Latest / Bad Decisions Podcast / Blender is REJECTING AI Companies, and is UNITY Winning the Race?
Transcript
- 0:00If you've ever wanted to make video games, there has never
- 0:03been a better time as Unity launches Unity AI for everyone.
- 0:08At the same time, Blender is rejecting AI sponsors for the
- 0:13next year. Making motion graphics now is
- 0:16easier than ever, and we're going to talk about why.
- 0:19Up until now you could only use cloud proprietary models within
- 0:23the desktop app, things like Opus, sonnet.
- 0:27But now you can use open source models inside your cloud desktop
- 0:31app. What is up, beautiful people?
- 0:33Welcome back everybody. It is Tuesday.
- 0:36That means we got another show for you right here talking about
- 0:39the latest and greatest in the world of tech.
- 0:42We come back to you every Tuesday and Friday.
- 0:44Remember that and. Are we ready for the show today,
- 0:47Farrahs? I don't think you are.
- 0:49Why not? Because if you could do me a
- 0:52favor and rotate your head 90° for us, no.
- 0:57In the morning today, Farrah was shaving his beard and I heard a
- 1:00little little moan from the battle.
- 1:05Moan, that was the scream. I was like, oh.
- 1:08No, that's not. No, you didn't do that.
- 1:09I did that you. Didn't even hear you.
- 1:11Were in the other room that was. The song I heard and then I went
- 1:15in to be surprised by the fact that she does not have sunburns
- 1:20anymore. No.
- 1:21What? Sunburns.
- 1:22Sun sunburns. Of course I don't have sunburns.
- 1:26You know that all these shavers, they have a comb.
- 1:28Do me a favor, rotate your head and then show us your side
- 1:31hands. No.
- 1:33What about the other side? It's the same so when you match
- 1:35them, when you're to match them. When you ruined one side, you
- 1:38had to. Yeah, I had to because, you
- 1:39know, there was a comb always on top of the shaver.
- 1:42I thought it was clicked in. Yeah, it was not.
- 1:44It wasn't. It was So when I was like, oops,
- 1:47everyone's gone. So that's when I heard the moan.
- 1:50Yes. No, not the moan, the scream.
- 1:54Can you show us anger? No.
- 1:55One more time, please. One more time.
- 1:56No, you can ask A. Question during the show.
- 1:58I will look at that. It's amazing.
- 2:01But I will look at the camera. No, but you won't see it.
- 2:03I'll try to do minimal movement. Great sideburns.
- 2:06Thank you. I I do like that.
- 2:08Thank you. How was your weekend guys?
- 2:10Let us send the comment. We had a good time.
- 2:12I I was sick if you guys remember.
- 2:13We had a good time. We had a good time.
- 2:16We had a good time, I was working the entire time.
- 2:17Yes, I mean, how do we know you were not in the studio?
- 2:21I was not. I was absent.
- 2:22I was working from home. And you haven't reported
- 2:24anything. You just said I've done a lot of
- 2:26things. I have done a lot of.
- 2:28Things. We didn't see anything.
- 2:29I have done a lot of Pratt. He didn't see anything.
- 2:31Hello. I didn't see anything.
- 2:32What do you mean man? OK, you know.
- 2:34What? We were here the whole time.
- 2:35I was. Working the entire time OK.
- 2:37What did he work on? I was working on our AI
- 2:39education program. OK, what else?
- 2:42That's it. I was also working on coding and
- 2:44application. What is the application called
- 2:46Styla? Where are you?
- 2:47Coding it or was it AI? Fashion try on Virtual fashion
- 2:51try on with AI. Do we get a demo?
- 2:54Sure. Next episode.
- 2:56Next episode I. Can do a demo.
- 2:58It's a revolutionary product, by the way, and it will dominate
- 3:03the fashion industry if I launch it.
- 3:05That's the. If I launch it.
- 3:06But the MVP is working and I did the MVP in one day with.
- 3:11Kodak, you keep talking about it.
- 3:12Show it. OK, I will show.
- 3:14It to you guys next episode. So shall we begin?
- 3:17Let's go. All right, first topic of the
- 3:20day is made for people who love making video games, who've
- 3:26always had the dream of making video games but never had the
- 3:29chance. Why?
- 3:30Well, if you ever dreamt of making your own video game but
- 3:34you never had the time, the money or the resources and the
- 3:38team, well now you can because Unity, the largest game
- 3:44development software in the world, has launched Unity AI
- 3:49publicly for everyone. You may be asking what this is.
- 3:53Before I describe it to you, I'm going to show you guys a video
- 3:55which will clearly show you what it is capable of.
- 3:58So let's take a look. Can you play it from the
- 4:02beginning with audio? Process.
- 4:03Can Unity AI actually build a playable prototype in minutes?
- 4:07Let's find out. I gave it my game design
- 4:10document and in the plan mode asked AI assistant to set up a
- 4:14third person controller and a fellow camera.
- 4:16With agent mode active, I can approve the plan and Unity AI
- 4:21actually adds game objects to my scene.
- 4:24In a few minutes. We have a moving cylinder.
- 4:26Next, I prompted to generate a drivable car and after a few
- 4:30iterations I have a vehicle that I can enter and drive.
- 4:34I even gave it an image reference of a town and it
- 4:37converted the layout into a playable scene using primitives.
- 4:41This is where I take over. I swapped primitives for 3D
- 4:45assets, added different effects, and focus on game design.
- 4:49Unity AI's context awareness helped me integrate animations
- 4:53and other elements with the scripts to speed up the process.
- 4:57The Unity Open Beta is out now. Try it out the link.
- 5:01Is in the description disclaimer this is a sponsored video but
- 5:08nonetheless it does a great job at demonstrating what the AI
- 5:13model can be capable of if used the right way now.
- 5:18If you still don't understand what this does, let me quickly
- 5:20explain to you pride. We're going to play this.
- 5:22Actually, you can play it in the background is fine, but Unity AI
- 5:26is an intelligent built in AI model that acts as a developer
- 5:32and creative assistant within the software.
- 5:35It lives inside the project with full context of how Unity works
- 5:39and what your project is about, and can help you create game
- 5:44assets such as 3D models and animations, can help build
- 5:47shaders and materials, write code and scripts for gameplay
- 5:53functionality, and can help you debug your problems, which is
- 5:57essentially all you're doing when you're working on a
- 6:00project. Now, why this is so interesting
- 6:03is because if you pay close attention Farah to the launch
- 6:06video itself, this is exactly what an indie developer or indie
- 6:13game studio or an individual creator needs.
- 6:17You know, you just always need a little bit more time, a little
- 6:21bit more money or just a few more team members to help you
- 6:24bring your vision to life. And that was never possible
- 6:27before. But now that you have something
- 6:30like Unity AI within the software that is built by them,
- 6:35which is the most important part, this is not clawed or
- 6:38ChatGPT codecs coming in and taking control.
- 6:41This is Unity's own AI tool. So they know basically all the
- 6:46things about the software. Exactly.
- 6:48And it's only going to get better.
- 6:50And when I look at this, what I see is the potential.
- 6:54Do you know how many people have ideas for video games?
- 6:58Just never had the chance? You know that simply by looking
- 7:02at Roblox and Fortnite, because they simplified the idea of game
- 7:08creation. You can see hundreds of
- 7:10thousands of games being created by 10 year olds, 10 year olds
- 7:15who simply just want to have a game play with their friends.
- 7:18Up until now, they had to go to Roblox or Minecraft or Fortnite
- 7:23UEFN and make their own games. But they were limited to that
- 7:27very world. They were limited to the
- 7:29graphics set by the developers. So what if we could open that up
- 7:32completely now? This is exactly what they did.
- 7:35Now you can make anything and the moment you have the idea, it
- 7:39will reduce all the friction that you had.
- 7:42Or I have to go and learn animation, I have to go and
- 7:44learn game mechanics or I don't know how to use how to make a
- 7:47first single player game. So it really reduces the
- 7:51friction. And I was thinking why Unity is
- 7:55ahead of Unreal Engine in bringing these kind of tools
- 7:58into the software actually, is it the first 3D software offer
- 8:01that bringing Maxon has it right?
- 8:03Maxon. Yeah, Maxon is.
- 8:04Officially. Fully integrated?
- 8:06I think they are the first ones, right?
- 8:08Autodesk has a department. So like bond dynamics?
- 8:12Yes. But it's not fully integrated
- 8:15into. It's not inside Maya, yeah.
- 8:17This is fully integration. And I think that's really
- 8:20powerful, yeah. That's why it's powerful.
- 8:23The reason why I wanted to talk about this is because what is
- 8:27the objective here? The objective is to lower the
- 8:31barrier for individual developers and small teams to be
- 8:34able to create games and also help you prototype new ideas and
- 8:39speed up production. If you just look at what is
- 8:43happening in this video or the previous video we showed you,
- 8:46the idea of making a game starts by you having the thoughts in
- 8:51your head, storyboarding it, of course.
- 8:53But once you have that, the first thing you have to do is
- 8:55bring in fake assets or bought assets or primitive assets, and
- 9:00then enable functionalities by writing code.
- 9:03So you start adding in characters, cameras and
- 9:05everything. And that part is almost
- 9:08universal to any game that you might make, right?
- 9:10You need a camera system you. Need a.
- 9:12Camera there. All of that was time that you
- 9:14had to take to do yourself, or you had to go and buy templates
- 9:18that were not always specific to your project.
- 9:21There were general templates, but right now the AI can help do
- 9:24all of that work for you, especially if you're not great
- 9:27at coding. The reason why I think this is
- 9:30very important for me, I'm a 3D artist who can create very
- 9:34cinematic scenes. We've taught hundreds of
- 9:37thousands of people, millions. If you consider YouTube how to
- 9:40use Unreal Engine in terms of creating cinematic scenes, I'm
- 9:43all set. I know how to create my 3D
- 9:45models, texture them, materials, lighting, all that, but I never
- 9:50felt the urge to make a video game.
- 9:53I've done it before for clients, but on my personal time as a
- 9:57gamer, by the way, I never felt the desire to go out and make a
- 10:00game because. Because of the mechanics.
- 10:02All of the amount of time it takes to go and learn the
- 10:05languages you need to create games.
- 10:07And then the biggest question, do I learn Python or C and all
- 10:10that? But now if I can have a
- 10:14developer inside Unity, I can work on the assets, I can create
- 10:19the lighting the way I want it, I can create the story the way I
- 10:22want it, and then all of the gameplay functionality, it can
- 10:26be added by my developer assistant.
- 10:29Then that means I can create a game out of any scene that I may
- 10:33have created. I can create a game anytime I
- 10:36have an idea. I can prototype the idea really
- 10:38quickly. I'll give you an idea.
- 10:41You just asked me what have I been working on on the weekend.
- 10:43Yeah, I just said, I literally created a fashion virtual try on
- 10:47app so that anyone can try on codes within seconds and see how
- 10:50it looks on them using AI, and I coded it in a day.
- 10:55The MVP is already there. My mom used it, she loved it.
- 10:58The fact that I made that MVP in a day is the power.
- 11:02So many developers have ideas. They just don't have the money,
- 11:05the time, the resources to prove to other people, hey, this game
- 11:08idea will work. But with Unity AI, if they can
- 11:12prototype the game within three days, sure, they can hire more
- 11:17people, get everyone involved for the final production.
- 11:20But the prototype, That's where every project fails.
- 11:22Because if you can't share and convey your vision, then you're
- 11:26never going to get funding, you're not never going to get
- 11:28investment, or you yourself will never have enough money to
- 11:31sustain. What will happen when now
- 11:36everybody is able to make these games?
- 11:39We. Can take a look at Roblox and
- 11:43Fortnite as an example. You'll have a plethora, a river
- 11:48flowing with games, 80% of which are.
- 11:50Garbage, garbage and only 1% are the 1000. 90% of them are
- 11:54garbage, 10% of them are great. They are sometimes more
- 11:58engaging. Than the official ones.
- 12:00Made by the developers because 10% of the community really
- 12:04knows how to create engaging experiences, and that's true
- 12:09with Fortnite. Yeah, we have the maps.
- 12:11Here with Roblox. Do we see pride?
- 12:12The number of concurrent? Yeah, you see now 4.8 K, Of
- 12:17course we have Star Wars at the top.
- 12:19Yeah, like these are by Epic, but if you go to Epic picks,
- 12:23these are user, user. Creative games 200.
- 12:28Usually the more stupid they are, the higher the number of
- 12:31followers. Yeah.
- 12:34Box PVP15K concurrent players. Wow.
- 12:38May I just add something for anyone who's followed us for a
- 12:42while now knows that we've been in the Unreal Engine world for
- 12:45the longest time. We've supported Unreal Engine
- 12:47since the beginning of this channel and we believe in the
- 12:50product. And there was never a time where
- 12:54I felt the urge to try Unity. Never.
- 13:00This is the very first time I'm feeling a strong urge and desire
- 13:05not just to try Unity, but also try to make a game.
- 13:09But why you want to make a game? First of all, I'm a gamer.
- 13:12I used to play video. Games.
- 13:13You like to make a game? Yeah.
- 13:15No, I used to play video games all the time and I love 3D.
- 13:20I've created cinematic scenes for the past six years, right?
- 13:24But I've never connected the two.
- 13:26I've never created. Well, I've done it for clients.
- 13:29We've created Fortnite experiences for clients, but
- 13:31I've never done it for myself. Like if I had an idea and I
- 13:34want. Like how would it look if you,
- 13:36you know, actually code it and? When you think about games,
- 13:38don't think about God of War and Horizon 0 down.
- 13:40You don't have to go that far when you're thinking about you.
- 13:43Know what your idea is? You know, Snake on Nokia?
- 13:46I think that kind of games you want to make, you're making.
- 13:48Fun of it, You're making fun of it.
- 13:49But no, really, that's true. Think about.
- 13:51Limbo. Can you pull up Limbo?
- 13:53Wow, OK. For Limbo is an incredibly
- 13:58simple game. It became so popular it was
- 14:02released across almost every device you can think of from
- 14:06Sony PlayStation to PC to Mac to iPads and iPhone.
- 14:11It's a silhouette style of game like the art style is relative.
- 14:16It's very cinematic, but it's relatively basic.
- 14:18There's no colors almost, and it's just this character in a
- 14:23puzzle game. Side scroller, kind of.
- 14:27Do you call this? It's a side scroller in a way,
- 14:29right? Yeah.
- 14:30It's not really a 3D experience. That game is purely just the
- 14:35idea and the guy truly spending time in figuring out a way to
- 14:38create an engaging experience. So many people have temple run.
- 14:42Temple run. Great idea.
- 14:43Was famous for I think 2 years swiping right left.
- 14:48Up and down, That's it, right? And it was the most addictive
- 14:50game. What I'm saying is there are
- 14:53many people in the world who have ideas.
- 14:56Kids. There are many creative kids who
- 14:58have ideas on what they want. They just don't have the skills
- 15:02to go and develop the game and create the access.
- 15:05Easier and easier. What this AI suite can help you
- 15:09with is fill in your gaps. I'm a 3D artist.
- 15:13I'm not very technical. I'm very cinematic in my output,
- 15:17but I can get technical help from the AI suite.
- 15:20I can help. I can get help on the
- 15:24development of the functionalities of the game.
- 15:26I want to create the character look myself.
- 15:28I want to do the lighting myself.
- 15:30But the game mechanics, the game mechanics, but.
- 15:32To connect the character so to set up the 3D camera, to follow
- 15:35the character, to set up specific animations for
- 15:38different interactions. All of that I don't want to
- 15:40spend time to make. I can have the AI suite help me
- 15:43out. Vice versa.
- 15:45You can have developers who love to code and they want to get
- 15:49really nitty gritty with the detail of the code they just
- 15:52don't know how to create. They can use Trellis and Tripod
- 15:55to create the 3D assets, and if you're making stylized games,
- 15:59honestly, I generally believe Tripod are very well capable of
- 16:02making that now, especially for prototypes.
- 16:06You can also buy your assets and then use AI to help with
- 16:09everything else. It is.
- 16:11It works because now you have all the different group of
- 16:16artists. Whether you are a texture
- 16:18artist, animator, 3D artist or even if you are a 2D artist, you
- 16:21can come now and make your game. Yes.
- 16:24Democratize for everybody. For everyone, I will add we have
- 16:29not yet tested this ourselves, so if you do go ahead and test
- 16:34these same prompts and they don't work out the way they are
- 16:37intended to, don't come and blame us.
- 16:39I'm not telling you, I've tested it and it works perfectly.
- 16:42I can almost guarantee it won't work half the time because
- 16:45that's how AI works. At least today.
- 16:47Even when you go to ChatGPT, you ask a question, you won't get
- 16:50the right answer all the time. But if you ask a developer also,
- 16:54sometimes you won't get what you want the first time.
- 16:56Same with the same signer. What I recommend you do is if
- 17:01you're interested in 3D and game creation or experiences, again,
- 17:07please do not only think in terms of video games as a game
- 17:11you play on the PS-5. We have talked and had meetings
- 17:15with thousands of clients within the this lifetime of bad
- 17:20decisions. You have no idea how many times
- 17:22we've been contacted by museums and art clubs to create
- 17:27experiences for them that are interactive for their shows.
- 17:30You remember the Museum of Egypt had an experience.
- 17:32That he. Wanted us to work on.
- 17:34No museums, art galleries, exhibitions, exhibitions,
- 17:39sports, you know, racing leagues, all those different
- 17:42categories of services that require immersive experience.
- 17:46Because they have thousands of audiences.
- 17:49They want they want to fan. They want the fans to be
- 17:52engaged. Thank you.
- 17:53And the only way to create engagement is to create
- 17:56interactive experiences. Prior to the year 2000 it was
- 18:01all physical installations with physical experiences.
- 18:04But right now a lot of the physical experiences are using
- 18:07digital enhancements to make them cool.
- 18:10This can be incredibly helpful if you're a 3D artist and can
- 18:15only create characters or 3D model and texture and animate
- 18:19but don't know how to code. You can use this to now create
- 18:23interactive experiences using Unity.
- 18:25Think about it that way, your characters can talk when people
- 18:28interact with them. Before you need to learn how to
- 18:31code. Now you don't have to.
- 18:32You can create that experience and work with a client you could
- 18:35never with work with before. So how long?
- 18:38How long for Unreal to bring this?
- 18:40Let's bet. I think when we talk about our
- 18:44second topic which is Blender. No, no, I'm not talking about
- 18:47that. But no, no how long you think it
- 18:49will take for Unreal Engine to deliver something like this next
- 18:55Unreal Fest June? One year.
- 18:58One more year. One more year?
- 19:00No. I'll explain why.
- 19:02From what I see, unit's CEO is heads down working.
- 19:07Public company. In to bring in AI into the tool
- 19:11set of Unity. What I see on the other hand
- 19:14with Tim Sweeney, most of his posts are about the fight
- 19:18against Apple because of the the fees that Apple is charging and
- 19:23monopolizing. But that case stop, You know
- 19:25that right? They are not going but.
- 19:27But he's still not, He's still posting about it really to this
- 19:31day. Yes, I saw another one last
- 19:32week. There's no problem with that.
- 19:34I just feel like the focus is elsewhere and fortunately or
- 19:41unfortunately I feel like Unreal is not putting a massive focus
- 19:48on AI at the moment, if you compare it to Unity at least.
- 19:52So you're saying more than one year?
- 19:54If you, if I would reframe the question in the AI race, I think
- 19:59Unity is ahead and it's very important to be competitive in
- 20:03this space considering how fast it's moving and how much it
- 20:07enables the new generation. I would love to move to the next
- 20:12topic and talk about it there, But generally speaking, you
- 20:15shouldn't just consider the current population of artists
- 20:18and developers using these tools.
- 20:20Don't just look at keeping people happy.
- 20:24The ones who are currently using Unity and Unreal, they have to
- 20:26think about the future. The new generation.
- 20:28The next generation who is currently in school and about to
- 20:31finish college and university and Start learning a new tool,
- 20:34which one will they choose? Are they going to choose
- 20:36Blender? Are they going to choose Unity?
- 20:39Are they going to choose Unreal? That is what is going to make
- 20:41the difference. It's not me, it's not Farrat,
- 20:44it's not Pratt, it's not you guys watching right now.
- 20:47It's the next generation because those are the millions of people
- 20:50who are going to be coming in and spending money and use these
- 20:53tools to create content around them.
- 20:55And so the question is, who's pushing innovation faster than
- 21:00the other ones? The reason they use Unreal up
- 21:02until now is because Unreal has been that one.
- 21:05Yeah. And they still do.
- 21:06It's not that they don't. It's just right now Unity with
- 21:10AI makes the most noise. Yeah, makes sense for the next
- 21:14topic. I think it's quite related
- 21:16actually. We can easily, yeah, seamless
- 21:19transition. Blender is rejecting AI sponsors
- 21:23for the next year after they got Anthropic as their patron
- 21:28sponsor in a few days ago. There was a huge backlash on X,
- 21:32people talking about how could Blender accept this donation and
- 21:36how could they accept the donation from an AI company.
- 21:40There were so many talks on XA lot of people were for it, a lot
- 21:43of people were against it. So Blender on their official
- 21:47website announced that recently Anthropic was joining the
- 21:52Blender Development Fund as a corporate patron member.
- 21:56Since then 2. 100. €40,000 per year and the way it works is
- 22:01they have to pay that money every single year and that money
- 22:04of course is used for Blender development.
- 22:07Since then we've heard concern from the community around how we
- 22:11engage with any AI company. We should have opened up more
- 22:15conversation and perspective from contributors before making
- 22:18this decision and for that we are sorry.
- 22:20So they took a step back and they mentioned Anthropic would
- 22:25be a one time donor, so they're not going to renew their
- 22:28donation for next year. A lot of people are angry, but
- 22:32why they did that? Why they are stopping the money
- 22:35to come to Blender? Because this money is used to
- 22:38develop more features to upgrade the software.
- 22:42And to pay for the developers who are working in.
- 22:44Blender on Blender Prada I want to show Andrew Price's tweet.
- 22:50He also shared the video of Tom. If you guys don't know, Tom
- 22:53Rosendalt, founder of Blender, Who stepped down as the position
- 22:58of CEO, I think a few months ago, two or three months ago.
- 23:00Let's watch this and then we're going to talk about this more.
- 23:04Like what is going to happen in the 30s, right?
- 23:06How are we going to deal with all this stuff with meta
- 23:09versions and AI's? And what is going to happen with
- 23:12computing and creation in the coming decade?
- 23:15And if Blender wants to stay relevant, we have to be ready
- 23:18for that. And you can't wait with that,
- 23:21because we keep making new versions of Blender day-to-day
- 23:25work keeps growing. And if we keep doing that, in
- 23:28five years we might lose it and then become obsolete.
- 23:31So I would. Like that is the talk that he
- 23:34gave in 2023. He was saying that if we do not
- 23:38pay attention to the latest in technology, of course, he gave
- 23:40the example of metaverse and AI, Blender would be obsolete
- 23:44because you can see now AI is coming into every single tool
- 23:48that we have. We have covered in the previous
- 23:50session, it's in After Effects, it's in Ableton.
- 23:53Now you can connect it to your Meta account.
- 23:55When software companies are refusing to integrate with AI,
- 24:00they'll be just left behind. The people who are coming in, as
- 24:04far as was explained, the next generation are looking for the
- 24:07tools that give them the flexibility that give them the
- 24:10access to create faster, to create better.
- 24:13And if for example, a company like Blender say, oh, I don't
- 24:17want to accept donation from AI companies.
- 24:20I think they kicked out Meshy also from their meshy AI, Meshy
- 24:24AI. But another thing that doesn't
- 24:27make any sense. Pratt, can you bring up another
- 24:28link if you really want to drill down?
- 24:31I'm referring to Andrew Price's post today all the time.
- 24:34Look how many AI companies are actually there in the chart of
- 24:38people who are donating, NVIDIA making chips, AMD making chips,
- 24:42Intel, we have Meta, we have Adobe Hyper 3D, we have Google.
- 24:47All these company are either developing their own NLM or
- 24:52developing the infrastructure for language models to work.
- 24:55So NVIDIA make or AMD make the chips.
- 24:58We have to be very open about the idea that this technology is
- 25:02here and it's going to be integrated to every single thing
- 25:05that we do. If you don't like certain things
- 25:09about this technology, if you don't like AI generated video,
- 25:12that is totally fine. I think nobody's forcing that to
- 25:14anybody. You can refuse to look at these
- 25:18kind of videos. You can go and watch different
- 25:20videos, but not having those integrated into the application,
- 25:24I think that is a concerning part.
- 25:27Enterprise has written here that if they cancelled all of these
- 25:32sponsors, Blender would lose €852,000 per year in funding and
- 25:38likely need to fire one in three developers.
- 25:41People have to understand something.
- 25:44The backlash from doomers and ragers online has already caused
- 25:52the loss of the money for four developers.
- 25:56From what I read on X, the amount of money they'd be losing
- 26:00not having entropic fund them next year.
- 26:04That's the annual. Salary of four.
- 26:08Developers who would be gaining bread for themselves and their
- 26:11families. No.
- 26:12Who would be working on features?
- 26:14Before developers for a year, Blender better, yes.
- 26:18Let's get deeper into this. I started my journey in 3D
- 26:22predominantly with Blender and then I I made this switch to
- 26:26Unreal Engine, but always kept Blender in my pocket because I
- 26:29use it to create models and then take them to Unreal Engine.
- 26:33So it's always been a back and forth between the two.
- 26:36But I've mentioned this briefly before my first month started
- 26:40with Cinema 4D and I want people to understand the reason why I
- 26:45made this switch from Cinema 4D to Blender was because Blender
- 26:50was easier to use. It was free, we had a great
- 26:56community, there were lots of tutorials and incredible
- 27:00resources like Blender Guru Enterprises Tutorials.
- 27:05Why did he make tutorials? Because when he went into
- 27:08Blender, it was free, it was easy to use, and there was a
- 27:11great community. Now, because I learned 3D in
- 27:17general, I then eventually stumbled upon Unreal Engine.
- 27:22I'm I'm talking about myself, but of course everyone in bad
- 27:25decisions included. And then when I learned Unreal,
- 27:28we all decided to teach that millions of people have learned
- 27:32Unreal Engine and are now making money using the tool in
- 27:36different industries because we made those videos.
- 27:39We made those videos because we learned Blender.
- 27:41We learned Blender because Blender was easy to use.
- 27:43You guys track what I'm saying. If Blender doesn't stay up to
- 27:49date, if Blender, it eventually becomes the software that is no
- 27:54longer the easiest to use, it's not the most up to date, it's
- 28:00not using the latest tools. Eventually the Pharaohs of the
- 28:04Future and the Blender Guru and the Fathot and the Pratt, they
- 28:08are going to go to the competitor that could be Unity,
- 28:12because they have Unity AI and the next generation of college
- 28:16artists, when they leave again understands.
- 28:18I mean, they don't know normal maps, they don't understand 3D
- 28:20models. They're day one, so when they
- 28:23learn how to make a scene, that's the first thing they're
- 28:27learning. They're never going to question,
- 28:29oh, how was this made ten years ago or 20 years ago?
- 28:31They get in, they won the fastest and the easiest.
- 28:33That's what happened. To me and they will make the
- 28:34decision in a day. Thank you.
- 28:36I want to learn something. Oh, I have Premiere Pro, I have
- 28:39Da Vinci Resolve. Which one I go to and whichever
- 28:42you go to, you rarely switch. You rarely switch.
- 28:44No, but. If one of them become obsolete,
- 28:47yes. If I am day one trying to find
- 28:51the tool I want to use for my career, I'm going to find the
- 28:54fastest and usually the one that is free, the one that is
- 28:57easiest, the one that has a community and lots of resources.
- 29:00If Unity AI has the tools that makes creation of video games
- 29:05and film and video and all that easier, then the next Blender
- 29:09guru will not fucking be Blender guru, he will be Unity guru.
- 29:13The next bad decisions is going to be teaching Unity if Unity is
- 29:16the easiest one to make because people want to learn that tool.
- 29:20It's just how the economy works. Every company that I know that
- 29:24has the slightest bit of strategy and competitiveness in
- 29:27them is learning how to integrate AI into their tool
- 29:30set. Actually, is there any company
- 29:32that says if you're not going to use AI?
- 29:34I I really don't think any public company can do that.
- 29:37First of all, no public company, public company, definitely they
- 29:40cannot do that. Public company is no because the
- 29:42investors are going to be in their.
- 29:43Ass even as I would say a big corporation that is not public,
- 29:47I'm not talking about a small studio of two or three that they
- 29:50can make a decision for themselves and it's fine.
- 29:52But if your business and when I say big company, not only in
- 29:56terms of number of employees, because when you are a big
- 29:59company, you have a lot of suppliers, you have a lot of
- 30:01customers, so you have an ecosystem around you.
- 30:05If you are just a free freelancer dealing with one
- 30:07client or two client, there is no ecosystem around you, right?
- 30:11If you cease to exist, nothing is going to happen.
- 30:13If you're a big company, you have a huge list of suppliers,
- 30:16you have a huge list of clients, They all rely on you at that
- 30:19point. You cannot say I'm not going to
- 30:20use AI again. You can say I'm not going to use
- 30:24AI generated photos in my game. You can't say that that is
- 30:27totally fine. But you cannot say I'm not going
- 30:29to use the technology in my company that that's, I think
- 30:33very absurd to say it. And one more thing that it
- 30:37really caught my attention was if people have problem with
- 30:42Anthropic itself, why not the other companies?
- 30:46Then Blender will lose a million $1,000,000 a year in donation if
- 30:50they go after every single company that is, you know,
- 30:53remotely connected to. AI native companies.
- 30:56Nvidia's focusing predominantly on AI at the moment, and Meta is
- 31:00the same, so technically they should be out too.
- 31:03And that will just cause lack of resources for Blender.
- 31:08Lack of resources means lack of updates.
- 31:10Lack of updates means eventually the users of Blender will feel
- 31:14fed up and they'll go to the next tool.
- 31:17And what I'm afraid of is that the next tool may not even be
- 31:21Unity, it may be Higgs Field or Magnificent tool.
- 31:26Right when I say I'm afraid of. It's not that they're bad, it's
- 31:28just that they don't have 80 to 90% of the functionalities that
- 31:32we have inside Blender and Unreal and Unity.
- 31:35So when you go to Higgs Field, you're at the mercy of their
- 31:37website, you know, and whatever feature they want to add today.
- 31:40And they change really fast. Isn't it much better to have
- 31:43those creative tools that are AI enabled natively inside our
- 31:47softwares so that we get to stay in the creative tool that we
- 31:51are? We have familiarity, we know how
- 31:53to work around it, we have flexibility, and then we can use
- 31:57AI whenever and wherever we want.
- 31:59It's just how I think we should look at this.
- 32:03I wanna make a prediction. Please go ahead after you've
- 32:07made a look at this. You can make your prediction and
- 32:12then I will talk. About it next year, by the time
- 32:14of renewing or not renewing Anthropics donation, everybody
- 32:20has changed their mind. They're gonna become a sponsor.
- 32:23They're gonna be integrated into Blender.
- 32:25They have to like one thing that everybody needs to understand
- 32:30too is like me myself. I'm a Photoshop and After
- 32:32Effects user, right? Heavy user.
- 32:35And I'm not user. Yeah, I'm not using Photoshop
- 32:37and After Effects. Anymore ex User.
- 32:39Yeah, I just go to anything I wanna edit.
- 32:42It's pretty fast. I just type it down, go to chat
- 32:44GPD and yeah, my job is done. No, I'm telling you, by next
- 32:48year, half of the people who were against it, I think they
- 32:51will change their mind. Blender Foundation will see
- 32:55that, hey, we need that money to stay ahead.
- 32:58Anthropic come back and you know they are working hand in hand.
- 33:02We showed the demo in the previous episode, right?
- 33:05How great it is. You have a tool that helps you
- 33:07with your material. That's it.
- 33:10Let's what is wrong with that? Let's talk about the opposing
- 33:13argument for a moment. I think it's fair.
- 33:16Because what is the opposing? I've got a few comments under
- 33:19our points. May I go ahead?
- 33:21I just want to see the comments on the let.
- 33:22Me explain really quickly the opposing argument yes.
- 33:25The opposing argument is that once these AI tools are natively
- 33:32inside the application, especially when they come from
- 33:34the outside, so if it's Claude or ChatGPT, they will learn
- 33:38everything about your scene. They're gonna use your art for
- 33:42training so that in the future they can charge you for the art
- 33:47that you gave them for free. The argument is that they've
- 33:51already trained everything on your data and they're going to
- 33:54continue to do so now that the AI can read your scenes and
- 33:57understand it. So let's say you're working on a
- 33:59project with a specific art style that you developed.
- 34:02It's going to read that, it's going to understand it.
- 34:04And next year, another person can use the art style that you
- 34:07made and 10 other people can do the same thing.
- 34:11This is perhaps the biggest and most understandable argument
- 34:16that I found against the use of AI within the 3D softwares.
- 34:20I wanted to share that with you. I want to get your thoughts.
- 34:22I have what it. Has to do what it has to do with
- 34:24the donation I OK sure don't bring.
- 34:27No. The donation states that Blender
- 34:31has to dedicate a developer and a member of their team to work
- 34:36directly with Anthropic, which means they will be developing
- 34:38tools to help Anthropic and that means that they will be helping
- 34:42them receive data from Blender. All those things you can fill in
- 34:46the blanks. But can I say that sure, OK,
- 34:49it's a valid argument, but if Autodesk, Maxon, Unreal Engine,
- 34:55Premiere Pro, Photoshop, every single one of them are giving
- 34:59data, I don't think the whole AI companies are waiting for
- 35:02Blender artist to give their. You see what I mean?
- 35:05They're not waiting. It will surely help.
- 35:06They're not waiting. That's where I come in.
- 35:09I have an idea about this. We've talked about this before.
- 35:14If we want to have this discussion, we all have to act
- 35:17as adults and grown-ups and understand one thing and accept
- 35:21it as truth. The general consensus about AI
- 35:27is true. Everyone keeps talking about how
- 35:29AI is trained. Is it ethical?
- 35:30Is it not? That's a very long and hard
- 35:34discussion. But the truth is, if you just
- 35:36look at it scientifically, AI and machine learning, they work
- 35:39by taking an input data and training themselves on the input
- 35:44data so that they can give you an output that is desirable.
- 35:48The input data has to be human content.
- 35:51It has to be the content that we've made.
- 35:53It's the text that we've written for thousands of years.
- 35:55It's the poems, it's the newspapers, it's the articles,
- 35:58scientific papers, it's the images, the videos.
- 36:01Fucking hell, every video we've uploaded on YouTube probably has
- 36:04been scraped for training. Is that cool?
- 36:08Probably not. I mean, it's not cool that
- 36:11everyone's work was taken and used to train, but it's already
- 36:15done. I will guarantee you one thing.
- 36:17Everything on the Internet has already been read, and multiple
- 36:20times, multiple times, perhaps more than five years ago.
- 36:23It's not even recent like we're using it now.
- 36:26But this has been done for the longest time.
- 36:29So yes, I understand that part is not cool.
- 36:32I get it. But I'll tell you about my
- 36:35personality, and because I'm living a pretty happy life in my
- 36:38mind, I hope and wish the same for you.
- 36:41But I think living life with misery and hate and resentment
- 36:45is hard, and I don't wish that upon anyone who's watching and
- 36:48listening. So I want to share with you my
- 36:50perspective. I always look at the glass half
- 36:52full, not the glass half empty. I know my characteristic.
- 36:57I'm never the kind of person who'll be like, this is my work.
- 37:00I'm not going to share it with you.
- 37:01I'm not going to show you because you're going to copy me.
- 37:03If that was the case, we would have never released free
- 37:05tutorials on YouTube. Literally our Star Wars series,
- 37:08when we released it, our friends were like, why are you sharing
- 37:10your secret skills that you could use with clients to
- 37:14people? I'm like, I actually look at the
- 37:16glass half full. I'm gaining millions of views
- 37:18from these videos and so much love in the comments of changing
- 37:21people's lives. That's more valuable.
- 37:23And in the long run, more people got to know about bad decisions.
- 37:26In the beginning, you could say, why did we release it for free?
- 37:29Why did we share all of our trade secrets?
- 37:32Again, you can have that mentality or just be open about
- 37:34it and actually gain an advantage here.
- 37:36For me, the glass half full is very simple.
- 37:40If it has been trained on my work and I believe my work is
- 37:42great. Sure, it was stolen.
- 37:44It was trained, it was copied without my consent.
- 37:47But now I have access to everybody else's work.
- 37:51Now I have access to code. I could never fucking code.
- 37:54I've made five applications that make my work faster and better.
- 37:58I can now use inspiration from other people's arts.
- 38:02If they can use mine, I can use theirs.
- 38:04The entire intelligence of humankind has been democratized
- 38:08for everyone to use. And I know the opposing argument
- 38:11is, Oh no, you're paying these guys.
- 38:13Sure, yes, you're paying, and probably.
- 38:15But how much? You're paying codecs and GPD,
- 38:18But how much? 20, Thirty, $40 a month?
- 38:20We spend way more on garbage that doesn't even matter.
- 38:23And this is fucking intelligence we're talking about.
- 38:26Again, I'm making applications that help us save so much time
- 38:30it's worth 20 fucking dollars. Yes, I'm paying them, but come
- 38:32on $20 it's like 2 cups of coffee. 200, you're paying 200.
- 38:37No, that's Claude and codec, because we.
- 38:39Codecs also, you're paying 200. You're ruining my argument.
- 38:42Finance guy, finance guy. But you're using 200.
- 38:44I'm saying I'm paying the 200. Thank you for.
- 38:46Shitting on my argument, we are using the most highest tier of
- 38:51the plan because there's three of us.
- 38:53Yeah, using the two things. The platform, yeah, you've got
- 38:58to be transparent about the whole thing if you're going to
- 39:00say that you got three people rape clod code here on the show,
- 39:07because we're all coding at the same.
- 39:08Time, no. And then we are developing
- 39:10another platform that takes a lot of tokens.
- 39:13But that's why we could be transparent about the whole
- 39:15thing, OK? It's like we're making five
- 39:18applications at the same time. Three of us.
- 39:21The point I want to make is simple.
- 39:23I believe that there's two people that is watching this
- 39:29show, creators or doomers. You're either going to be a
- 39:35person who's going to use these tools to create.
- 39:37You're going to ameliorate your life, You're going to live a
- 39:40better quality of life because you're working faster, you're
- 39:43working more efficiently, you're excited, You keep using and
- 39:46experiencing with these tools, you create things that you could
- 39:50never create before. Or we're going to be a doomer,
- 39:54we're going to be resentful. You're going to keep looking at
- 39:57these tools and say, oh, it's taking my job, it's copying my
- 39:59shit. Yes, that population exists and
- 40:02I could do the same. But I don't think that's fun and
- 40:06I don't think it's helpful to you, to me or anybody else.
- 40:11So. I want to I want to read a tweet
- 40:14by Blender Bob. Yes, he mentioned.
- 40:16I've made many applications, many plugins using codecs and
- 40:22GPT, over 40 / 40 plugins. I gave it for free to the
- 40:26community, Everybody used it. Nobody complained that it was AI
- 40:31because I gave it for free to the community.
- 40:33Now that the AI companies come and donate to Blender, why
- 40:38you're angry about that? No, he actually people were
- 40:41angry at Blender Bob. For support, of course.
- 40:43Yes, yes. And then he said, guys, you
- 40:45don't understand, same people who are using my plugins.
- 40:48I made it with AI. Why didn't you complain?
- 40:50Yeah, not complaining to. Me.
- 40:52Because they were not just complaining to Blender, they
- 40:53were complaining. To him and he was saying that
- 40:55how he could use codecs to make these plugins because first he
- 40:59said I'm 50 something I'm not going to learn coding.
- 41:02Second, I'm not going to hire a developer to make a free plugin,
- 41:05but now with these tools, I could make plug insurance.
- 41:08This wasn't possible 10 years ago, five years ago.
- 41:12There are a lot of benefits that we are getting from this
- 41:15technology that we don't even realize it in our daily life
- 41:18nowadays. One example and there are so
- 41:21many more. I'm going to use another
- 41:23example, I'm just thinking about it right now.
- 41:24Perhaps it could register better.
- 41:28I work with Farhad and spend a lot of time with Farhad.
- 41:32If I have an argument with him, I can let that last for a couple
- 41:37of hours, but it cannot go past a day.
- 41:41I know brothers don't talk for weeks or months, sometimes
- 41:43never. We cannot let an argument last
- 41:46longer than a day because we are starting again the next day.
- 41:49We have more work, we have more projects.
- 41:51It requires us to come and record.
- 41:53We cannot be in an I'll mood against each other and record.
- 41:58We, we have to finish arguments earlier because it just, it
- 42:01doesn't work. It we have too much exposure to
- 42:04each other. It's not healthy, it's not
- 42:07enjoyable, it's not joyful. And that's not the kind of life
- 42:10I want to live. So anytime we have any sort of
- 42:12arguments, usually about work, opposing ideas, it's healthy.
- 42:17But there are arguments, right? Never fights, just arguments.
- 42:20And we always end it within a few hours maximum.
- 42:24The same is true with AI. There's so much exposure to AI.
- 42:29AI is like your brother at the moment, because I'm thinking
- 42:32about it. Every company, every tool, every
- 42:34person around you is probably already using AI.
- 42:38And if you want to constantly fight it and have tension, it's
- 42:41just going to make living life impossible.
- 42:43You have to let go and try to understand how to work with AI.
- 42:47Try to understand how to work with your brother.
- 42:49You just compared me to a higher intelligence.
- 42:51That's it. Thank you very much.
- 42:53But dude. Doesn't promote this doesn't
- 42:55make sense because I'm exposed. To you every day.
- 42:58Yeah. If you want to be better, if you
- 42:59want to make more money, if you want to have more clients, you
- 43:02have to work together. Yeah, we can say that we don't
- 43:04want to work together. Pratt, we are not going to work
- 43:07with you anymore. I want to go do something else.
- 43:09Yes, you can do that, but it would be slower.
- 43:11Yeah, you will be slower. And then you're not gonna,
- 43:13you're gonna make something great again.
- 43:15The whole purpose of the show is why we started to tell you guys
- 43:19what is good, what you can use, what is actually working.
- 43:23The reason we are also researching all of this is
- 43:26because we want to be better. And then we transfer it to you
- 43:28guys. So there is no point fighting it
- 43:32at this moment. No.
- 43:34We should have debates and there are people who will do things
- 43:40that are misguided and potentially evil at times.
- 43:44There are companies who do that. I'm not denying it.
- 43:46No. And we also have a lot of
- 43:47thoughts about creation, about art.
- 43:49How many times we sat down and thought, Oh my God, are we gonna
- 43:53have another Lord of the Rings? Are we going to have another
- 43:56matrix? Are we?
- 43:57Going to spend the time, yes. To create because now we have
- 44:00all these tools and it's sometimes it's sad.
- 44:02Or are we going to have? A music composed for film.
- 44:06Can you expand that entire discussion?
- 44:08Because it doesn't make sense. You just say it in a line.
- 44:11People need to know the kind of discussions we also have off
- 44:14camera. So it's not just always, Oh yes,
- 44:17let's go AI, let's go. One of my biggest concern for
- 44:20the past few months is thinking about the concept of creating
- 44:25great art and the concept of fandom to that art.
- 44:30Somebody creates a piece, whether it's a movie or it's a
- 44:33piece of music, people look at it.
- 44:36People sacrifice something to see it.
- 44:38That sacrifice could be money or could be time.
- 44:42I wait or I travel a long time to see that art, or I spend
- 44:46money to see that art. That art is so valuable, whether
- 44:49it's a movie that I want to watch in IMAX, whether it's a
- 44:52concert that I go pay or I stand in a queue to see.
- 44:55Creating great work previously was very evident to people.
- 45:01You could say that oh, Lord of the Rings was a great a movie
- 45:04that was made or Star Wars soundtrack.
- 45:07The score still to this day holds.
- 45:09It's an orchestra. It's amazing.
- 45:11People really want to listen to it.
- 45:13With the advent of these tools that now everybody has access to
- 45:17it. Do we really have fandom or can
- 45:20we really create amazing work that if you look at it 10 years
- 45:24from now, 15 years from now, 20 years from now, you look at that
- 45:28AI generated video or film or music and say, wow, the person
- 45:32who has created this must be through a lot of pain and
- 45:37pressure to make this work. We always thought about it, is
- 45:42that going to happen in the future?
- 45:43And sometimes, oh, it's never going to happen.
- 45:45So I say no, it might happen. We might see an AI movie and I
- 45:48think you watched one a few days ago.
- 45:50You told me. Oh no, this.
- 45:53Runway released the video. Pratt, can you pull that up?
- 45:56Don't play it now. Just have it ready.
- 45:58Runway. It's the black and white video.
- 46:00The guy using. It the watch to change time.
- 46:03I'll I'll tell you when to play it in in a moment.
- 46:05Thank you, Pratt. So we have also some
- 46:08contradicting ideas about AI and about this technology.
- 46:12But as far as said, we want to make it big.
- 46:15We want to really deliver the best to our clients and to our
- 46:18audience. So we have to use this.
- 46:20Tool can I come in beautiful by the way, what you said the
- 46:24reason why we're even saying this is because you have to
- 46:27understand me and Farrad are huge Star Wars fans, which is
- 46:32why our first Unreal Engine program was based on Star Wars
- 46:37massive Lord of the Rings Harry Potter fans.
- 46:41We devoured these films dozens of times through our childhood
- 46:47and as Farrad was discussing with me the concept of fandom,
- 46:51why were we such huge fans of these films and why are we not
- 46:54fans of anything right now? Any films?
- 46:57Why can't we feel that sense of attachment anymore?
- 47:01We remember because when we were back home, our mother was
- 47:05travelling to Dubai. We were very young and Farah
- 47:08specifically requested that she buy lightsabers, find them and
- 47:12buy them because we couldn't find them at.
- 47:13Home. That's 20 years ago, by the way.
- 47:15And our mom had to go and look for an entire day at different.
- 47:20Toy. Shops and stores to find and
- 47:22locate lightsabers. And she finally did.
- 47:24So she spent a lot of effort to find him and we played with
- 47:27those lightsabers. Like I had bruises from him
- 47:29because he would hit each other so hard for a year, 2 years
- 47:32until he broke. Why were we such huge fans?
- 47:35And so we started doing some research and it has to do with
- 47:40the amount of energy spent on the project.
- 47:45I don't know how much you guys believe in energies and all
- 47:48that, but when we think about the feelings and emotions of the
- 47:53artists, whether that's the director or the producer or the
- 47:57writer scene or the makeup artist, the.
- 48:00VFX. Artist bro, anyone who has
- 48:02touched that film in any way shape or form, they have
- 48:08transferred their energies into it and the best example I can
- 48:12give you is from Darth Maul in the 4th or he's in the 4th.
- 48:18One first prequel, 4th actual one timeline based.
- 48:23Darth Maul the villain has how much of a screen time?
- 48:2811 minutes. 11 minutes total screen time in the entire film.
- 48:32And that's the first Star Wars film being made ever, by the
- 48:35way, so they're not even sure how well this.
- 48:37Is going no, no, no, no. That's the 4th.
- 48:38One, that's the 4th one. Sorry my bad.
- 48:40That's the 4th 111 minutes of screen time.
- 48:42How much did he practice for the sword fight?
- 48:4520 hours every single day for the last three months of the
- 48:49shoot. 20 hours out of 24 hours every single day in the studio,
- 48:56three months out of the shoot. It's mind blowing.
- 48:59When I told Pharaohs these stats.
- 49:01So I, I asked you to bring it up because that level of energy
- 49:07that he put into the craft, if you believe in science as well,
- 49:12energies, frequencies, right? He's literally transferred all
- 49:15of that energy, the sweat, the pain, the suffering of staying
- 49:18up sleeping for four hours, but finally mastering sword fights.
- 49:23And then he brought that in front of the camera.
- 49:25You can feel it with the action, with his opponents, Qui Gon and
- 49:30Obi Wan. Which is why that fight scene is
- 49:32so fucking remarkable. I get goosebumps every time I
- 49:35see it. And then the guys who worked on
- 49:36the music put in the same fucking effort.
- 49:38John Williams. All of that energy gets
- 49:42channelled through when we watch it versus let's pretend someone
- 49:47just puts in a. Prompt No, no, no.
- 49:49Even the last Star Wars. Let's not even bring AI in.
- 49:52Let's just compare. That's a good sword fight.
- 49:54Pratt, can you just put pull this up, pull this up, pull this
- 49:57up, pull this up. Let people see.
- 49:58Put Darth Maad Quag on fight. We need to.
- 50:03Show you Obi Wan. Yeah, Obi Wan fight Phantom
- 50:05Menace. This is gonna blow your mind, by
- 50:09the way, if you haven't watched it.
- 50:11Yes. Are we gonna?
- 50:15Yeah. No, no, no.
- 50:17We won't get copyrighted. Come on.
- 50:20For Star Wars, no. Yeah, put.
- 50:23No, no music, don't music, no audio.
- 50:25Audio will make us copyrighted. Fast forward a bit, just for the
- 50:291st 10 seconds of the fight, they've trained.
- 50:32Yes, yes, go The choreography of this fight.
- 50:38First of all, every single actor has been trained the the actor
- 50:43for Darth Maul is an actual martial art artist who knows how
- 50:47to fight with swords. Now when you compare this to
- 50:52what they did in the 2000 Kylo Ren sword fight, Kylo Ren sword
- 50:58and what was the girl's name? Kylo Ren versus Ray versus Ray
- 51:02sword fights. Bro, it's shameful.
- 51:05It's shameful. The one near the water, The
- 51:08third? Yes, that one.
- 51:09Try. Holding the whole time holding.
- 51:18This looks like sea dance. Pushing down and conversation,
- 51:25Conversation, we talk. Jumping and then that's it. 20
- 51:31years later. This is, yeah, 20 years.
- 51:33Later, 20 years later and a lot of budget I imagine.
- 51:37No, the budget is on water simulation.
- 51:41Yeah, so literally just use that, No.
- 51:43Sword fight. Actually there is no sword
- 51:45fight. OK, pause it now.
- 51:48Again, if I wanted to compare 2 live action.
- 51:50Can I just say something before you even say that?
- 51:53Have you seen anyone talk about this sword fight?
- 51:57Like Oh my God, have you guys watched the Kylo Ren sword
- 52:00fight? It was the best part of the film
- 52:03and they keep talking about it for weeks to come.
- 52:05Type Battle of Heroes, since we are on it, if you want the fight
- 52:09that everybody talks about is this fight again.
- 52:12The training for this fight was also.
- 52:14I had all these stats. Yes, yes.
- 52:17The second one. The second one, just the start,
- 52:20just the intro. I actually tried to do it when I
- 52:24was 15 with barrels. We choreographed the first 3
- 52:28seconds. Fast forward, Fast forward a
- 52:30little bit, a little bit more. No, no, carefully.
- 52:33That's it. That's.
- 52:34It you will try. Actually, I can do the voice
- 52:37over. You're not gonna get
- 52:38copyrighted? Yes.
- 52:39Only a city deals in absolutes. If you're not with me, then
- 52:44you're my enemy. You can Fast forward a little
- 52:46bit. Press.
- 52:48Yes. Yes.
- 52:53Wow. Dude, look.
- 52:55At that, just look at this speed up bro, look at the movement.
- 52:58So I'm comparing. All choreographed but performed
- 53:02by the actors. This is not double s performing
- 53:05for them. They learn how to do this.
- 53:07When an actor learns how to do it themselves, they put in the
- 53:10time, the energy. It shows.
- 53:11That's why people love, for instance, Keanu Reeves as.
- 53:15Well exactly for John Wick and all the works that he did.
- 53:18So now we are comparing 2 lives, exactly 2 live action movies,
- 53:22different amount of efforts. When AI comes in, I'm not even
- 53:26considering it as a new technology.
- 53:28Proud copyright. Let's let's not let's not give
- 53:31them more reason. Let's not watch the whole Star
- 53:33Wars year. The revenue is shit anyway.
- 53:38I'm comparing two different films now.
- 53:40Bring AI in. I don't even care about the
- 53:42technology. If the amount of effort that you
- 53:45put is not equal to the one that you have done previously, people
- 53:50will feel it. And I'm not saying the amount of
- 53:53work equals to amount of output. It's not always like that.
- 53:57But the amount of attention, the amount of focus, the amount of
- 54:00skill required, the amount of knowledge that you need to have,
- 54:03these are really, really important before, before.
- 54:06You show that I think we all have to change our definition of
- 54:11AI slop. You have to understand for the
- 54:14AI to create something, there's a human behind it.
- 54:16So I believe you have to call it AI and human slop.
- 54:20We have human mastery, which is what we just saw with Star Wars
- 54:25initial films. And we have human slop.
- 54:27We have human. I'm not calling the latest.
- 54:29Star Wars. No, no, no it's not.
- 54:30But. We have slop.
- 54:31We have a lot of human slop. Again, Even myself in the
- 54:34beginning when I was using Blender and Unreal, I created a
- 54:36lot of fucking slob ugly shit that I would never share with
- 54:39the world. People still do that 10 years
- 54:42into their career if they don't want to improve.
- 54:44So now you have those humans who do not want to improve and they
- 54:47want to be lazy with AI. They're going to create human AI
- 54:52slop. And then there's going to
- 54:54people, there's going to be people who are going to create
- 54:56human AI mastery, human AI, phenomenal work.
- 55:01Those are the people who are going to put in the time to
- 55:03actually create something meaningful.
- 55:06This is perhaps a closer example to that.
- 55:09This is a video created by Runway.
- 55:12Runway is a platform that allows you to create AI videos, AI
- 55:17imagery very similar to Higgs Field and Freepic, which is now
- 55:21called Magnific. Runway was one of the early
- 55:23ones, by the way, before Higgs Field and.
- 55:26Runway and luma, I remember, yeah.
- 55:28This is a proof that they made proof of concept to showcase
- 55:33what AI films can be like. And what's amazing is when I saw
- 55:38this film before we even watch it on Instagram.
- 55:41I saw it when they first released it.
- 55:43Didn't have a lot of comments and likes and shares.
- 55:46I saw it, I knew it was AI because it was by runway.
- 55:49But when I watched it I was immediately hooked.
- 55:52Loved the story to the point that I was like, this video is
- 55:56going to fucking make it outside of the AI community.
- 55:59No joke. The next day, check my phone.
- 56:02You know that new feature on Instagram where it shows you
- 56:05some of your friends who've liked that same video, which is
- 56:08a pretty cool feature. I saw all of my friends who have
- 56:10no fucking clue what AI is or what runway is.
- 56:14They like that video, which means they didn't even realize
- 56:17it's AI, or even if they did, they just liked it.
- 56:19We're going to get to that point where there's going to be people
- 56:22who create great work with AI and there's going to be people
- 56:24who create slop with AI. Let's watch this video and
- 56:27you'll get you wanna watch the whole thing with the audio?
- 56:29Yeah, it's very short with the audio.
- 56:31Is it copyrighted? No, I don't think so.
- 56:35No, sorry. Very cool concept.
- 56:49Very cool concept. Yeah, the club,
- 57:07you know, I thought he's going to make himself hold there.
- 57:10He gave ideas. Imagine if this was a watch
- 57:15commercial and done the right way.
- 57:19With C dance 2.0 you can actually make it look really
- 57:22good. Most people also realize.
- 57:24Even if I don't think I could, I could.
- 57:26No, I could. If you could want, Probably you
- 57:29wouldn't know. Yeah, There wouldn't know.
- 57:31There wouldn't. No way.
- 57:32What I'm saying is again, you can create good work if you put
- 57:34in the time and you come up with the right story.
- 57:36Don't just go to chat GBDA, give me the story, go to C dance,
- 57:39paste in the prompt. That is where most people fail.
- 57:43So please, please, please understand there is human AI
- 57:47slop and this human AI great work, just like before we had
- 57:50human slop and human great work. The reason why they call it also
- 57:54slop because slop, you need to have a lot of it for it to be
- 57:57slop. AI does allow you to create
- 57:59faster, which is why we have a massive amount of human AI slop.
- 58:04And I'm not going to deny that that will exist.
- 58:06We have to do something about that, which is why we are going
- 58:10to continue our education on bad decisions.
- 58:14The reason why we're actually working on our educational
- 58:17program, which we've told you guys before, is because we want
- 58:20to take that same mentality we've had when we're working
- 58:23inside Blender and Unreal and Cinema 40 and Houdini.
- 58:27We're going to take that same mentality and bring it into
- 58:30using these AI tools. Because if you have the same
- 58:33mentality, if we do our pre production, right, our research,
- 58:36right, our storytelling right, then we can actually create
- 58:40something meaningful, whether it's for ourselves or for our
- 58:43clients, it is possible. And what we're actually doing
- 58:47with this AI program is proving that to you.
- 58:49So something like what Runway did, They created something that
- 58:52is meaningful. Many people loved it, shared it.
- 58:54I think it has maybe a million view or maybe millions, I don't
- 58:58know on Instagram, I'll check it or hundreds of thousands at
- 59:00least. We want to teach you guys how to
- 59:03create something meaningful using these tools.
- 59:05And just like the show, we're going to tell you what's hype
- 59:08and not teach you that and teach you what is actually practical.
- 59:12If you guys haven't signed up yet, the link is right here.
- 59:16AI dot baddecisions dot studio down there.
- 59:18And you're gonna go to this. Website.
- 59:20If you Scroll down, you're actually gonna see if you.
- 59:22I haven't. No.
- 59:23No. There is one.
- 59:24There is. There are some.
- 59:25You're gonna go more. No more details.
- 59:28No, not now. OK, There are some details.
- 59:30OK, Go a little bit more. A little bit more.
- 59:32Yeah. That's it.
- 59:33OK, It's fine. I have some more.
- 59:35You're gonna have it soon. Yeah.
- 59:37Let's see. Cuz we're this last weekend as I
- 59:40was sick at home, I followed and Platt said I didn't work.
- 59:44I was actually working on one of the playbooks that we have for
- 59:47you guys. I will just can I just tell them
- 59:49a little bit about it. OK.
- 59:51This playbook is so exciting. It's so big.
- 59:54We've been working on it for at least three weeks.
- 59:55Just this one playbook. This playbook is called launch a
- 59:59product. Now you might be thinking, hey,
- 1:00:01I don't fucking want to launch a product.
- 1:00:02Hey, I haven't launched a fucking product myself.
- 1:00:04It's the title of it, but what you're learning.
- 1:00:08By launching that product is literally A-Z of using every AI
- 1:00:12tool out there because you have to ideate and come up with your
- 1:00:15product using LLMS. And then you have to come up with a brand DNA
- 1:00:19and you use LLMS to do research to do that.
- 1:00:22And then you need to go ahead and create photo ads for your
- 1:00:25product, which uses a lot of image tools.
- 1:00:27And then you're going to have to create video commercials for
- 1:00:30your product, which uses all the video generation models and
- 1:00:32tools. And then eventually you're going
- 1:00:34to need to create a website for your product and then you can
- 1:00:37fill in the blank. So you're going to learn
- 1:00:38everything by launching a product.
- 1:00:40We made it so exciting. This is just one of the
- 1:00:42playbooks. We have so many of them.
- 1:00:44I'm so excited about it. Sorry, I just went.
- 1:00:46On a you tease it, but I'm going to update the website.
- 1:00:48The link is in the description files.
- 1:00:51Are we ready for the next topic? Yeah, next one is me, right?
- 1:00:53Yes, making motion graphics. I will say that again, making
- 1:00:59motion graphics is not easier than ever.
- 1:01:03Remotion has added an all new feature which is HTML in canvas.
- 1:01:09That may sound confusing to you. In simple terms, they are now
- 1:01:14enabling you to create all sorts of sophisticated effects like
- 1:01:19blurs, glitches, specific transitions that were completely
- 1:01:24impossible or incredibly difficult to create.
- 1:01:28Now things that required lots and lots of time can now be
- 1:01:30easily done using remotion. Now what does this actually
- 1:01:34mean? You can go to codecs or.
- 1:01:38There is a plug in. No there is a plug in in codecs
- 1:01:40so easy to use. You just ask it what you want
- 1:01:43and that's it. We'll give you a video MP4.
- 1:01:45If you're making a simple presentation or a video for
- 1:01:49Instagram, you can go ahead and literally tag Remotion as a plug
- 1:01:53in at Remotion and then tell it what you want.
- 1:01:56Give it some references and then it will create the motion
- 1:01:59graphics and then you can continue that discussion with
- 1:02:02remotion to create the effect that you want.
- 1:02:04And then use it on the website, use it in your videos, use it in
- 1:02:07your slides. The point is making motion
- 1:02:10graphics is not available for everyone, which is crazy because
- 1:02:13before it was kept behind a door, either you had to learn it
- 1:02:18with After Effects yourself or have someone do it.
- 1:02:21But if you couldn't do that, now you can just use this.
- 1:02:24We're going to test this. We've already tested for you
- 1:02:26guys in our previous show, Hyper Motion and Remotion.
- 1:02:30But beforehand, for all the people who care about the
- 1:02:32details, Remotion use React by enabling HTML in Canvas, they
- 1:02:37just enable you to create a lot more beautiful effects visually.
- 1:02:40So we're going to test this out. One of the playbooks I can
- 1:02:43guarantee you that we'll be doing inside the AI educational
- 1:02:47program is teaching you how to create motion graphics.
- 1:02:50You shouldn't call it motion graphics.
- 1:02:52I just realized these are great presentations.
- 1:02:55Depend, you can use it for many. Things yeah, so so even for
- 1:02:58again, I come from a corporate world background and coming to
- 1:03:03the creative world, you guys have no idea how ugly the
- 1:03:06presentations are, bro. The training I'm not OK The
- 1:03:11sales presentations are fine. Template template on canvas.
- 1:03:17The internal presentations, the reports that are not going to go
- 1:03:21outside of the company. Horrible, horrible.
- 1:03:24And now with these tools, you can just upload the excel and
- 1:03:27say give me the graphs, give me give me the charts.
- 1:03:30It would help. It will be a more pleasant
- 1:03:32experience. Sitting in those meetings.
- 1:03:34Yes, yes, thank you. OK Farah, shall we move on to
- 1:03:38the next topic please? Something that I found really,
- 1:03:41really interesting when I was scrolling Twitter was that you
- 1:03:45can now use open source models in your cloud desktop app.
- 1:03:51Let me explain. When you download cloud desktop
- 1:03:54app, you will have Cloud Chat Co work and Cloud Code 3 different
- 1:03:59tabs that you can click and you can use the models.
- 1:04:02Up until now you can only use cloud proprietary models,
- 1:04:07Sonnet, Opus and Haiku. You can switch between these
- 1:04:12models and do your work. And Opus 4.7 is the most
- 1:04:15intelligent one, but it's the most expensive one.
- 1:04:18That is why exactly we moved to the 20X Pro Max tier.
- 1:04:23Because if Pratt is working on the show, I'm working on the
- 1:04:26educational platform, Faraz is working on his own app.
- 1:04:29We are going to go out of credit very very soon.
- 1:04:32Every time we hit the limit and we have to wait for another 5
- 1:04:35hours to get there. Now with Olama you can bring
- 1:04:39open source models. And for those of you who don't
- 1:04:41know, open source models are models that are free like Quinn,
- 1:04:44like Gemma, into the cloud desktop app.
- 1:04:48So you can switch between these models.
- 1:04:51For anybody that doesn't know, let me give you a little bit of
- 1:04:54background. What is Olama?
- 1:04:56Olama is a free, easy to use tool that gives you the open
- 1:04:59source models. You have two options, either
- 1:05:02download these models on your laptop or computer locally,
- 1:05:06which is completely free. Imagine like a software, you
- 1:05:09download Gemma or you download Quinn on your software.
- 1:05:12Your data stays on your computer, doesn't go out and you
- 1:05:15don't have to pay a penny to do anything.
- 1:05:17Of course, these models compared to codecs or cloud Office 4.7
- 1:05:22are not that strong, but they can do the job.
- 1:05:25The second version is Olama on cloud.
- 1:05:28Olama will give you all these models on the cloud.
- 1:05:32If you don't have a good laptop or you don't have a good
- 1:05:34software, what you can do is you can connect to Quinn or you can
- 1:05:37connect to Gemma for 20 dollars, $30.00 a month.
- 1:05:41But these models are much much cheaper than chat GPTN codecs.
- 1:05:46Now what has happened? Ulama has integrated itself into
- 1:05:50Clod desktop application. Now you can ask Clod to do
- 1:05:55planning to come up with tasks, but you perform those tasks and
- 1:05:59activity with these open source models for a much cheaper price.
- 1:06:03So imagine if you are building something huge, if you're
- 1:06:06building an app that is very complex, instead of making every
- 1:06:10single thing with codecs or, but you can now divide the work, do
- 1:06:15the planning, strategizing, decision making with the
- 1:06:19expensive app, do the execution with the cheaper 1.
- 1:06:23So now your overall cost would be much lower.
- 1:06:25This is the first version that they released.
- 1:06:27I read the comments, everybody is asking can we use local
- 1:06:31models? So then it would be free, free
- 1:06:33to use. They said they're going to work
- 1:06:36on it and that would be next, but that is something that we're
- 1:06:38going to test immediately after this, going to connect Aloma to
- 1:06:41our cloud code. Faraz and Pratt, you're going to
- 1:06:44be using Quen from now on. Quen 3.5.
- 1:06:47They're actually pretty. Capable.
- 1:06:49They are pretty capable. I do want to ask you guys one
- 1:06:51thing. We posted a story yesterday on
- 1:06:53Instagram. If you guys don't follow us on
- 1:06:55Instagram, please do because we do ask a lot of questions.
- 1:06:57We talk about a lot of things that we do on a daily basis on
- 1:07:01Instagram. Update you guys there,
- 1:07:03especially when the show sometimes might get cancelled
- 1:07:06because we're going to an event or something.
- 1:07:08We always update you there on the.
- 1:07:09Story X and Instagram. I posted a story with the guys
- 1:07:13yesterday that asks you what is the one thing you want to learn
- 1:07:16most with AI, and I want to ask that same question here.
- 1:07:20I would love for you guys to go ahead and answer that in the
- 1:07:23comments because that will help us shape the AI education
- 1:07:25program to your needs. One of the things that I'm
- 1:07:29particularly excited about that I use it every day is codecs and
- 1:07:34cloud code. Now that is not our primary
- 1:07:36focus in the start. We're not actually starting your
- 1:07:39program by teaching you how to code and make applications
- 1:07:41because that is slightly more on the intermediate and advanced
- 1:07:45end and there's so much room for experimentation versus the stuff
- 1:07:49that we're working on in the beginning.
- 1:07:50We want to make it beginner friendly and then we slowly work
- 1:07:53our way up. But for us, because we've been
- 1:07:56working with AI tools for the longest time and also we, I
- 1:07:59study computer science, I do understand code.
- 1:08:02I cannot write it. I understand.
- 1:08:03So it's easier for me. What I'm saying is agentic AI or
- 1:08:08coding, all these things are so exciting.
- 1:08:10Things like Olama, using local models on your computer to run
- 1:08:15agentic tasks for free. These are things that we're
- 1:08:18doing right now ourselves. Would you also be particularly
- 1:08:21interested in these topics? Let us know specifically what
- 1:08:25you need help with most, and that way we can take in all of
- 1:08:28your feedback and design the program so that when you come
- 1:08:32in, you get exactly the kind of thing that you deserve.
- 1:08:35And to add what Pharaohs mentioned, things like ulama or
- 1:08:39open source model, they might sound complicated, but if you
- 1:08:43know the foundation and if you know the basic concept, the
- 1:08:47moment you understand how these open source and closed source
- 1:08:50models work, it will become much much easier for you to use at a
- 1:08:55certain point of time. These AI models are also getting
- 1:08:58easier to use as the time goes by, so if you have the
- 1:09:01foundation right, you can use all of them.
- 1:09:04Perfect. I think that about wraps it up
- 1:09:06for today's episode. We had long debates, yes.
- 1:09:09We'd love to hear your thoughts on everything.
- 1:09:11Again, every single one of them, yes.
- 1:09:13Go ahead and let us know what do you think about everything.
- 1:09:15You can number it as well, like 1-2, three, so it's per topic.
- 1:09:20I want to give a shout out. I hope you're still watching by
- 1:09:22the end of this video. We said you want to give a shout
- 1:09:24out to all of you guys watching for today's shout out, it's
- 1:09:27going to be Noxi Entertainment. You're always there with your
- 1:09:31Popcorn Time comments. Shit, so I.
- 1:09:33Want to give you some love. Thank you so much for watching
- 1:09:36for you and for everybody else. You guys are amazing.
- 1:09:39We love you. See you guys on Friday.
- 1:09:41Can't wait to see you guys on Friday.
- 1:09:42Until then, have a great week. Chat.