Phil Rice 00:59
Hello and welcome to And Now For Something Completely Machinima, the podcast about machinima, virtual production, whatever you want to call it. We don't care. This episode. We are going to, well, first I'm going to do, I'm going to go ahead and introduce us. How about that? Let's, let's actually pretend this is a professional show or something. My name is Phil Rice. I'm here with Tracy Harwood, hello and Damian Valentine and debuting on the show today. Generative AI Ricky Grove, Ricky, say hello to everybody. Ricky. This is something. What

Phil Rice 01:53
was that? Well, he'll, he'll maybe, maybe it needs more, it may need more input to to get a proper response, so we'll just but anyway, Ricky is away. Everything is is good, but with his cooperation, we are implementing this, this generative AI version of Ricky. We're all very excited about that technology, very optimistic about it. And this is a great way to kind of show off probably the future of podcasting. It's going to enable us all to take vacations when we want to. It's exciting times. So we will be consulting. We'll throw it over to Ricky here and there. AI Ricky, excuse me and see what he has to contribute. So this episode is all about the news. What's been going on. So Tracy, why don't you lead us through this?

Tracy Harwood 02:47
Oh, my god, yeah, I've got quite a bit to tell you about, but let's start with something that I thought. I haven't really found too much out about this, but it's something we need to talk about. So let's, let's sort of reflect for a moment on the launch of the Nvidia machinima app, which was, when was that launched? Not that long ago, a couple of years ago. Okay, well, we've not heard very much about it for quite some time, and frankly, I can't find very much about how it is being treated, I think, as a legacy app, because it does appear that it has disappeared and has been probably subsumed into Composer. Now, the only thing that kind of alerted me to this was that I caught sight of a short that apparently celebrates its demise, which is by a little video by Pekka Varis, and that video is called restaurant RTX 4K Camilla Toon, and I'll share the link on our show notes for that. But maybe you guys have seen something about this, because I really can't find very much out about what's happened to the Omniverse Machinima app.

Damien Valentine 04:15
So I can answer this, because I use Nvidia Omniverse for Heir to the Empire. So Nvidia Omniverse is a complete set of various pieces of software, which included the Machinima app. And so what happened is, you got machinima and you've got create, and there was a couple of others, the there's not much, there was not much difference between create and machinima to the point I was trying to think, what's the point of developing two apps that do almost the same thing? There was the small differences in features, like create was the one that got the priority new features, and then they would get put into machine learning, but later. And there were some things that machinima could do that create couldn't, which is more like. Can to create them more for if you want to put something together, and you can render animated scenes in it. But if you just wanted to bring in some props and, you know, decorate them, you'd want to use create and then machinima, if you wanted to do a lot more animation yourself, within the software, Machinima is probably the better one to use. But there was such so much overlap between that. I think at some point, because I can't remember exactly when the switch over happened, I just it just did, and then carried on using it. They merged all these apps into one, which is now called Composer, right? And it makes a lot of sense, because why are they developing two apps that do almost the same thing when they could just do it with one app and then incorporate some of these other features as well that some of the other apps did. So it's not a bad thing that it's gone apart from maybe the names disappeared, which it was nice to have a an app called machinima. But as far as actual usability goes, it and, you know, keeping the software updated, and it is still. Composer is being updated quite regularly. There was an update a couple of weeks ago, which I haven't actually tried yet, because I keep using the previous one. But you know, it hasn't disappeared. It's just been merged into one app. And I think that's the right thing to do as far as development goes, and as far as keeping everything together because, you know, switching between apps unnecessarily. I actually rarely used machinima app itself. I did everything in create because create could do everything in because I'm also using iClone to do all the really advanced animation, like moving the characters around. I didn't need to worry about that as far as machinima went, because they'd already done it with iClone. So you can still do that. You can animate your scene in icon. You can export it over to Omniverse, and it just loads up instantly in Composer. So it's not a bad thing. It's gone

Tracy Harwood 06:59
okay. Well, so here's the thing I just want to say, then, thank you very much to Nvidia, because what they have done with the development of that tool, using machinima as a name, was help us take back what machinima was all about in the beginning. So you know the fact that it's kind of disappeared doesn't mean for me that it's been a pointless exercise at all. I think what they've done has been hugely important to the Machinima world, and I think I think we probably all reflect on that as being a worthwhile exercise, and I'm very grateful to them for having done it. I

Phil Rice 07:46
agree. Yeah,

Tracy Harwood 07:48
okay, well, okay, so what I then wanted to talk about, given that machinima, is, you know, I think what amazes me is just how many films we're seeing of so many different varieties come through to us each and every month. It's just it's really tricky to kind of make our selections and and think about why we're making the selections, and what we're trying to illustrate by making those selections. So I wanted to just highlight some of the ones that didn't quite make the cut for our full discussions, but are nonetheless some really cool machinimas that we've seen. And not necessarily always machinima, but some some just fascinating projects. So projects that I want to sort of flag are one called Endgame, which is an upcoming James Bond inspired short. There's an absolutely superb soundtrack, which includes music by Dark Machine Studios, whom we've also reviewed in the past. The aesthetic for the video is absolutely awesome, and I think that's something I can recall that we discussed in the last video that we looked at for Dark Machine Studios too. You remember the one that he used AI for with the with the yellow, the yellow filter? Do you remember that

Phil Rice 09:15
Mark Johnson? Is it? Yeah, yeah. Yeah. Really

Tracy Harwood 09:19
cool. Well, this one's a more James Bondy type psychedelic key sort of thing. Really interesting. Very much looking forward to seeing the whole film and it kind of comes out. And then another artist that we've reviewed before is actually JP Ferre, and he's released a film called DCS: Spitfires - Cinematic. This is a really powerful film, actually. It's used an AI generated voice which presents the imaginary thoughts about the recollections of a Spitfire pilot from the war. Yeah, it's really very moving, and actually, I think it's especially moving. So don't know if you've seen this on the news over in the US, but just a couple of weeks ago here in the UK, that the Battle of Britain Memorial flight was was grounded because one of the Spitfires crashed and it killed the pilot. And that all took place near to where the thing was was based, which was incredibly sad, but, but this film is really moving with what it what it's portraying. So definitely take a look at that. I think he's done sterling job with with the with using AI as well. Then Pink Floyd has announced all of its winning music videos. Now, the one I was especially impressed with was created by theDavedood of Scratby films, and it's been done in blender, and it's been done for the single time, which was from The Dark Side of the Moon, as part of their 50 Years celebration contest. It's really beautiful work. It's, yeah, some really interesting portrayals of time and death and whatnot in it. You've got to take a look at it. It's really interesting. And then another one, Anomidae, have I said that? Right? Annomidai, I think so. Yeah. He's posted the next episode of his Half Life supernatural series called Interloper. It's another cracking instalment. This one's interloper nine something under the skybox. It's quite, quite a length this one, but it just deepens the mystery even more he's flying with this series. It's really worth watching. And then, Phil, you'll be really interested in this. I found two Fallout inspired works interestingly. I mean, I guess this is partly the aesthetic of the game as well. But both of these films, I think, are taking what I would call a sort of Tom Jantol anymation approach to the way that they're presenting the material. So it really seems to have captured the imagination of of the of the creators, although they are completely different types of creator. One of them is JT Music, and he's done a they've done a rap called All in With the Fallout. Now, obviously it's taking footage from the game Fallout four. And quite interestingly, when I was looking at it, I noticed that Harry101UK, did the video for it. Now, if you recall, a couple of weeks back, we reviewed Harry's portal film The Sound of Science, which was brilliant. So I didn't realise Harry was doing the videos for JT Music, which I think is really cool. And then the other one is called Fallout Dream On (a tribute) by Couch Patrol, two completely different films, but the way they've done them is quite similar, really, quite interesting. So take a look at those. We'll put the links on the on the show notes. And then one final project that I wanted to highlight, and which leads me nicely into the next part of what I wanted to talk about is one that's using generative AI. Now this one is an animation of Anthony Hopkins, and it's called Fables of the Foolish, and it's by Dreeko. It's just delightful and so very witty. It's an animated character of Hopkins retelling what Dreeko refers to as the fatefully foolish exploits of idiotic protagonists who meet their untimely, ironic demise thanks to their own oblivious actions. In other words, known as the Darwin Awards. It's a sublime comedic parody. It's brilliant. You've got to just see it. It really sounds like Hopkins. I'm guessing he's used ElevenLabs or something like that, too.

Phil Rice 14:00
Yeah, I'll bet it's an 11 labs voice profile.

Tracy Harwood 14:03
It's absolutely brilliant. Spot on. It's it? Yeah, I really loved it. I guess that might be a point at which we can ask the Ricky AI what he thought about it. Yeah, Ricky, what

Phil Rice 14:17
do you think about this? This Anthony Hopkins film, the one about the Darwin's awards,

Ricky Grove 14:25
Salvation Army, is a great place to go to get fish. A lot of people don't know that.

Phil Rice 14:35
You know what I think? You know what I think AI Ricky is doing here is very clever. Is he's kind of, I don't, I don't know what he's doing there. Yeah, we

Damien Valentine 14:49
about to think about that one a little bit.

Phil Rice 14:50
Ricky. Ricky has always been a fan of the theatre of the absurd, so maybe, is that what you're trying to tell us, Ricky? Or are you? Are you? I. Making some kind of a absurdist commentary on this, slowly,

Ricky Grove 15:05
Slowly the cat came up the road and told me that Sally is going to be late for school. Now, why a cat would do that? No, no, no, no, no, no idea.

Tracy Harwood 15:23
No, I think, I think, no, I don't know what he's talking about there. Okay,

Phil Rice 15:28
well, thank Thank you. Thank you Ricky for that brilliant Okay, let's move on. Yeah,

Phil Rice 15:39
Ricky, what do you think about if we, if we modify your generated AI version to be skippity, Ricky, is that something that that you'd be interested in?

Tracy Harwood 15:39
let's Okay. All right, so AI is genius in it. My AI updates. Then, firstly, ElevenLabs has released a dubbing tool set tutorial, which actually makes, I think, what makes it quite helpful for doing different languages and different language videos even more accessible. Can't wait to see what you all do with that actually. Then Google DeepMind has announced a new video generation model called Veo which produces pretty high res videos actually for over a minute in length and using a whole range of different video styles. So that looks quite interesting. Stability AI has launched stable artisan to a wider user group on Discord. It's a tool for media generation and editing. So it's not probably quite our audience, but it looks pretty interesting for you know, maybe creating promotionals and what have you. And then the one that really did interest me is a TV show generator, which has been launched called Showrunner now that's been created by a studio called The Simulation, and it's basically a text to episode generator. Now these are the same guys that released that weird South Park AI episode last year. So I can't wait to see what folks do do with this. There's already a few shows up on their channel, if that's what you want to call it, although some, in fact, are referring it for referring to it as the Netflix of AI. Now there's a waitlist operating to get early access to it, and I'll put the link on the show notes, but it looks quite intriguing. I think. Then the winners of the second AI Film Festival were announced during the month by Runway. The last month, the best film is called Get Me Out, and it's been done by Daniel and tebby. It's an, it really is a beautiful film. It's actually done with a with a realistic central character, but it's an incredibly moving and powerful story. It's definitely worth watching. It's been it's told in Japanese, actually, as well, which is which somehow seems to make it even more compelling. So definitely have a look at that. And then finally, on the on the AI side of things, on LinkedIn, day job, so to speak. It's not on YouTube at all, but this this post that I saw by a guy called Luke Shurgers, which I'll put a link to. It includes a video parody of Sam Altman talking to an interviewer, I'm guessing Luke, about how to make it big and also how to generate an audience using social media. And it's, it's been made using SigmaGPT. But what's such fun about it is the creator has used Altman's head sticking out of a toilet, which he's created using Replikant, and he's called it Skibidi Sam. And it's, it really is comedy gold. I have to say it just absolutely it is hilarious. You definitely should, should have a look at it. I'm hoping that a link to a LinkedIn post takes you there, whether or not you're on LinkedIn, so I'll put that on the show notes as well, and that's it. That's all my news for this month. I don't know what you guys have got?

Ricky Grove 19:35
I wonder how much longer it's going to take the sun to explode. Don't you think it's been a long time now it's over due.

Phil Rice 19:46
Hello. Okay, this is, this is not going how I expected. But thank you. Thank you. Thank you, Ricky. That's That's great. Maybe

Damien Valentine 19:57
just didn't like this. Give it a Ricky thing. Uh. Could

20:00
be,

Damien Valentine 20:01
yeah.

Phil Rice 20:02
Did he sound angry? It's really kind of hard to read.

Damien Valentine 20:06
Yeah.

Phil Rice 20:06
Damien, What news have you got?

Damien Valentine 20:08
I got a few pieces of news. I'll start with the AI one, since we just had some a news. This is from Sony Music. So it's not necessarily about machinima itself. But they have opted out of AI training for all of their entire music catalogue. So they've gone to all the AI training companies and said, Please do not use our music to train your AI because we were not interested. Now, I posted this in our little group chat that we have for us. And I saw yesterday that I've noticed half the posting, because I saw yesterday, Sony have also taken the unusual step of this is going in the complete opposite direction of using AI in the upcoming TV shows and movies, but not their music. I didn't have a chance to read too much into that. But I imagined that they're going to somehow use AI in their future visual productions. But they don't want their music as part of it. And I guess we'll we'll have to wait and see what's gonna happen there. Some may say the quality of their recent spider verse may be improved by AI. Maybe given some recent experiences with more AI superhero films, that says a lot. Alright, so moving on to some more interesting, more machinima related subjects. We talked about the Backrooms last year. The the origin, the start of the whole thing has been discovered by some internet sleuths. It was a picture that no one knew where it came from, of the rooms at the yellow floor on the

22:00
walls. And there's no way I remember it. Right? It

Damien Valentine 22:03
turns out, that was from a hobby shop that was had been emptied out because it's in the process of being refurbished. And that was taken back in 2007. I believe it was posted online. And then the people who own and run this hobby shop, they still exists still runs. Now. They had no idea that a decade or so later, someone would find that picture and make this whole Backroom thing. They had no idea about any of it until these detectors, tracked them down and contacted them. And I think they weren't quite sure how to react to that. But I think they're quite pleased. And then they've actually gone and posted some pictures of what the shop looks like now. That's ruined it. Yeah, I mean, they've got windows now. It's all brightly lit. And you know, they've got things there's a huge sort of remote control car track that people can take their models on race on. And it looks very different from what we've seen in the Backrooms.

Tracy Harwood 23:07
Disney would be up for doing a theme park with the Backrooms.

Damien Valentine 23:15
That would be very creepy. But yeah, I can see that. Creepy,

Phil Rice 23:20
very, very low budget.

Damien Valentine 23:21
Yes. We talked about the Backroom for for I thought I had that to do it, because it kind of part of the story of where it originally came from. Well,

Tracy Harwood 23:33
it's a nice wrap. Thank you. Yeah.

Damien Valentine 23:35
And last bit of news I've got last month, we talked about a spiritual successor to the movies called Blockbuster Inc, which, at the time recording is a few days from release. By the time this goes out, it will be available on Steam. Obviously, we haven't had a chance to look at it yet, because we're not there yet. But there is a playable demo of that she could try out. But I found out that later this month. So after you see this, there is another spiritual successor to the movies being released called Movies Tycoon. And there was a demo of this as well which I have not had a chance to try out my had a look at it. And that looks very similar to the movies as even a there's a picture I saw of the the main menu. And there's a option for just movie making. So imagine this is going to be like the free from the mode in The Movies where you just have everything and unlimited money. So you still build your studio. And you can recruit the actors and do all that kind of stuff, but there's no restrictions and you can just do whatever you like as far as movie making goes. So this is a month for The Movies, spiritual successors. So I'm quite intrigued to see what happens with both of these games. Because I did enjoy the movies. I know it had some limitations of what you could do with it as far as information goes. Technology has obviously moved on along a lot since then. So I'm interested to see what these two games are going to do. As far as making things more accessible. I know that Blockbuster Inc has built in modding capability that they really want people to bring in their own content, and costumes and sets and animations. I don't know how that's gonna work, because obviously, we haven't seen it yet. But I'm hoping the Movies Tycoon will be something similar, where it's very easy to bring in custom content. And it allows a lot more flexibility with movie making than The Movies had. Because these are going to these have the potential to be too new is to use machinima tools. So I look forward to seeing what people do with them. And the games themselves.

Tracy Harwood 25:50
Yeah. It's interesting, isn't it? Because on the one hand, you got AI, which means you can create your own show. Yeah. Showrunner. Or make it up in a game. Yeah. Which would you do?

Damien Valentine 26:06
I just the game, because then you get more control over it. And you're gonna get the result you want.

Phil Rice 26:12
Yeah, that that Showrunner thing? did? Did you guys watch the most recent season of Black Mirror? No, no. Okay, so there's an episode. In that most recent season, it's on Netflix. And in that episode, Netflix. This is a Netflix show. But in the show, Netflix has in this episode has a quantum computer that is generating AI shows in real time. individually tailored to the people watching. Yeah, this was released about a year ago. I

Tracy Harwood 27:03
want to say this is Showrunner, isn't it?

Phil Rice 27:06
I mean, it sounds so much like that right now in of course, black in typical Black Mirror fashion. It gets really creepy. And actually, they do a thing with it. I don't want to spoil it. But it goes deeper than that. Like it's it's hard to explain. And I there's there's no way to do so without spoiling it for anyone who hasn't seen it. But you know, the general premise is that this big media company Netflix, and they actually call them Netflix and thing. This girl is watching Netflix. Her and her boyfriend are watching Netflix, and the show that comes on up in there and the algorithm brings up for the recommendation and they watch it is what happens to her that day. As a show, with actors and everything. And at one point, one of the actors is Salma Hayek who if you don't know who that is, by name, you would know if you saw her face, very well known actress and she is upset that Netflix used her persona and voice and everything without her permission. But and when she goes to complain about it, it turns out that she signed away those rights and didn't didn't read the fine print kind of thing. I mean, it is it's insane. When that came out that was really before that was just prior to AI really kind of you know the horse breaking out of the gate. And now it's like it's it's not nearly as implausible as it might have seemed, then you know, it's it's crazy. That how precient that was. So that's worth seeking out even if you're not a fan of that show. Generally, I believe I don't know what the episode title is. But there's only there's only a handful of episodes in that season. Just look for the one that has Salma Hayek in it. It's hysterically funny what they do with it. But the premise there of this, you know, AI generated just the other day on a podcast that I listen to. It's a comedian does a podcast, and he was talking about having dabbled with AI music. You know? Suno I always mess up the names. Yep. Suno and was raving about that he could just type in. I want to hear a song from the 1950s about a car and it will just create it, and that he just loves it. And he basically just said, the media, the statement that he had said, that really stuck with me. And that should be a little bit alarming is the median quality of an AI song in this current technology is better than the median quality of human made music that's out there right now. And yeah, that's an averages thing. But it's one way of saying there's a lot of, you know, there's a lot of garbage music out there, there's a lot of, you know, either established artists are just not making good music, or, and then there's also like, a lot of just amateurs who, you know, maybe shouldn't be putting their music out there, you know, quite so. So boldly, because it's not great. It was it was just really jarring to hear. Hear someone actually articulate that. I guess, it has to do with what your attitude about music is. If it's just about filling your ear with with a pleasant sound mixed with, I don't know, immediate gratification, then nothing's gonna compete with AI in that regard. But if you're actually, you know, if your connection with music is something a little bit more deeper, or you're wanting lyrics that are meaningful, or a song that's from the heart. You know, it has that's not a nice strong suit. It's the lyrics are. They're great for comedy, I think. But they're not really. I don't know, there's the the missing soul is evident in the AI music, but in terms of can it make a cool beat and a neat sound? Yeah, yeah, it can. And I think the same is true for some of the AI video, the way it's going, is that it's going to make really good looking pieces of candy. But, you know, if you want to, you know, a chef prepared meal. Not sure you're ever going to get that. So it's I don't know, it just really stuck with me as this sense of there is a mark, there is a market for it,

Tracy Harwood 32:24
I think, yeah, I agree with audience. It depends what your attitude is to the, to the use of it as well, because I think, if you are using it as a tool in your creative process, and it's, you know, co creating with you, which, you know, which then gives you more options through which you can select stuff that you might not have been able to create yourself. Or you're not, you might not have been able to get to as many options yourself on your own. Without it. I think that's one set of thinking with it, if it's replacing entirely what you might want to do with it. That's, that's another thing entirely. But my my thoughts really are is it's a tool, you look at it as a tool, and you look at it as a tool within your creative process.

Phil Rice 33:20
I'm not sure if I've said this on this show before, or if it was elsewhere, but I remain of the opinion that the most effective. At the end of the day, the most effective users of these AI tools be aI generated art or video or music, or ironically enough going to be artists. Absolutely. Because if for no other reason. Artists have gone through the process of learning what the creative process is. And so they know the language of it. And this is all about language. All these models are the generative AI models, they're all about giving the the the model the right language. I mean, that's why they that's why the that's in the name large language models. That's what they are. And so if you if you are a master, I've told my son this, as you know, he's he's in college right now and at university and we've had discussions about what skills are going to be valuable, you know, hey, I'm taking this course and I don't, you know, seems like a waste. And the thing that comes up continually that i i advise him on is that probably the single most important skill other than maybe learning how to think, which I hope universities still teach that have questions about that sometimes, but is how to write which and frankly, there's there's more than Few very smart people that would tell you learning how to write properly is learning how to think that it's the same thing. Really, it's the same same brain activity. And I think applied to this AI stuff, that's, that's, it seems to be undeniably true that if you're going to adopt these as tools and navigate them, the most important tool is language is mastery of language. It is, and maybe mastery of a specific type of language, too, it's not the same mastery of language that would make you a good social worker, or make you a good counsellor, or make you a good lawyer. But getting these tools to do what they need to do, the people who do it best are those who take the tough that you can't just go and talk to it, like you talk to your buddy, and expect to get the kinds of results that the people at the top of this game are getting, they're doing that because of lot of trial and error. It's a lot of work. It's just, it just looks different than other creative work, but it's a lot of work. A lot of experimenting on the language level to get to get out of these things. What what happens? So yeah, it's it's, it's, it's interesting,

Tracy Harwood 36:20
it is what I was gonna say was that, you know, that because it's because it's it's using, you know, data datasets, you have to remember what the big data set is that is that it's drawing on. And a lot of it is also it's not just text, and images that it's drawing on, it's drawing on the metadata related to the description of the images and what have you. The, the hidden part, and the part that people are now beginning to ask questions about is what they call the, you know, the para data, the data in relation to the process of how, say, a visual piece was was created. So there's actually different types of descriptive data that, that that can be used. But, but there isn't an awful lot of paradata data that describes the process of making something. And I think increasingly, as people start to integrate these kinds of tools, in the development of their work, there will, there will emerge a new kind of way of thinking about how to create stuff that is already having to think about how we make sure that we capture the human side of the creative process, using AI's in order to attempt to claim copyright of what comes out of it. That's apparently that's kind of a parody to type. Were thinking about it, but the fact that that paradigm doesn't actually exist that much at the moment is a problem. And how do you how do you capture that? It's really hard to figure out how it's, it's captured, but it needs to be captured in order to sort of understand how to how to develop new ideas using all this, you know, all these different different types of of information if you like.

Phil Rice 38:26
So, yeah, I think that I think that's important. And I also think I think a smart way for this stuff to go would be that every so if you've ever done AI generated artwork, I would say, particularly with the Leonardo platform, when you create an image in there, or generate an image in there, all all kinds of metadata is assembled around that object that you've created, you know, what, exactly what engine was used? And, you know, all that information is documented and attached to that image for all time. Well, the the information that needs to be added to that metadata body is attribution. Absolutely. Because if they can figure out how to do granular attribution data for these generations, then at least the info is there to then figure out now what do we do about this with it with regard to rights, royalties, all that I think the engines have that information? Yeah, when they generate I don't think it's just some sea of conglomerated. Info and I suspect that these engines know all the things that they're pulling from and maybe it's just overwhelmingly large, you know, the list of things that have been pulled to generate an image, maybe it's, you know, most images that are that have any complexity at all, are probably not pulling from two or three specific items in the training database. It's probably many, many, many more than that. But at least show your work. You know, I feel like that that could be the beginning of how to solve the artist rights issues that emerged from this is at least at least document where it came from. And okay, so then that creates all this extra data that's hard to parse through, well, what what better thing would an engine like, let's say chatGPT be good at then parsing and making sense of and summarising and conglomerating that data, so that any given artist can say I want to see, you know, I want to report of all the places where my artwork was, was used, but starts with that information then to figure out what if anything, can even be done. Monetarily there? You know, that's, that's, that's tough, because I think you're, you know, Ed's that's a, that's a, that's a problem, ironically enough that only AI could solve. There's no human institution that could navigate through all that and come up with, I'm thinking in terms of like, you know, comparing it to like a royalty sheet that would be generated at a music publisher. Yeah. So you know, Joe Smith has such and such songs, and they've been played on radio, and he's, he's a member of ASCAP. And so they're collecting all these royalties, and all that kind of trickles down into it, ultimately, a statement and a check he gets, right. And I think that's what maybe people that are hoping that this is monetizable in a fair and equitable way that they're hoping that would happen. But we're talking about data, that's just, it's way too big for those largely human systems, you know, those those systems were all formed when, you know, radio disc jockeys used to write the stuff down in logs by hand and mail them in. They were doing that in the 50s and 60s, there were no computers involved. And all that's being done with that stuff today is just a little more efficient versions of the same thing, it's logged. This is different. This is where the actual piece of work is composed of a training dataset that may span hundreds of other works by dozens or even hundreds of other artists for an image that may not even be commercial in nature. So yeah, boy, it's, as soon as you get into that level of detail with it, it's like, you know, you, you immediately started thinking, well, you need an AI engine that that would tackle that something that can super efficiently go through textual data and summarise it correctly. You know, that's the key. It is some of the stuff with the chatGPT it's like, it needs some guardrails? You know, there's certain things that when we ask it to do it, it's either hallucinating or guessing, or there's evidence that it's not computing as much as it is. Trying to be a predictor, which is the whole point of generative AI, right. It's not actually there's no intelligence there. It's just simulated intelligence. It's prediction of words and stuff. So but an AI like engine that is hardcore data processing of all that is what would need to be needed to solve that. So yeah, this this ball of yarn is only just beginning to unravel. It's it's like, with all the problems that Ayar might arguably be solving in terms of expedience, in terms of how quickly you can get things done, et cetera. Boy, it creates a whole new set of problems to solve, you know, so yeah, it's an interesting time. I hate to keep defaulting back to that. Ricky, why don't you finish this off here? What's the final word on on all this AI related stuff? And you know, how do what do you what do you say what's the verdict?

Ricky Grove 44:43
If there's a man in the house, you can damn well be sure there's a lot of condoms too

Damien Valentine 45:00
Okay then.

Phil Rice 45:00
All right. That's it for our show today. Thanks for joining us. I'm Phil and with me has been Damian Valentine and Tracy Harwood and kind of half with us has been AI generated Ricky say goodbye, Ricky.

Ricky Grove 45:19
Oh know why they do this, but they keep sending me money through the mail. Yeah. Oh

Phil Rice 45:30
thanks for joining us. We'd love your feedback in the comments or at talk at completely machinima.com We'll see you at our next episode and have a great day. Thanks. Bye