hear those stories as well, because I have, uh, uh, I don't know the case law that comes before. Speaker 00: Okay. Andrew Mayne: I mean, following the Waymo, follow the Waymo stuff. You had the first interactions of when Waymo drivers were supposed to be accountable. The rollouts that's going out through there, you look at through what's been going on the lawsuits against Tesla versus full self-driving when they've been trying to sign accountability. I'd say like, there's just, it's been, it's been, I'd say that we've been watching this rollout with autonomous driving. And it does seem to be still actively litigated. Like people are still trying to figure out exactly. Yeah, but I'd say that the question, the original hypothesis, what question was, who is responsible? And I would say that, that right now we see that sort of, it comes into two different places is you can have an operator, but it ends the relationship between the operator and the person who's engaged with it. In the case of, like I said, with a Tesla, when you sign in there and you click, you click, I take a troll, you're accountable because you're saying you're operating and Tesla says you have to handle accountability. You get into a Waymo, fully autonomous vehicle, whatever. Waymos have had accidents, hit people, whatever. However, Waymo accepts the liability for that. And I'm not saying it's completely settled or whatever, but that is why we have these things on the road right now. This is why we have robotic vehicles driving out there at high speeds right now is because we've reached this sort of idea of who gets assigned responsibility for that. My issue with this was like, well, this was literally an RC robot, which was sort of a different example. I know I got very pedantic on it, but I'm in a world now where that matters. Yeah. Yeah. I think, Brian, are you, I think the more interesting question here is who is like in the scenario that you are painting, who is morally responsible as much as legally responsible? Yeah. Well, exactly. Like, like, like in this snapshot of a moment, uh, uh, this is the closest to a real life Michael Crichton story, uh, that I've seen so far. And, uh, uh, like nobody signed an agreement. They were there to see a cool creation. And then, and then the robot attack, like it's a literal, like, uh, like I'm just curious to see where it goes. Uh, I just, I guess for me, it's like, I'm, I'm, for me, it's like, I'd be like, how would I parse this versus a drone show in China where the drone started falling down? And that's another good story we could talk about. I'm interested in that one too. I would say there's a ton of these. So I'm just saying like, that's the thing for me is that like, there's, there's this, but I would say that the, the, the liability thing, whatever it's ongoing, but it is going to become a point because we're going to enter the Isaac Asimov world where these things are going to do more complex behaviors. And you're going to say, no, I didn't tell it to do that. It decided to do that. Speaker 01: Yeah. I think it's God's fault. Andrew Mayne: Got him. Got him. Got him. No, I don't know. I mean, like if you look at, all right, so let's say this is, it's more along the lines of like, that there is a robot theoretically that had about the capability of like a Waymo today. Right. And so now you were looking at, you know, and in many ways it would be fractionally smaller, the damage that a unitary robot could do than like a Waymo, right? Because like Waymos have had examples or autonomous vehicles have had examples where they've just hit people and they've died. Like, uh, but if this was okay, uh, doing a Taekwondo demonstration, a unitary robot boots a kid in the chest, uh, and imagine that it is indeed totally autonomous. It's like, all right, well, who is it? The creator? Is it the person who put on the event and programmed like, Hey, do the Taekwondo demonstration? Is it the, uh, the, the, yeah, people who decided where the people were going to go, where the crowd was going to go and put them in harm's way? Uh, I mean, I assume like any. Is it the mayor who insisted that it's 4th of July weekend and that the beaches needed to stay open? Yeah. Is it, uh, uh, I mean, I think that's, uh. Or like, I mean, like, is it somebody who didn't patch things correctly? They're like the nose sidekicks patch didn't go in. And so now there was a dangerous, like X, Y, Z, like, I don't know. Uh, uh, I think that it is, it is more diffuse, although I don't know if it's as diffuse as we might think. Like we certainly, there are more interesting questions to be had about it, but ultimately the person who says, go robot, go is going to be the one who's responsible or the person who puts on the event no matter what happens. It's like, it's not the idea of like a robot kicking a kid is the first time that there would be danger in a crowd. It's like people are lighting fireworks or, or, you know, hell, look at the. Fireworks is a good comp, especially any kind of like Rube Goldberg-y kind of machine. I mean, there, there have been. Yeah. I remember there's WrestleMania where fireworks lit something on fire and like trestle fell on, uh, people in the crowd. Uh, there was that Travis Scott concert in Houston several years ago where. It was not anything crazy on stage. Uh, there was just gate jumpers that had crowded the floor so much that people, uh, got crushed and suffocated and died. Like there's, there's a ton of different ways that people unfortunately can, uh, get hurt with these things. I got a link for you. That's the last time of the rope this happened with the Chinese robot in a parade. Uh, and you can see it's another kid, Amy Chinese robot hits kids, but it's saying this is the one where you see the guy run up to it. I'll send a link to you, Brian. You see the guy run up holding the remote control as the robot goes out of control and you see the guy go and stop it, which not as photogenic as, by the way, the other one that you showed me, the rope, the clown, it's been all of my feed and literally as I open up Twitter in my feed, um, because it's like, we're getting, you know, something like each week, but here's a, I just sent you the link. You can see this is another example of a robot hitting a kid. It looks, it looks like he's had one too many where he's just like having a good time. Oh, wow. And then you see, watch this. You see, oh, did you get the link I had? Cause then you see that you'll see the angle there. You'll see the guy come running up. Speaker 01: Yeah. Andrew Mayne: There's the remote control. Yeah. Hey buddy. Hey buddy. Let's get you, let's get you back to your trailer. So again, my, my point is saying, I was trying to say before, and I know we, we, we, we are, our enthusiasm for robots got very enthusiastic is, um, it like here I go. Like, yeah, I'd say the guy at the remote control. This is my favorite. This is, this is the dancing robot in the restaurant. Watch this. It's dancing. It's dancing. It's doing a little dance. It's pointing to its head, pointing. It's pointing. It's adorable. Everybody look at the robot do his little robot dance. What's the problem. Why does anybody have a problem with this? Why could you, why would you object to this? It just starts smashing the table and then he's just like, I'm great guys. Speaker 01: I'm having the time of my life. Andrew Mayne: They've got the, in their hand is the remote control because they put it into dance mode or whatever. And yeah, they need to stop dancing. They need commander Riker to turn off data. Yeah. Oh, another one. Oh my God. Brian's now found the magic feed. This is this prominent, this lady brought up this, the most janky robot walking on stage and the robot waves and it's human sized. Speaker 00: I know. Andrew Mayne: Well, I mean, then we get back into my fascination with how people manage narratives in that moment. Whether or not they, they act respectful for what appears to be a human body simulacrum passed out and trapped or what have you. I'll give you like a little, like kind of behind the curtain here. Is it like, and again, I'm biased because it's just now investing in the robotic space and stuff like this. And sometimes I want to tear apart other stuff that I think is garbage and, you know, maybe you need to be more kind. There are a number of robotics firms that have been around, have been together, and they have people with strong mechanical engineering background, strong mechanical engineering. They understand a lot of different things that are needed to do bipedal walking, different algorithms about like this, but they're not really strong in the neural network side or the, you know, that part of it. And we're seeing some of these demonstrations and stuff. You go, that's really cool. I think walks, whatever. It's kind of all it does because the next step is really tricky. And I think we're going to start seeing really, really complex systems come into play. And I do think that, and to your point, Brian, the stuff that I said, hey, autonomous driving answered a lot of different things. Because that's fine when it says my instruction is take me from point A to point B and do this. And any deviations, either response for me is the driver supervising it or is the company providing the car. But when it comes into, and I think within, we put a robot in your house, you assign some disclaimer. But when it's like, robot, go to the grocery store and go do this, there will be a lot of scrutiny on the instructions. There will be a lot more attention towards how those things are decided. And that is sort of, I think, the world that Isaac Asimov was foreseeing was the idea that, hey, listen, these things could go there. Well, and it's funny, like, iRobot is almost the story of an actuary who's hunting down the source, the cause of an industrial accident, right? Like, you have a system with a decision tree that went afoul at some point. And all of that. Brian means the book, guys, not the movie. Yes, exactly. Speaker 01: Sorry. I did not murder him. Andrew Mayne: Yeah. But it's amazing to see that get closer. And it always happens because however good the guardrails are, that is how far event planners are going to push them. You know, the mayor is going to want to keep the beaches open. And then, and that means these, each of these incidents, and yes, they're misrepresented by the media as being, you know, autonomous or whatever. But the reason they are is, of course, because we can all see the universe a few weeks, years from now where these problems do get more serious. And like, like at some point, I don't know, it's, it's interesting. Did I tell you, I saw a thing about iRobot, the movie with, uh, do you know who played the robot? Yeah. He plays all the robots. If you need a lovable robot or somebody to die. There was an interview. Yeah. It was an interview with Alan Tudyk. Like, and he was saying that, uh, when the movie was testing, his character tested like out of this world. Everybody that saw the movie was like, oh, the robot, the robot's amazing. My favorite character was the robot, right? Uh, to the point where the studio wanted to kind of market the movie almost as kind of like a two-hander that it was like Will Smith and Alan Tudyk in as the robot, right? Like, like that this was like a futuristic, playing into the futuristic side of it of like, here's, you know, one of the greatest performances via CGI that like, it's so good. You'll feel him through the camera. And that, um, Tudyk suggests that it was Will Smith who was like, uh, no, like I'm not, I'm not splitting billing with the robot. Like this is a Will Smith movie. It'll sell as a Will Smith movie. And that's, that's that, but that Will Smith put his foot down and Tudyk was a little salty about it because that was definitely going to be the biggest studio kind of, uh, uh, backing for anything that he had done up till then. But, and I guess if, if you got beef with Will Smith, it's been a good couple of years to air those grievances. You've had cover. Yeah. No, now's, now's the time, you know, uh, his star has, uh, gotten complicated. He's up there next to Michael Jackson and, and, and stars with asterisks. And it's, it's frustrating because it's like you had some great Isaac Asimov stories, you know, the, uh, the Elijah Bailey and the Daniel Oliva, you know, robot team up like the, the, one of the, the first human robot team up stories that. Oh, the caves of steel. Yeah. Yes. Yeah. As a matter of fact, caves of steel, there's a plot beat where, uh, there's a whodunit kind of classic mystery vibe to it. And at one point it's like, it had to have been you, you're the robot. He's like, no, I'm the one person. It can't be. It was like, but you're not a robot at all. You're a human. Why do you say that? Because when we went to the bathroom, you had a wee wee and why would a, why would a robot have that? He's like my, I told you I'm an exact duplicate of the guy who made me. And I mean, exact. I'm like, Oh, okay. Anyway, moving on. I guess they did some. Oh, they did a BBC did a version of the caves of steel back in 1964. Speaker 01: Huh? Andrew Mayne: Oh, that's, and then they almost a naked son. So very cool. Uh, gentlemen, hold on. Brace for it. Okay. Ready. Totally not stalling. So I can pull up the email and actually get the facts right and just not make stuff up. I mean, it's the internet. I know. I know. Um, so there's been, um, this is, this is sort of a really interesting talk. You know, we're talking about dark matter. Everybody talks about dark matter, but have you ever seen dark matter? Anybody have any dark matter? I mean, I keep squinting. It keeps not showing up amongst all the conversations that I've had about dark matter in the last year. I have never seen it. Well, you know, makes one suspicious. So there is a primordial black hole candidate. The idea is a primordial black holes would have been started like very, very early on in the formation of the universe, right? Very, very early on, you know, kind of has different patterns for that. And there is one called Phoebe, which was visible towards the large midgenellic cloud. And they watched as this thing moved across an event. And it took like an hour, which was sort of this, this transient event, which, you know, does this thing where it warps the light around it. And the timing of it and the size of it made it very, very suspicious because they felt like for the size of it was like a lunar mass that would be more explainable or understandable if this was made of dark matter than if it had been like a traditional collapsed matter black hole. And this might be one of our first direct observations of a black matter object. So dark matter, as I understood, deep in the periphery, is largely like a mathematical. Like we've got all of our observed data of the universe and we got some stuff that doesn't match. And then you hypothesize. But if there's some matter that we can't see, some kind of dark matter, then the equations work out. And so it's like until we know something more, let's keep running with that. But so all we know is literally all we know right now is what you just said. Speaker 01: Yeah. Yeah. Andrew Mayne: That is that we have we look at the galaxy and we go not enough visible matter to hold together like this. There needs to be. So some other matter that we can't call it. Yeah, that's that's it. And that's the thing has been trying to look for. Well, and keep in mind, it's so interesting because and I adore this story. Einstein died believing that his fudge factor, he, you know, had a bunch of math that wouldn't work out. And he was like, I don't know, maybe there's a cosmological constant. And then he was like, no, that's dumb. That's like a fudge factor. And he died calling it his greatest mistake. And then in the time since he died, it turned out to be the most cited piece of work he's ever done. More cited than even the laws of special and general relativity. So it's remarkable when you open your mind to, I don't know, the numbers, the math works. Speaker 01: Yeah. Andrew Mayne: I think it's going to be interesting to see. We now have AI that we can start throwing a lot of data. And there's going to be ways in which we can start, you know, looking at a lot more information and a lot more stuff and processing that and looking for signal. Other question is GLP drugs. Been hearing about all the other benefits. You know what? But I saw, well, first of all, whatever story about GLP, I want to hear it. But there may be another remarkable therapy on the horizon. Do you hear about this cholesterol gene therapy? No. Yeah, I only, I look forward to reading the full article, but early tests seem to indicate promising gene therapy that would make, instead of ongoing treatment for cholesterol, it's a gene therapy that would be, as the pull quote in the New York Times said, a one and done solution for cholesterol. So it's like alter the something something gene and then you don't have as much LDL. If I remember correctly, there is a subsection of humanity that has a gene that essentially just keeps LDL cholesterol low. Like there's nothing that those people can do to raise their LDL cholesterol. They are, it is very, very hard for them to do it. And part of it is just that they create an enzyme at a different level than other people. And this therapy, theoretically, just kicks off that enzyme production. And it just, once it's going, it's just going. Andrew, you almost certainly know much more about this than both me and Justin put together. But when we hear gene therapy, like it's not like you take a pill and your DNA is all different throughout your whole body. What, like what, what do you, what do you know about gene therapy? Because I don't, all I know is it's not the way I imagine it. Well, Brian, it's an oral, not necessarily a pill, which could be, I have no idea. Um, I mean, you, you, the way we do that, the, one of the ways we do gene therapy is often is you use a virus. You use a virus because remember a virus goes in and takes over the transcription stuff. And there's different categories of gene therapy, but much like a virus goes inside of your takes over. You remember, we forget how freaky viruses are. They literally hijack into your cell, take over your DNA, the replication machinery there and start saying this. And sometimes it gets screwed up and bits of the viral DNA get trapped in our own DNA, but once you use the same machinery. So, so gene therapy can mean either going into the cells and saying, hey, just start producing this or using something that actually tries to rewrite the DNA of the stuff. So, um, there's forms of it that will then extend to the, some things you just take, whatever, it's not going to get passed on to your kids. Some things can actually rewrite your DNA. So there's a few different kinds of gene therapy. Yeah. Sometimes it's just literally take this pill and it's going to be producing, you know, helpful genes or whatever in there. And it just sort of leaves. Yeah. Uh, I'll tell you what, I, I asked my doctor about it. Oh. The cholesterol drug. She had not heard of it. Um, but, uh, I was like, yeah, you know, cause I'm in a situation where they want to lower my cholesterol, but it's, it's a weird thing. I have like a, uh, I have a gene that like will just always keep my LDL high. My whole family has it. Um, but everything else that they test me for, I'm healthy. Like, so all the remedies that they would normally give me, like radically altering my diet or working out more. It's like, I'm doing it. Like my blood pressure is normal. It's just my cholesterol. Uh, so I was like, look, to be totally honest, uh, I'm just trying to foul off pitches until, uh, AI cures death. So, you know, whatever we need to do just to, just to keep the wolves at bay for another 15 years. Like I think we're going to hit escape velocity and, uh, we're all going to be here for another several thousand years. This ship of thesis is ready to, ready to, ready to roll, sir. Yeah, I'm like, I'm like, look, I just need to not get hit by a bus in the next three years and I feel like I'm going to be fine. Uh, yeah. Locks, blocks, GLP ones though. Yeah. They've been finding a lot of other benefits from it, from, from joints, maybe anti-aging, other conditions from this. And it's, I would say that for, on one hand, we have to acknowledge for everything you do, everything Michael Greit would say, for every time you alter a system, there is going to be perturbations to it too. But we don't need to be animous about it and feel like, oh, I got this magic here. It must have come at great cost somewhere else. Sometimes the cost is much more minimal. Sometimes you're, you're, you're basically just trying to avoid entropy. And, and this has been, there's really interesting data coming out and showing this. We're already seeing effects. Like people say like lingerie sales is up, which, you know, is that attributable to GLPs and whatnot? It's a nice thing to think about. And I would say too, a thing that we have to adapt to is the, like less people gamble. I don't think less people gamble, but like Las Vegas is, you know, not as cool. I've, I've got a, I've got a friend who went on GLP ones and dropped maybe a hundred, 200 pounds or something crazy. And also stopped his compulsive degenerate gambling. Like, uh, we're, we're six figures minimum. Like all of the casinos won't stop calling him because he was one of their high roller whales and he just lost interest. And in fact, while we were out in Vegas, he's like, I think maybe they haven't figured out that I'm cured. And so he would get, he would get comps to things, but it's, it's real. It's very real. Yeah. And we've seen this already kind of declined and sort of like that idea of that, that, you know, gambling in general. Uh, well, I don't know. I mean, I don't know if we count for what's going on in sports online, whatever, but, but certainly certain behaviors like that. And, you know, you hear people in the alcohol industry lamenting that younger people don't drink wine. A lot of people are like, ah, they're doing animals. And I was like, well, maybe, but not like all the time. I think people just, just like smoking that being drunk all the time is less accessible, you know, less likely, you know, and that's changes. And these things do have economic impacts, both positive and negative. And sometimes I'll, I'll sit with people who are involved in the alcohol industry. Like, yeah, what do we do? And I, and the first problem solver, like, well, and I'm like, no, this is good. This is good. I'm sorry. Change industries. Get into something else. You know, this is, you know, this is, this is the path, but I think it's going to be transformed. I think we're going to be seeing more of these things and we need to closely monitor them and measure the effects. But, uh, it is, it is, it is crazy too, how somebody pointed this out that like a number of, you know, body positive entertainers who were very body positive have now suddenly slimmed down considerably. And it was like, oh, if there was a switch to make it as easy for some people as it is for others, then that is okay. Yeah. Yeah. I mean, I think there's probably a larger cultural question about exactly what those movements like are and were, but, uh, it is without a doubt, uh, healthier for you to be at a slimmer weight. I think that scientifically there's not much argument there. The GLP one side of it is fascinating because that is stuff that we've seen since the very beginning of it is that compulsive habits fade. And, and, and if you look at the compulsive habits that tend to define people eating, drinking alcohol and gambling are three of the biggest, probably smoking right afterward. Now, and all of them, all of them are damaged by GLP ones. Uh, smoking and alcohol both have baked in pro social aspects to it. So in the before times where there wasn't vehicles to find like-minded people who aren't geographically located next to you, or to have rituals to do it, you know, video games, et cetera, uh, then smoking and drinking were a way for introverts to open up and, and reach out to extroverts. Everyone had to extroverts. Eating too. Yeah, I guess. No, that's right. Yeah. Uh, and, and I think that the lament, I, I think all things being equal, the restaurant business isn't so affectionate to alcohol for alcohol's sake. I think that they like a high margin thing besides the food, because the food doesn't keep the business afloat. So they're, they're hungry to say, we don't really care what the high margin thing is. Uh, but, but can we find a high margin other thing? It's just the thing is that high margin thing, the more you consume the high margin thing, the more it sells itself. That was a, that was a, uh, uh, a lucky benefit for the, for them for a long time. Yeah. I, I mean, also, I don't economic driver, you know, I mean, I think like, like restaurants, not exactly famously run by the most, uh, uh, penny pinching people on the planet. They tend to be, you know, the kind of artists of the entrepreneur class, but like, uh, you know, there's, if you had a busy bar, you could make some mistakes on the food side. You know, you could, you could have price spikes on certain, uh, ingredients and, and weather it, you know, because there, there was margin to be eaten into on the alcohol side. It'll be interesting if this is something that continues to go on or there are just broader trends that bring down alcohol consumption, how feasible are kind of sit down restaurants in, in the same way. Speaker 01: Well, I think, I think that we're, uh, starting, well, I don't know. Andrew Mayne: Um, I'm, I'm unprepared for this Ted talk, uh, but, but I know that everybody's racing to figure out bespoke, whether it's, you know, uh, weed infused cocktails or, uh, the non-alcoholic variants is, uh, burgeoning developing industry. Uh, you occasionally see hookahs and vapes, uh, as novel things. Well, maybe, but maybe, I mean, Brian, I, I, I, I, you and I are both problem solvers and we look at the thing, but maybe we just want to get rid of the thing. You know, maybe the idea is that we, you know, maybe it's just fine dining, you know, maybe, maybe the idea that we, we do are, I don't know. Like, yeah, I, I kind of feel like, I'm like, maybe I, it's like when I, when I want to lose weight, I realize don't get diet candies, don't get like the sugar-free versions of stuff. Cause all that teaches me to do is just crave candy. And, you know, is there, you know, I don't know. I, I just think like, what if, like, I don't drink. Like, so, you know, I, I, I think of like, what is that like for me when I go to a restaurant in my experience, you know, and the dress runs just adjust pricing, you know, just increase, make the, make the dining experience more. I think it'll just be more expensive. Yeah. I think it'll definitely drive prices up because you're going to have to make more margin on the meal, uh, than, than you would otherwise, or at least, you know, it'll be interesting. I don't know what it is. I mean, if you're an entrepreneur now, if you're a, uh, somebody who would, is into food, you know, there's just as many, uh, uh, people that like go into the food truck situation or ghost kitchens or things like that, where, you know, that's, that's the idea. Like that's a little bit more manageable. You don't have to pay. But then again, that doesn't have the downstream effects of hiring people, hiring servers and dishwashers and cooks and, and a bartender and stuff like that. Like, and, and especially in a world where you want to onboard more and more people into the kind of workforce, like those are great entry-level jobs, uh, that go away. And I'll say something too, is that I'm a Taco Bell guy. Like I'm like, like, I love my Taco Bell. Yeah. But, uh, you know, as is a birthday tradition, my wife and I will go to French Laundry and sometimes we'll take friends there. And as much as I'm, you know, uh, the fast food guy, I, I like it. I appreciate it. I really enjoy it. And I would say that if you thought of all the people who you would be able to take to French Laundry and have them go, oh, I really like this. I would have put my name on the list. But then all of a sudden, when I see that amount of attention, the amount of detail that goes into it, and you do look as popular as a show like The Bear, you know, and then on other things. I, I do think there's a world where the, the restaurant industry does fine because people stop and like guys like me step back and go, oh, when they really care about a thing or really care about the, the, the food or the dish that goes in there, I will notice. Yeah. I mean, French Laundry exists at the, uh, uh, delta of a lot of different things that allow French Laundry to exist. UNKNOWN: Yeah. Andrew Mayne: Yeah. Like there's money, there's prestige, there's the Napa of it all. Like there's, there's, there's a lot. Um, I don't know though. I mean, every, every, every, every town has the best restaurant or the better restaurants though. Every town has the places that are better. The question is where is, does that better extend beyond where you would think of just like, oh, is that worth it? Like that's ultimately what people, especially when you get into higher margin food stuff is like, like, oh really? $50 just for whatever the, the, the dish is. I mean, yeah, I don't know. I mean, cause I think that I, and I, cause we all have friends that, that will deliberate back and forth for 30 minutes. If you should get the $8 nachos, then order five more rounds of a beer they didn't really care for, but because it was some, they said it was a special or whatever. Like we've watched people do the, the, that, um, I do think that you might see these things strata out a little bit more. I think what, what might kind of be the victim are, uh, lower priced restaurants that could kind of count on a certain level of alcohol traffic. Um, so maybe chains like Applebee's or Chili's or something like that, where, you know, you, you have a fixed price for your food and, but it comes in pretty cheap cause they want to get families in the door. Or, but also the center of all those things are one thing, a bar, cause they want that value of a higher margin business to come in. Maybe those kind of, uh, uh, are, are the mushy middle. And then you see the Buffalo Wild Wings style, like, no, this is a place to drink and watch sports and eat wings. You limit that menu, maybe even a little bit more than it is right now. So you have cost assurancy and then on the other side, maybe you do see more like, oh, this is an experience. This is a, a, a, a, a, a reason to go out, um, that, that we are going to charge you a little bit more for, but you're going to know where the money went. And, uh, maybe you pay the waitstaff a little bit more. Maybe it's a little bit more concierge, but it's not, um, budget. You know, maybe, maybe that's where we go. So I, yeah, I don't, I don't know. Like, yeah, I, I think about this too, cause I do think we talked about before, like in Brian, your, your description, your friend, by the way, is fascinating. Just the, the other behaviors that just, I've heard this anecdotally, but not noticed anybody that does that. Uh, uh, uh, what's interesting is, uh, uh, when, when I started losing my weight, uh, uh, lost 35, 40 pounds. When I started on that, Bonnie casually brought up, Hey, don't do those drugs. I don't want them to kill your ambition. And, uh, and I understood intuitively what she meant. Like, like who knows whether it would or wouldn't. But the threat of that is like, uh, especially with us, the three of us being in independent creation fields, you kind of want whatever that, that hunger, that drive to go out and kill something is. I, I totally relate. Like, I, I, I've had friends that've used it and it's been a wonderful experience and transformed for them. I look at it and I'm also at that point where I can't, you know, for you and Miles to be like 30 pounds, like that's, it's hard for you and I to sort of justify it in a sense that we know that some lifestyle changes. There's, you know, we, we, we, I, or I can, I can only speak for myself. I knew for a fact that I knew what it took to get back to TV level fit. And even now I see myself phoning it in on that last 20%. Uh, and so, um, yeah, I knew the recipe and I had a track record of being able to do it. So it was just a matter of like, right, I guess, hello, five 30 in the morning, my new best friend. Yeah. Where it's like, yeah, I have, I have friends that they're, they're, their genes are different and it's a different thing for them. And I get it. And it's also, they're dealing with a bigger thing where, you know, for me, when I look at my fluctuation, I'm like, yeah, um, don't eat dessert every night. You know, like, don't do this. Like get out of your chair. Um, and other people, it's more complicated than that. But I would say that, yeah, that, that for me was a thing too. It was just like, I don't, that's why I never did drugs too. Like I never, I never did. Cause I was like, man, I'm pretty high output creative and I'm not sitting there thinking I need to be more creative. And I'm very lucky that I know how to flip the switch and zone in and do something. So I'm terrified of messing with that. Of shaking the engine. Like whatever it is, it's working. Cause I've seen people too, being in LA, like people like, oh no, it's great. And I'm like, you haven't done anything in three years. You feel great about not doing anything. You feel really good about the fact that you haven't done anything, but you haven't done anything. The, uh, the, the, the recursive loops that placate us or mollify us to, to do like whatever they are, it's, it's dangerous. It might be the only villain you ever have to face in life is just get out of that, whatever thought process that begins and ends with not doing a thing. If you can just get to the world of doing it and to get finally back to AI, cause we don't talk about it enough on this show. But that's kind of what I know for me working with Codex recently has kind of unlocked. And, and you, so you heard a lot of people with Claude Code, um, or Claude Cowork, uh, a little while ago from, from my money Codex is kind of the, the, the top of the pops, but. That's the producer of the open app podcast. Yep. Chill alert. Uh, but I, I don't think that it's a, a, a, a super controversial opinion these days. Um, no, I share it with you. I'm just kidding. Yeah. Yeah. Uh, uh, but it's like, it just, the barrier for just annoying things like Brian and I, like world's greatest con world's greatest con had a website problem because an old service we were using was SEO outranked to the service we were using and it didn't have a great website. So it's like, also it was sending out the inaccurate signal that we were a one season podcast that ended in 2020 because we moved hosts after the first season. Right. Um, so it's like, all right, well, we need a new website. That's fine. Um, but also it's just a lot of like little things, little things that I don't know, little things that I could find out. But it's just, I just know it's going to be hours of my day and codex just being the grease to kind of like get things going. So now I am like, I'm never without, I'll put it this way. Anytime that I face a problem, I can ask codex, how do I fix my problem? And it explains to me how to fix it. That's what went from a bad situation to like what Brian's showing right now is a website that was vibe coded from scratch and looks better than, uh, in my, in my opinion, looks better than, uh, what we would do on, um, on Squarespace or, uh, any other kind of like make it, uh, piece by piece. That looks amazing guys. That looks amazing. My favorite part is it exists as opposed to the hypothetical one that we'd still be with. And here's, here's where it started. It started with just a Reddit thread that said, you know, images to actually does this certain kind of art style. Well, and then codex being really good at producing multiple images. So Brian, we were talking before we started about like codex and images. That's one of my favorite thing about codex and images is that you can just say, Hey, make 20 using image gen and it'll just make 20 things using image gen. And so I was like, all right, uh, I didn't even tell it what to do. I just said, go to the RSS of world's greatest con group, these things by seasons and stories, uh, and make this style art about it. And then it did. And I was like, Oh wow, these are pretty good. So then the next thing was, Hey, what would a website look like if it was kind of wrapped around displaying this kind of art? Speaker 01: And then it spat out that website, worldsgreatestconpodcast.com, uh, spat out that website. Andrew Mayne: And then the other thing is like, all right, so what else is annoying about this process? Notes, you send it to friends, friends have notes. They want to change things like I'm working with Brian. Brian, obviously very invested in, in the world's greatest con, uh, uh, website. You want to know what codex makes really easy notes? Cause I just, in the verbiage that the person tells me to do a thing, I just say, Hey, more of this, less of that. And it's like, it's spit something out. I show it to Brian. I'm like, Hey, does that work? And he's like, eh, more of this, more of that, less of that. Boom, boom, boom. Uh, uh, it just made the whole thing, uh, uh, really, really like fun. I guess that's the, that's the key. The key is just the fun part is that it just kept breaking. It kept getting me back to the moments where my human intuition actually mattered. Uh, and we could get feedback, the ability to speed iterate and, you know, to grow a project rather than, you know, have scheduled at bats to maybe make your point this time and get it to be heard. Yeah, I, I got, I, I, we could do this in after things cause I think Brian and I collectively probably saved over like $10,000 like using codex to go through billing and stuff. Uh, I got a bill from liquid web, which was the, the latest post to inherit weird things and it was $270 for the month. Jesus Christ. Now we have a podcast kind of popular, but it's not $270 a month hosting feeds. And, uh, first of all, I said to myself, how did this happen? And well, it's, you know, we started off with on a wired tree back in 2007 and then you get these acquisitions that go back and forth. And the next thing you know, they're like, Hey, you, you're, your WordPress site, you know, we're using WordPress. Hey, you should buy this service to help with this because you'll get errors. And the next thing you know, you're spending a hundred dollars a month on stupid WordPress services because WordPress is terrible. Yeah. And I said $270 like that. When it was like 80 bucks, I was tolerable. I was okay with it. When I got to 270, I'm like, this is stupid. Hold codex. Go look through the entire work that work. You know, the weird thing site, get download every single podcast we've ever done. And then it found some stuff that was also like on archive, whatever. And then I said, uh, go through the whole archive, the whole site, everything, download every bit of data on their server. And it did that. And I said, build a weird things.com website, create an archive for eye tricks with the same thing. And built a very simple weird things podcast site. Every single episode now is there. Now with an RSS feed served through Cloudflare, which I love for hosting this stuff. And, uh, my $270 a month bill is now, let me check the math, 80 cents. Speaker 01: That's less. Andrew Mayne: Yes. For those not into math. Mm-hmm. That's, that's, and that's the world we're living in. And I would say that it took me some time to do it and to be thorough to make sure we got all the episodes. And it's also like, if somebody said, yeah, but you screwed this up, whatever. Cool. I'll tell Codex to fix it. Fix it. You know, I switched, I switched the RSS feed, everything. I got it completely off of WordPress, had it build its own RSS handler, went in there, did all of that. And, you know, people are listening to this episode. It worked. If it's not, you're not listening, then you'll never know. I mean, that's, I think when you look at the economic impact of a lot of this kind of stuff, what I think is going to come to a close is a lot of gravy trains, like a lot of, a lot of like real gushers that were pumping a lot of money. I would imagine if I was in a cartel that had somebody on the hook for a monthly fee for a thing, I might be a little bit nervous. Well, I mean, especially if it's not delivering. And that's, that's the other thing is that it's like, okay, I don't know if, you know, us shredding a bunch of money on subscriptions is also ending up in people losing jobs per se. Because I think that a lot of these professions that have this gigantic money factory, I don't know if they're hiring. Because if they were hiring, maybe their service would be better. Maybe they'd be offering different things. You look at something like Cloudflare, Cloudflare is a remarkable company that has only continued to offer more and better things. And like, as I've gotten more into the vibe coding thing, it's amazing how much all roads lead to Cloudflare. That codex would just be like, I'm like, hey, I want to do this. Cool. We're just going to plug it into Cloudflare. We're going to do it like, and it's where it works right in codex. You can just deploy it right through there. And it's like, wow, it's, and guess what? Their prices are competitive. They have phenomenal customer service. They continue to offer new products that stay up with the market. And they're doing great. They're doing fine. If you're a hosting service that is relying on the idea that people are just too bored and lazy to move a thing, and instead they will just justify the cost in their head as a lazy tax, you know, the lazy taxes might go down. You know, the churn, I think some of these subscription services for like TV, you know, now that it's easier to cancel, it's easier to resubscribe. Like, I'll tell you what, like, it's a good time. If you want to save a lot of money, it's already been a good time to add and to subscribe and cancel subscriptions very, very easily, especially within, for example, the Apple ecosystem or your Amazon ecosystem. They've already made it very, very easy. But now we're at a fascinating time, as Justin and I were talking about the other day, where a doodad is becoming a layer. And right now, there's a lot of doodads, but soon there'll be a layer where you won't even know what you're subscribed to, because the very notion won't make sense. You'll just know that you want a thing, and then it shows up there when it comes to all of your video content. Or somebody says, hey, check out that whatever, subscribes to it, and maybe, for all you know, it unsubscribes the moment the credits roll. Well, there's, OpenAI's launched, if you go to the Codex site, if you go to, well, where do I find this? If you go to developers.openai.com slash showcases, excuse me, showcase, rather, you can see, one, they've launched a new feature for Pro and Enterprise, which will eventually make its way to everybody else, I assume, which is to make it even easier to launch apps or websites, so they'll host it for you. Yeah. Which I think, and I think that's the thing, for a lot of it, stuff for small users. I think for stuff like we do, it makes sense to go push it to Cloudflare, but they have a lot of games, a lot of other examples of stuff. And I think that I'm just excited about the things that we experienced as builders, you know, and then you guys now as coders, as other people get to do it, and then as younger people just treat it as second nature to just build a thing. Speaker 01: Yeah. Yeah. It's a beautiful time. Andrew Mayne: And you think about, I think we three have been through that period of just having so much energy and want to create something, but the tools weren't quite within our grasp or the knowledge. Yeah. And you just want something that pulls you forward a bit. And I think we're just really at a great place with that with generative AI. So they have those examples here. If you go to developers.opening.com slash showcase, when you click on stuff, they'll show you the prompts that were used to create stuff. There's some fun games and different things. They have things too, including, you know, like the sites that people are creating and then the games, like landing page examples. So when you click on, you know, one of these examples of like a game or whatever, it shows you the prompt that was used to create it. And it's also like crazy to like, I go back to when GPD 3.5 came out and I wrote my breathless blog post about being able to make the dumbest games and how many steps it took me to go through to get it to work. And now it's, you know, neon style, build it, do this, this feature, this, whatever, and then boom. You know, we're, we're verging a little bit into after things territory, but there's a sense that, that I had that maybe I was too late for X, Y, and Z, but the barrier to entry is just getting lower and lower. So if you have been on, on, on the edge of like, man, I don't want to start now, it's too late. No, it's the, the on-ramp is shorter than ever. You just start today with anything. Yeah. Speaker 01: Yeah. Andrew Mayne: You know what I did? So now that Codex can run on PCs and actually control PCs, yesterday we did the, the weird, or the bonus podcast. And, uh, normally it's like, I don't know, 30 to 40 minutes of tedious stuff. But this time when I did it, I just opened a project in Codex and I said, Hey, I'm going to show you what I do. And I want you to see if you understand it and then tell me if you could do it. And so first I took a screen grab and I showed the file and where it is. Then I moved the file to where I moved the file. And then I opened audition and then I selected the preset. And then I unchecked the two redundant parts of the preset to sweeten the audio. Then I sweetened the audio. Then I exported just screen grab of each of these. There's maybe 20 screen grabs at the end of it. And I, and I say, were you able to piece together what I did? And then it, boom, came out with a 20 point checklist. And it's like, well, can you do it? It's like, yeah. Although I would advise you to not let me publish, let me get up to the point. And then you could check X, Y, and Z to publish. I'm like, that's great. I'll see you in a week. And then next week I'm going to say, are you ready to try this together? And then, and, and this is all, it added maybe, maybe 15% to the already budgeted time for things. But then within a week or two weeks, it'll be not a thing I do after lunch with Justin. It'll be a thing I begin and then go to lunch and then come back to find out that it's done. And watching it happen on your phone. Yeah. Yeah. Connecting your phone is amazing. Brian does not have a lot of time. And I think everybody sort of justifies going, well, I don't have the time to do this. I don't have anything to do it. And it's like, we're, you're, there's a concept called technology debt, which is where you build so fast that you create all these problems you have to solve later. There, there's also technology debt in it. You took so long to wait to do anything that you now find yourself, the cost of catching up is too, is so high. Yeah. The, uh, if, if you're on the fence, you can maybe set a rule of like, uh, however much time you have, you could probably add 10% to see what's possible. So just, just make, you know, and, and again, the cognitive leap is a lot. It's a lot to rethink your entire workflow of stuff and it's exhausting and I don't blame anyone who doesn't want to. There's a solution for people. So there was a meme going around, uh, in an anti AI sense that was like, Hey, we wanted robots to do our laundry and balance our checkbook, not to do our art and write our poetry. Uh, I'm here to tell you with codex, have it do the figurative laundry and checkbook balancing. Yes. Don't look at codex as doing the cool stuff. Speaker 01: Yes. It's not the thing that steals the thing you love. Andrew Mayne: Yeah, exactly. Do, do the, do the stuff that's annoying to you. Like I have to pay out multiple shows like, uh, every month it's coming up. It's like right now, this is like this weekend I have to pay out shows because you do shows with friends and everything comes to one pot. Somebody has to be responsible for divvying it up. A lot easier with codex. I can have it go hit the button, go to the different things. Uh, and I'll try to automate more of that process this week, but I tried to automate a little bit of it last month. Uh, do it with that because a, that's a bigger lift than trying something creative. Because if you're trying something creative, there's a lot that goes into it of your own judgment and that can be fun, but you got to trust the tools first. Um, it's way more rewarding to try the tools to do boring or frustrating things. Uh, because when those break down easier, you're like, wow, now I'm saving time. Uh, I, I, I don't know if I told you this, Andrew, but, um, one of the items that has been a 25, 27 year, uh, albatross around my neck was when the eight, six, six toll free exchange opened up, I was able to nab eight, six, six, four magic four and eight, six, six, eat fire. And I knew that those were pretty good vanity phone numbers. And for decades, I've spent 30 to $40 a month on a now antiquated thing for maybe one call per quarter, but I'm not ready to give it up. And then when, once I started talking to my money as a cloud, it said, well, why don't you get rid of it? And it's like, well, you, instead of that, you could pay two bucks a month for Twilio. It's like, great. And I got as far as the, oh, my kryptonite, the bane of my existence here, Brian, just fill out this five page feed PDF. That would have been the end of the discussion. I would have reached in my wallet and 35 to $40 would have gone out again. But instead I said, but what, will you take out the laundry and do the dishes? And it said, yes, it took me 30 seconds. And now I've transferred them and it's done. Yes, I, I, I, I've got, I get these domain updates. Like I'm paying like 24 bucks a year for like shockmagic.com. Not a lot, but more than I should be paying like 10 bucks elsewhere. And over the years, you know, that adds up. Now it's easy enough for me to press the button or write the code to tell Codex, go move this to Cloudflare, you know, dot, dot, dot. And I, I've been counting up how much I've been saving on all these little things, all these little things. I go, eh, it's a little inconvenience, but then you get to a point where it just becomes so easier to do it. It's, it's, it's, I think, yeah, I think, I think some other companies are in for a big reckoning. And, uh, I think that we're, we're in a, it's a, it's an amazing point to be in. Um, I will say this, um, I want to go to such stress and brought up the people who are like, we didn't ask AI to do our art thing, poetry. Like, guess what? You can still do art and poetry and write. I still write books. I still write novels. I am writing a book right now that I'm even not even using dictation. I'm literally typing it with my fingers and I get to write that book because I have more time to write the book because I'm letting AI do a bunch of other stuff. But also when I hear people say, we didn't ask it to make our art or whatever, it's like, eh, you're, you're not talking about your art. You're talking about the idea of having to compete with it, like other people's art. And I think we have to get into a point where we acknowledge. Yeah, I don't know. I mean, I, I, I kind of, you know, I, I almost don't even want to acknowledge that kind of stuff. Like Brian and I have had this running thing over the last week of just like, you know, uh, it's like, how did, uh, how, who made the art on the new world's greatest con website? Magic. If you're going to be mad about it, it was magic. Speaker 01: Yeah. Would you have a magic trick? Andrew Mayne: Like, like whatever, like, you know, like whatever, just, if you're going to be mad, you're going to be mad. Like the only thing that's different is I've made a million, not a million, I made a dozen websites that were terrible. And guess what I use for the art? Either things that I copied without copyright from the internet or terrible pictures that I took because I don't have any nice pictures of myself or the one set of pro photos that I took of myself. So that's it. Did you, that's, did, did you see the Martin Scorsese ad for the company that's using generative AI for helping like film planning? For, yeah, for, for storyboards. Yeah. Yeah. And, and, you know, in the middle of a lot of people saying things that they have to say performatively to you, that's part of it is just some people who are like, oh no. So he's advising for Black Forest Labs, which makes flux from great video models and he's getting backlash, but I also think that if anybody's earned the right to say, hey, I understand my role as an artist and where I begin and the tool ends. It's, um, yeah. And I was glad to see that. I think he's at the zero Fs to give point. I know, I know some Grammy award winners. I know some Oscar winning directors, some other people that feel the same way. They're less willing to come out about this because they're at certain points in their career, but I will tell you that some of the more, the more, the more talented somebody is that I've known are some of the more prolific they are, the less afraid they are. And there's some very, and they're very deeply into this stuff too. Hey, who's got on, uh, I'll say on after things, by the way, if you're a subscriber, that's the podcast we do after this. I think Brian and I will go into some of the ways that we've been using codex to save money and some of the methods we've had. And I would say that like, it will be the one episode that will justify having been paying for this ever. Yeah. Yeah. Uh, my pick is spider noir. The first episode is, yes, it does exactly what you'd expect. It says it does. It lays out all the pieces of a story and you get to gawk at, oh my gosh, that's a riff on this character and that, that character. And I wonder if we're going to find out how he got his powers, uh, by the end of the third episode, you realize, oh, I'm really fascinated by this power dynamic and these alignments do make things complicated. And, uh, wow, I kind of do want to know how he got his powers now because it's not this. And then, uh, and it gets really good. I'm on episode seven right now, really enjoying it. Um, they, uh, yeah, that's, uh, you get to see a de-aged Nicolas Cage move like a not de-aged one, um, not an anti-pick, but, um, we're finishing the boys, you know, it's the end of the boys still doing the, the, did they finally shock you? Did they finally come up with something that you couldn't believe? Did they finally cross the transgressive line? Justin, yeah, I just, it lost the interest of being a superhero show at some point. This is a point I've made on, on, I don't know if I think I made it on great night, but like all of the scenes, the superpower scenes, you could pretty much tell the same action set pieces if they just had guns, there's like nothing like super interesting or fascinating or like pushing the limits of the various powers, which are usually the hallmark of like really cool superhero stories. It's just sort of like, oh, boom, you're dead. Uh-oh, I could kill you. Uh, my superpower is stopping you from opening that door. It's like, I'm scissors, I can cut anything, especially paper. Can you cut rock? Like what? Oh no, I can't. Nothing can cover me. Except what about me? That would be cool. That would be cool, right? That'd be an example of a good, because then there's only certain ways you can do it as opposed to just everything being like, oh, my superpower can kill you or stop you. It's like, okay, well then that's a gun. We might as well, that's every action movie. Yeah, the, the early seasons and when they were kind of a lot more favorable to the comic, not to say the comic has to be the benchmark for what you do. There was a lot of, every power might have a weakness or they might be unaware of a thing. You know, powers make you arrogant. And, and I love that idea of like, you know, Lex Luthor knew, hey, uh, Superman's super fast. What do you got to do? Oh, make him find Lois Lane, who's in the middle of peril and about to be dropped into the cement mixer. Superman's busy doing that. Well, I get to rob him. And that was clever. It's like, oh, how do I, like, how would you, and you would see that later on. I'm like, well, how do you fight somebody who's super strong with your brain? You outsmart them. And that, that's, that's the, that's the great thing. But then you get into, I like Zack Snyder, a lot, a lot of things about him as a person, some of the stuff he's done, but a lot of his movies just been, it's now we punch more punching, more punching and more superpower punching. And you're like, where's the clever parts? You know, where's this? And that's what the boys was just, I haven't watched the latest season, but I felt like you're in a role of powers, powers are cool, powers can also be used against people because they get lazy. Well, there's also, we haven't gotten to the final episode, we're on the second to last episode, but there's a character, there's a sequence of events in which a character demonstrates a power and then makes a decision that completely invalidates the drama of the decision that he has just made, because you know, the power can instantly undo it. So it's like, oh, there's no consequences to him making this decision. Like, this is just, you know, totally useless. We're just, you know, literally doodling on a piece of paper until class ends. Uh, so I don't know, it's sloppy. I, you know, I think certainly that cast seems like they, everybody except, um, Carl Urban seems like they're kind of over it. You know, Anthony Starr still, I think, has a really good time playing Homelander and Carl Urban playing a butcher is still somebody who's like putting his whole hundred percent, but like, everybody else feels like they are just hitting fast forward and, you know, it's just one of those things where it's like, the mute character started talking and it's just, it's, it's like, did the masked character show his face? Yeah, it's like, it's like, it's like, okay, how much did the agent bitch or like, was it in the contract renegotiation that like, the actress wants to talk because otherwise everyone's going to, she's going to be typecast as a mute or something like that. And it's like, oh, okay. All right. Well, I guess that makes you less interesting as a character, you know, cause that was the cool thing about you is that you had this limitation and now you don't. I, I hear you. I think that it's just, you, you, the longer these shows go on, there's a problem. There's the reason sometimes showrunners just want to do like five seasons or they want to do that because they know at a certain point, your actors are going to get rambunctious. You're going to have more demands. It gets higher. And you're only going to make so much money per season and it's just, and the costs go up so high and it's hard. Network's like, oh, we want it forever. And I think there's ways to do that forever. But then also you just sort of write yourself into a different place. My favorite story though, is that, uh, Anthony Starr, they were shooting a scene in like some big auditorium filled with people and Anthony Starr stresses Homelander and somebody's cell phone goes off and he just yells at the crowd, like, shut your goddamn phones off. And they're like, we're watching Homelander scream at us. This is so awesome. That's amazing. There was, uh, there was one story, uh, during the monoculture, Happy Days was a cultural tour de force. And there was, this is in the before times where giant events would happen and more people could show up than could enjoy the events. And there was a scene sometimes mob frenzies would occur and with 5,000 people in this one area all there to see Ron Howard and, and, uh, Henry Winkler, uh, there was a moment that they were in a dangerous spot. They didn't have any bodyguards. There was an over promotion to the event or whatever, and they needed to get from point A to point B. And if it was the kind of thing, if one person started getting paw-y on them, it was going to be a problem. And Henry Winkler did something he's never did again. The notoriously introverted, shy Henry Winkler adopted the persona of the Fonz and he leaned back and goes like, all right, listen up. Here's how it's going to be. You're going to make a tight line all the way down. Now move it to the side. One person touches the leather you're done for. And everybody responded to the character and it was like parting the Red Sea. It was amazing. He had such power as the Fonz. I love Henry Winkler, but he could have done so much more good in the world. Mm-hmm. Mm-hmm. Did you see the Russell Crowe, uh, clip that went viral? Of him with the crowd in front of, like, the hotel, I think, in Italy or France? No. Should I look it up? So TMZ says, oh, Russell Crowe, you know, look how rude he is to his fans or whatever. It's him stepping out of a hotel, like, just being besieged by autograph sellers. You know, let's be clear. These were autograph sellers who already had, you know, gladiator posters, things like this. The whole idea that this was just an organic group of, like, fans. And also, somebody conveniently recording the whole thing and knew how to sell stuff to, you know, TMZ, right? And, oh, listen to this audio. I want you to hear this, Brian. Start it from the beginning. Listen to the audio. Okay. There we go. Yeah, I'll go back, uh, from the beginning. Oh, no, I can't. Well, I'm in manager mode. Here, I'll just refresh it. Speaker 01: Are you listening? Are you listening? That's what I said. Are you listening? Yes. Stay where you are. Look, dog. Ocean on me. Sure. I'll come to you. Everybody's space. She was listening, somebody's a dick. Okay, okay. You got me? Yes. Clear? Yes, sir. Yes. He's signing. He's signing. Andrew Mayne: He's signing. Everything being given to him. You can add Maximus. One guy asked him, like, would you do a sign Maximus? And he's like, no. Maximus. Speaker 01: No. Andrew Mayne: And he got, like, TMZ wanted to say, oh, look how, I'm like. Wait, that's the. He's signing. Okay. Yeah, they were overselling it. That weren't nothing. That was good snake management. Yeah, because he said, I'm on my way to the airport. I step out. He doesn't have any staff with him. There's this whole crowd of people who wants his autograph. The hotel staff's like, what's going on? People want to get out? He's bandaged it. I thought, like, he was in control. He didn't say no. Wasn't rude. He said, here are the conditions. If we're all cool on this, use, you know, Russell Crowe language. It's fine. And then he signed everything. Even though knowing full well, these are not fans. These are people going to put them on eBay or mark it up or sell this. I was like, to have that kind of control and discipline and to get everybody's needs met. That was assertive. It wasn't aggressive. It was, I will help you just follow these rules and everything's going to go smoothly. The moment you break up, you create a problem, whatever, spawns. The moment you touch the leather jacket, whatever, we're moving on. It was great. Yeah. What's your pick, man? I got two picks here. Okay. Pick number one is, I'll just do one pick right now. Um, I watched the Hulk Hogan biography on Netflix and, uh, it's a four, I think it's a four episode Netflix bio. They did spoiler word, uh, Hulk Hogan dies and, uh, in the middle of production of this thing and, uh, loved it really, really loved it. They did. They went deep into a lot of stuff, a lot of the controversies. I think you came away with a pretty good understanding of the guy, people close to him might feel differently, what have you, but I actually thought it because like, there's a lot of stuff like later on when he went political, which Hulk had never done that before and whatnot. They got flack, handled that, handled, you know, some of the different controversies and stuff. I thought it was really well done. That's great. Speaker 01: And that's on Netflix. Andrew Mayne: Yeah. There's a tweet that made me laugh because there's been a rash of wrestling documentaries that have been kind of mainstream on Netflix. Uh, it's just like wrestling documentaries are great because it's the rock wrestling, somebody dressed as a volcano. And then Vince McMahon comes on and he's very serious to the point of crying saying that man was the greatest volcano that ever lived, just like his father. It's, it's funny to like, the funny thing for me too, is when you go to Wikipedia and you read wrestler bios and they will talk about these matches like they are Olympic events, you know, for this, he was up against so-and-so and he defeated him in this, like, all right. It's like, you know, it's like Patrick Stewart playing King Leo and third act, it was a little bit unsure if he was going to be able to deliver the powerful speech that would rally every, like, listen, everybody, you're talking plot, it's fine, like, we're, we're, we're shifting from their career to plot points of characters they've played. Yeah. So that's my pick, uh, the Hulk Hogan wrestling, the biopic, oh, and my other little add-on pick is, uh, they mentioned in there, you know, how the Rocky appearance kind of pushed him over the top and, uh, so, you know what I haven't watched in a while, Rocky Balboa. Hmm. The 2006 film, when's the last time you guys watched that? I think 2006, I saw it in the theater. Probably that, yeah. It is, it is really worth a visit because, um, it's probably one of my favorite Rocky films and, you know, it's, it's a great, cause it's, you go watch it and then you watch Rocky, like, also the first Rocky, there is the version of Rocky in people's heads and then there's what Rocky is. Rocky is a story of a broke guy with no family in Philadelphia clinging to his community and the people around him. And by the way, it's a really cool job offer at the end. And, you know, he's a smoker, he's out of shape and whatnot and has to sort of meet, meet, meet the opportunity he never got. And the opportunity is a boxing match, but that's neither here nor there. Yeah. And the funny thing too, though, it's like, you go watch Rocky and you watch Rocky Balboa, no disrespect to Philadelphia, but you watch Rocky and you're like, man, Philadelphia looks really rough. And they talk about how it's declined. Then you watch Rocky Balboa and you're like, oh my God, I didn't realize Philadelphia could look worse. Oh, wow. Because like they'll have like, they'll be like, ah, this is where so-and-so live. Fast forward 30 years later. Oh yeah, that building next door is burned out. It's fire. So they have to fix the fire damage and buses don't come here anymore. There's no guys on the stoop doing singing doo-wop. It's just, you know, you know, meth heads and whatnot. Amazing. Yeah, it looks musical the street toughs are these days. Yeah. So, anyhow, gentlemen, it's been weird. That's weird. UNKNOWN: Yeah.