AI Video is getting UNREAL... (GEN 4)

So Runway just announced their latest thing, and we've seen amazing looking releases before, but then when the actual users get their hands on it, it's a little bit underwhelming. But so far, as far as I can tell, people are producing some pretty awesome AI video with it. So let's take a look. So this is the video from the founder of RunwayML. They're introducing their latest thing, which is the Gen 4, and they're going to showcase some new things that they have in store for us. What makes this model a little bit more special, a little bit more new? And again, as far as I can tell, the users that are getting their hands on it, it's available on all the paid plans so far. The response seems to be very, very positive. So let's take a listen. Hey, I'm Chris. And I'm Jamie. And today we're very excited to announce and release Gen 4, our most powerful AI model ever. Gen 4 is a completely new standard for both image and video generation. And it's the first model ever that we're putting out that achieves world consistency. That means that you can create consistent worlds with consistent environments, objects, locations, and characters. Gen 4 is a completely new standard for both image and video generation. And it's the first model ever that we're putting out that achieves world consistency. That means that you can create consistent worlds with consistent environments, objects, locations, and characters. When you can do that, you can start to tell longer form narrative content with actual continuity. You can generate the same characters, the same objects, the same locations across different scenarios. So you can block your scenes and tell your stories with intention over and over and over again. It's been very fun to design and develop Gen 4 for the last couple of months and see what the model is really capable of doing. We believe that the best demonstration of Gen 4 as a creative tool is in the stories that can be told using the model. We also put together a series of short films and experiments to showcase what you're capable of doing with Gen 4. So now we want to show you the behind the scenes of how those films were made so you can understand how powerful this model is for creating stories with consistent environments. Let's get into it. The first short we're going to take a look at is The Lonely Little Flame. It's actually a story that our director has been wanting to tell for over a decade, but never had the means to do so. With Gen 4, they were able to tell this story of a lonely little flame who just couldn't find a friend. I love this film. Yeah, it's beautiful. The opening shot is really interesting because it kind of like sets the tone and the feeling and the mood for the story itself. Yeah, and this scene is really all about this skunk searching for something. And with Gen 4, you can actually direct the subject across the scene. So you'll see we give the skunk two marks. So in this case, we wanted the skunk to go first in this side and then return, making it feel like he's looking for something. And you can see that throughout the scenes, it's the same character with the same light, with the same mood, with the same condition. And here we're introducing the character, the protagonist, right? Yeah. Again, like all great animation, you see so much expressiveness in the way that not only the characters are designed with the model, but also moving within these scenes. Okay, so what we just saw was a really great example of character consistency, bringing those same characters across different scenes, different lighting conditions, with different motions and actions being directed. But you can also just take objects, objects from the real world. So we have a little toy here. We're going to take a photo of it. And then you can basically bring that photo of this toy into Gen 4 and then bring it into any environment you want. So I'm going to do a kind of little demo here. I'm just going to take a picture of this particular object that we have here. Again, just from my phone. So you'll see it come here. I'll do that. And then what I'll do is I'll take that picture and I'll run it through a set of different environments. So now that we have that picture, I'll just drag that into Gen 4 over here. That's my reference. And I'll also use this photo of New York that I took a couple days ago. And now I'm going to type the kind of composition that I want. So in this case, I want this wooden toy, maybe leaning or next to the sidewalk of this street in New York City. Now Gen 4 is combining those references. And there we go. We have four initial images that we can browse and select. I kind of like this one. So it's animated. Maybe I want people kind of like coming in front of the toy. And now you can see how we get the motion and animation that are required, both with the toy and the city used as a reference. And of course, you can do this with any location. So we can take the toy to perhaps the mountains. We can bury it on the desert. Basically, you can do with it whatever you want. Having just seen that, I want to show off the next piece. You out there, boy. Holland, I don't like killing cows. But if I have to kill every last one of them, that would be worth it. I don't like killing cows. I don't like killing cows. I don't like killing cows. So let's pause right here. This is a shot I really like. I don't know if you noticed, but we can actually see the reflection of the character in the eye of the cow. Okay, this moment right here. This does it for me. This is so lifelike, so realistic. The physics here as the fire catches across the grass. You're still nice and dark over here and in the back, but you're getting this nice fall off. Super realistic. So the really interesting thing about Gen 4 is that you can use consistent characters and objects and environments all across your scenes. So in this case, we can start with a character with design, with cast. We can use another character and environment to basically create the mood, the look and feel, how the character would want to look. You can create a net new image out of that, and then you can branch again out of that image. So you can prompt the model for different shots, different angles, or maybe some changes on how you want that person, that object, or that environment to react. Which is really powerful because you can take anything from real images to generated. You can look at anything around you and use it as a prop, either for the characters, the environments, or the objects that are in your shots. Okay, so this next one is called New York. Is this new? And it's a really great showcase of the visual effects capabilities of Gen 4. And this is combining real photos we took of New York, different parts of the city. And we're taking real photos of animals and we're basically merging those together to get all these interesting compositions of animals in impossible locations. Yeah, and it also is a clear demonstration of its understanding of physics, the weight of these animals, how they move across surfaces, the way they interact with those surfaces. It's so fun to make this because, again, you're making this in just a couple of hours. And it feels very addictive to be able to get your ideas out of your head really fast by having a model that can understand or be your creative partner in a way. I think one of the most powerful and fun things about Gen 4 is that it allows you to combine anything with everything that you want. Well, sometimes it starts with one image. The next video that we're going to show started with a generation of a monkey in a white studio space. And from there, it just snowballed into a music video that we created. So let's go check that out now. I hope by now you get a better sense of how powerful and fun Gen 4 is to make everything from consistent worlds to characters, locations in ways that you couldn't do before. Gen 4 is now available to all of our paid users and enterprise customers. Scene references for consistency with character, locations, and objects will be coming soon. Here's one from somebody on the Runways team. Some they made this one from a prompt they had laying around. And it's looking really good, really lifelike. I mean, this thing is obviously some sort of a cloth or cotton or something like that that's crawling across a rock. But it looks really, really good. It looks like those macro stop-motion animations. This wolf looking good. Great shadows. Great field of view. Another style snake. Again, you can tell that the texture is some sort of a, what do you even call that? It's some sort of cloth-like material. Looking flawless. So this is definitely replicating those macro zoomed in stop-motion animation videos. And it's looking good. Excellent reflections. The water dripping from the deer's mouth. Absolutely terrific. Here's somebody testing some of their go-to prompts. And it's looking pretty good. So they're saying it seems flawless. So definitely this space theme is looking very, very good. Very high-tech. All the characters seem consistent in those shots. I mean, there's no weird morphing that's happening. It's a little bit more like an anime style. The fingers looked weird. But other than that, I haven't spotted anything that's obviously wrong or off. Here's somebody that's saying this is their first generation in runway Gen 4. This is not cherry-picked. I mean, that's looking really good, right? The pillows, the force in the background, the movement shot off a train. It's really lifelike. This is some sort of a fast and furious remake with whatever, I don't know what that character is, the blue cat thing, but terrific. So what happens if we take an image like this? So we do have like the bottom of here, but notice that the top, there's nothing past this. But this is what runway Gen 3 generates. So it figures out kind of what's above it, builds a whole world around it, reflective buildings, billboards, etc. It's looking really good. Here's one about parental relationships. It's looking, yeah, very realistic. I mean, you can tell this lady seems a little bit off, a little bit, you know, stiff or something like that. You can tell it's like animated, just a little bit off, not 100% smooth. But other than that, looking pretty, pretty good. So I got my hands on Gen 4. It takes a while. So you have to have the paid plan. It did take about five minutes until I had access to it. But let's, let's try it out. So they recommend, you know, starting out with more simple prompts. Don't underestimate the power of simplicity, as they say. So don't get too complicated. Try maybe a simpler prompt first before getting a little bit too crazy, too specific. But since the thing that they're working on is kind of having character consistency and, you know, controlling the camera movements and stuff like that. Here are some elements that tend to work well to refine the output. So we have subject motion, camera motion, scene motion, and style descriptors. So here it's a handheld camera. Tracks the mechanical bull as it runs across the desert. The movement disturbs dust the trails behind the mechanical creature. Cinematic live action. Alright, so we're going to try this. The fit man covered in tattoos climbing up a rainforest tree. I found this insanity on mid-journey. Let's see if we can replicate that. I think the mask is a nice touch. Alright, we'll just do a five second clip and let's go. How about this angry cappuccino kitten? Should we have it give an inspiring motivational speech to various sales agents? Maybe saying something like, Coffee is foreclosers. Yes, I think that's perfect. Cat with coffee giving it an inspiring motivational speech. Now, let's try a little bit more of an action shot. How about we take this reference image and we have it follow the skier down the slope. So we'll see camera following the skier down the mountain. Alright, let's try a hard one. So we're going to have this lady spinning one way and the camera spinning the other way. So we're going to see woman spinning clockwise, the camera going around her counterclockwise. So here's a heart. Let's have the camera kind of shake a little bit every heartbeat. So kind of like that heartbeat reverberates through the camera. The heart is beating and shakes the camera with every beat. A few days ago, Sam Ullman politely asked all of us, Can y'all please chill on generating images? This is insane. Our team needs a sleep. I replied with something that I hoped would be motivational for him and his team in their quest to not melt their GPUs as the entire world is gibbler-fying everything that they can get their hands on. So here's my artistic rendition of the OpenAI management team motivating the employees into producing more gibbler images and going without rest. I think I really captured the essence of it, but let's see if we can animate it. The hamster pulled carriage goes ever faster. Meanwhile, some of our videos came back. So let's check those out. So this is the man that's supposedly climbing up a tree. I think a rain forest tree. All right. I mean, that's pretty good. He's kind of walking on air a little bit, but everything else is looking like the initial few frames are phenomenal at taking that mid journey image and very faithfully kind of turning it. And then, you know, it goes into another scene there. Maybe I would mark it down, you know, because it kind of gets a little bit more different and inconsistent. But man, those first few seconds, it really keeps that character very similar to what the mid journey image is. So it almost seems like all the tattoos and everything else as he's turning. Notice how. So this is the original image, right? Like it's right here. So notice as he's turning here. I mean, it really made everything very accurate in terms of like how it would turn in 3D space. I'll give it like a four out of five because the second part is a little bit. It's not bad. It's not bad. But I mean, he's walking on air and and maybe since I said climbing up a rainforest tree, maybe that's why we have kind of that up shot. We got to play around with it, but it's not bad. All right. Here's our coffee is for closers. Motivational speech. Hey, I believe it. I don't know about you. I'm motivated. That's an aggressive looking kitten. All right. So let me blow this up a little bit. So this is following the skiers down the hill. So it's definitely has a character consist. I just as about to say character consistency, somebody's red leg got ripped off. I don't know what happened there. So again, I feel like the first few seconds are exceptional and the last few seconds maybe get a little bit weird, but still not bad. So notice this is the original. So it's a macro photography snow, not Lego people, but some sort of a plastic little figurines. And as it continues, I got to say, I mean, they really capture the essence of it. Like it's, it's very, very good. Then they continue here. It's still, you can tell that it's macro photography, right? Still plastic pieces. I don't know what happened to this person's leg right there. Uh, I don't know. Without that, I think I'd give it like a five out of five, but, uh, that whatever that is ruins it a little bit. I mean, is that, what am I looking at even? I don't know. Here's our spinning lady spinning camera. Let's see how well. Okay. So she's going clock. She's going counterclockwise. The camera seems to be standing still. So this is something that a lot of these image generations, video generations have a problem with, right? The kind of like morphing effect VO two from Google. It was excellent, but it also had some issues with this. So this is what, like a two out of five, I would say. Um, I mean, very consistent character in terms of like, it's the same person, but doesn't show the 360 spin. Ignores my camera spin instruction, the beating heart that pulsates and, uh, shakes the camera. So again, we're not seeing the camera shaking the pulsating hearts kind of, I guess, not quite how a heart beats, but I mean, it's getting the motion in there and the hamsters. Let's see, please tell me in nailed hamsters. That's a, that's not bad. I mean, you can really tell those hamsters are moving the two people controlling the carriage. They're kind of smiling. That's not bad. I like it. Five out of five. All right. So next let's just do in rapid fire succession, a whole bunch of a mid journey images with no descriptive text. So there's no instructions. We're saying here's an image animated because really this is what I think the new thing is really good at. It's sort of like taking a characters and then creating consistent characters in video, having them follow certain instructions, camera angles, et cetera. So by the way, all the stuff that you saw before this, it was not cherry picked. So every single image that I uploaded, every single description that I wrote in there, you saw every single one of those. So if we looked at whatever, seven or eight, that's because I uploaded seven or eight, whatever that number was. And now we're doing the same thing just with 18 mid journey images. And these are with no description. So the first one, you can kind of see the first frame is that original image. So this is how it got animated. So runway decided to zoom in as the creepy skull is kind of beginning to slowly turn towards you. Very, very cool. Definitely kind of captured the creepiness of the scene. And now we have a pizza delivery on a scooter in Italy. I'm guessing. Let's see how this one turned out. And there it goes. The kitty is keeping up with the scooter. Very, very cool. That's looking pretty good. Simple, but I like it. It figured out that since the image sort of, you got that vanishing point, the street is going this way. The only way it would make sense to animate this is like this, right? It going forward. Anything else would be a little bit weird. Next, we have this spirit wolf or whatever you want to call it. And it begins slowly walking through the grass. I mean, it's looking good. So here we have this wonderfulness going on. So let's see what it decided to do here. And wow. So first of all, it's clear, sharper, and more realistic. I feel like then, than the original image, the fluid pouring down is very lifelike. It's got the reflections. It's pretty good. What's some wizard magic. Let's see how well it did that. Great looking fire, right? It's fire as you would expect it to be in that kind of like fantasy setting. So it's maybe not photo realistic, but it's definitely realistic. Like if you saw this in an animation, you would be like, that's yeah, that's what that animation would look like. Then we got this thing. So some sort of a cartoonish character, leaves of grass. He's looking around. It's kind of a pixel style art. I think they nailed it. In fact, when he turns, it's almost like a higher resolution than the rest of the image. But all in all, pretty good. Here's a, some sort of a space deity space. None. I don't know. But a robot. None. I don't know. When she turns, I got to say that is flawless. They capture all the face features, the sort of that paint stripe on the side of the hood. Like how well they capture everything when they turn is phenomenal. It's shadows, the reflections of the light on her face skin, robot face. I don't know, but it looks good. And this is some sort of a pencil sketch of a cat creature. Um, I guess could be a bobcat or one of those Sphinx cats. So let's see, is going to start moving. So it turns away. So it rendered it as a Sphinx cat. One of those hairless cats. It did a great job doing it. How well it's able to kind of recreate the 3D rotation is, is pretty amazing. I gotta say. Here's a lady is smiling in the sunlight and she kind of had crazy eyes there for a second and kind of a weird jerking motion moves a little bit like a dog, a little bit like an animal. So that's not the best one so far. Maybe four out of five, three out of five. I guess maybe she's laughing, but yeah, it's, it's a little bit jerky. It's a little bit weird. All right. Next we've got this horse kind of a jumping over the viewpoint of the camera. Let's see how well he did that. Uh, so that's what they're seeing. Now it turned into barbed wire. The horse changed a little bit, kind of a lossome definition. The style of the image changed. So this so far, I feel like is the worst generation so far ever, like the lowest quality. It's not horrible. It's not bad, but this is closer to what we would expect to see with most of the other image and video models. Like I would say this is bad, but it's rare to get something even this, this bad, so to speak. So it's, it's not, not that horrible. I would say like the horse didn't fly off into the sky or anything like it, the fence kind of kept its texture, et cetera. Not horrible. I'm going to say 2.5 out of five. I think that's fair. All right. I was very interested how it's going to do this. So we've kind of got a, uh, formula one indie racing, what on Saturn on the rings of Saturn. All right, let's see what it does here. And there we go. I mean, that's looking pretty good. It's definitely capturing the, it seems like, you know, it was turning left and it started turning right, which I think maybe it's not quite how the rings of Saturn go. But I mean, again, these are with known text instructions. So, um, in terms of everything else, it's looking pretty good. You can still, you know, that it's a formula one or Indy car, wherever that is. It keeps that consistency. It understands it's in space. So it's showing all the other planets kind of, um, in the vicinity for four out of five, I would say. I saw this one up voted at mid journey and it kind of cracked me up. It's on one hand, it's such a wholesome family portrait. The other hand, the, uh, the golden grill, uh, it's, it's, it's an interesting, it's just an interesting image. Let's see what runway does with it. I mean, it's supposed to be like a family. So it doesn't do too much with it. It kind of animates them a little bit, but they barely move. I mean, it's supposed to be like a wholesome family picture, which I guess that's what you do for you hold still for a family picture. Okay. Well, we'll give it to it. It's not bad. I didn't do too many photo realistic ones yet. So I kind of try to want to squeeze a few more like photo realistic. Like photo realistic ones as well to see how well it's able to capture that. So here's one kind of a holding a pomegranates. That is terrific. Wow. That's actually incredibly good. So notice how, I mean, there's a reflection of what appears to be the sun in the pomegranate in the first shot. So when she turns the sunlight falls on her, on her shoulder, something about the movement is a little bit like, it's almost like it's accelerated a little bit, but other than that, this is just absolutely stellar. Absolutely phenomenal. I wanted to see what it would be able to do with this one. Let's see. So it didn't really animate that. All right. I would give it a low score just because it didn't really animate. It just zoomed in, but okay. Here's another one. That's sort of a photo realistic. Let's see. Oh, that's interesting. So it kind of put them on a train. The train is rocking back and forth. That's really good because notice. So this is the original shot, right? So it's not obvious that they're on a train. I would not have guessed that they're on a train. I would have guessed that they're, this is like a lodge or some sort of a cabin or something like that. All right. This looks like wood. This is the roof, right? Overlooking snow. Runway decided to go with something moving like a train, which is totally cool. And they definitely nailed it. So the scenery is moving past them kind of a parallax perspective. There's some highlighting. The kid is like a super happy. Wow. That's five out of five. I got to say this one. I picked specifically for the camera angle. So we got the massive buildings. We've got the massive, what is that? It's sloth, right? Wearing a hoodie and the, we're looking at it from like way down below. So this is like a human hand here. So this thing is, you know, a skyscraper height. Looks like there's some destruction going on here. So let's see how well it did here because this is kind of a hard thing to do. It's not doing too much, unfortunately. So, I mean, yeah, it animated it, but it didn't do anything too exciting. All right. And the hand is just sticking up, I guess. Okay. And this one is the last one from this batch. So we kind of got a rainy, foggy day and a massive, massive plane flying overhead over some cars, maybe a bus. There's a person there. Let's see what it did with this thing. All right. So the plane is kind of slow. The bus is moving. Two cars collide and disappear. That's interesting. So you can see right here, they're kind of like collide and disappear. The person is just walking down the middle of the road. I don't know what to give it here because it definitely captured something very interesting. I mean, there's a lot of things here that visually look good, but logically don't really make sense. So I don't know if I want to give it too many points off for the logic not being there just because again, so these are without instructions. In fact, let's rerun this with some text instructions and see where it takes us. All right. So we're going to say the camera floats up as the plane flies head on into the camera view. So we'll try that. Then we're going to try the camera kind of tracks the plane going overhead, right? So kind of like tilts up as the plane is going over us. The plane moves in slow motion as the cars move. And I'm blanking on the word where stuff moves super fast. What is that? Time lapse. So time lapse speed, like a super, super fast motion. So we're going to say the plane moves in slow motion as the cars move in, let's say hyper time lapse speed. And we're going to say the camera is still. Okay. So we have three different sort of ideas about how the camera should move when it floats up, when it pans up and when it stands still. And last one, we're going to say the camera moves back to keep the plane in view. All right. So let's see what it did here. So this is kind of the first one that it generated for us. So this was without any sort of instructions in our next one. We asked it to. So this is the camera floats up as the plane flies into the head of the camera. Let's see how it does that. So I wouldn't call this floating up. This is more what we want from the next one. But I got to say, this is a great shot. Nevertheless, it's not quite the camera action thing that we're looking for, but man, is that a good shot just on its own. It's, it's, it's phenomenal. So next we have the camera tracks the plane going overhead. That's what we wanted. Like the previous one is kind of what this prompt was supposed to have been. So let's see what it does here. Okay. So here it seems like the camera is moving back to keep the plane in its view. So I'm a little bit confused about how to describe the proper sort of camera motions in a way that it understands. Obviously it's really good at doing the various camera shots. It's just the text that I put in doesn't seem to match up to the actual camera shot. But as far as the camera shots go there, they're terrific. They're excellent. And finally, the plane moves in slow motion as the cars move in hyper time lapse speed. So let's see what it did there. That's not what we were looking for. So this one, I would give it a low score. I mean, it's still generated everything, but it's like one out of five in terms of following the directions. And finally, we have the camera moves back to keep the plane in view. Oh, that is actually pretty good. Okay. So this is kind of exactly what we wanted. So the camera is moving back at the same speed as the plane. So everything kind of receding into the distance, but the plane more or less, the perspective stays the same. This is five out of five. This is terrific. All right. Next, let's see how well we can get the characters to behave on screen. So we got this gnarly cat. Let's see what it can do. So cat takes off sunglasses and smiles. So it's, as you can see here, it's like a full cat. Let's do cat hides the knife. How about cat picks at his own claw with the knife and how a cat sidekicks the camera. Oh, I didn't expect that. Did you? Here's another one that's ready. The camera follows the car as it races at high speeds through the city. So let's see how well it did this. That's pretty good. I mean, it's following the car high speed kind of maybe. Yeah. Camera motion nails it. This is very, very good. I feel like the style changed. The style changed pretty drastically. There's a lot of great things going on here. The reflection off of the palm trees in the sky is, is phenomenal. The reflections of the, uh, taillights. Excellent. You know, the only thing is this is the style that we were going for, right? It's kind of a more gritty pastel. I don't know what, but it's maybe like a kind of a GTA style, maybe a little bit. And then this is what we're getting. I mean, it's very close, but it's not the same style. It's more smooth and beautiful. The other one's a little bit more gritty. And we got some like moisture splattering on, on our viewpoint. We got this little glitchiness thing going on here, a little bit of a lens flare. So that's kind of lost here. So, I mean, in and of itself, it looks great. It follows the instructions very well. Loses the style. But other than that, it's phenomenal. All right. Here's our naughty kitty taking off the sunglasses and smiling. So, hey, that's pretty good. That's pretty good. It definitely does everything really well. The fact that it gets the, the, the, the knife so close to its eyes is a little bit unsettling, but I mean, yes, it's a smiling cat. Yes. It took off its sunglasses. I would give this a five out of five. This is terrific. Next we have the cat hides the knife. So, you know, it hears sirens in the distance and what happened? It, what hides the knife. Okay. So we, okay. So kind of like turns around. It's like, Oh, I don't know. I don't know what's going on here. Where did it go? I mean, that's pretty good. That's not what I had envisioned, but certainly I feel like this, this nails the prompt. It's kind of acting a little sheepish, like, Oh, like trying to figure out if, if somebody saw it, do it or not. The view from the back of the head, the sunglasses kind of wrapped around that I got to say, that's a very, very, very good. I'd give that a five out of five, even though that's not quite what I had in mind, but still it nailed the prompt. I feel like, all right. So here the cat picks at his own claw with the knife. Let's see how, Oh, that's what, what is happening? Okay. So this is, um, another one where I would have to kind of grade it poorly. This one in the horse jumping over the fence. Like it, it kind of is, you know, like, I don't know, two out of five, one out of five, something like that. I mean, I don't even know what happened there. It's a different cat all of a sudden. So this isn't good, but you know, so far we've seen not that many of bad generations. They're, they're mostly pretty high quality ones. All right. Here's a cat sidekicks the camera. Let's see what happens here. I'm very, very curious. All right. Not what I envisioned. I would give it like a three out of five, maybe a four out of five. I would give it like a three out of five for this one. Cause it captured the sidekick. Maybe having it kick like the, the camera is very, very difficult. I don't know what's happening with the knife there, but yeah, like maybe a three out of five. I think interestingly, I accidentally clicked on sidekicks the camera twice. And, uh, so this is kind of another interpretation. This is much better. So this, I would give like a four or four and a half out of five. I get to give it a lot of credit cause it looks good. It's a cat side kicking the camera. Then, you know, the tail moving, even though everything else stops, like it's an interesting artistic design. It's wearing those like, you know, jeans or pants or whatever. Great sidekick, great form. The whole, I don't know what happened with the rings around it, but there's a lot of great things happening there. Consistent character design. Again, not kicking the camera, but that prompt, it just might not be the right prompt to, to be testing out here. If this was, you know, a cat doing a sidekick, this would get a perfect five out of five stars from me. All right. Here's another one they did from an image from mid journey of a person flying this airplane through the clouds and absolutely terrific. Even the instrument panel. I mean, yeah, you can't really see what it says. You can't see the numbers or anything, but I mean, it looks very legitimate. It looks very real. The things that are circles are look kind of like circles. They got something blinking there. Uh, the pilot is tweaking the buttons, changing the buttons. Uh, you know, if you don't look too hard, everything is looking phenomenal. Certainly the sort of like little scratches on the, the cockpit on the windows are very realistic. The sun shining and kind of changing how much of those scratches we can see is phenomenal. The clouds are phenomenal. There's a lot of just incredible, incredible stuff happening here. And I love it. This is a five out of five for me. So that's a pretty much it. You saw pretty much every single thing that I uploaded in there. Nothing was cherry picked. There was just one project, one image that it refused to do because it had some copyrighted character in it. Other than that, it did everything phenomenally. Well, there was maybe like one or two generations that I would rate poorly. The rest of them were either phenomenal or, or maybe they missed just some portion of the prompt, et cetera. But overall, like this is a, a very, I'm very happy with how it generates stuff. I'm very happy how well it's able to keep the consistent characters. As I'm recording this, I'm going to see if I can take this video and use another one of other tools called lip sync video. So I don't know. Will I be the lawyer cat or will I find something else to test out? I don't know. But let me know what you thought about this whole thing. Does it look good? Is it seem like AI video is just getting better and better? Are you excited about the fact that the character consistency and all that stuff is developing really well runway seems to be making a play for kind of the movie development studios. So this isn't just for, you know, funsies for everybody to play around with. They, they have a real business plan. They want artists creating on this platform. Do you think we're pretty close to the point where people can create actual movies and enter them in competitions? Do you think that's really going to catch on? We've already seen a few examples of that, but do you see this as just a fad or is this like the start of the next big thing? Let me know in the comments if you made it this far. Thank you so much for watching. My name is Wes Roth and I'll see you next time. Bye.