AI Video From OpenAI Just Blew Everyone's Minds!

Remember when we got that huge leap in AI art from mid journey version 3 to mid journey version 4, where all of a sudden all of the images looked realistic and just insane? Well, today is the day that AI video generation just made that same leap. This is mind blowing. This is AI generated video right here. Today, February 15, 2024, OpenAI announced Sora, an AI model for creating video from text. This, this is the best AI text to video generation tool we've ever seen. Here's the official announcement from OpenAI on their Twitter introducing Sora, our text to video model. Sora can create videos up to 60 seconds featuring highly detailed scenes, complex camera motion and multiple characters with vibrant emotions. Up until now we've only been able to create three or four seconds at a time. And then if we wanted longer, we can extend it and extend it up to about 16 seconds. This is a single prompt, 60 second video, super realistic. Just check out this video that they shared as the example video when they made their announcement on Twitter. It looks super realistic. I mean the, it's a little pixely, but I think that's more due to Twitter's compression than the actual video itself. This is just absolutely phenomenal and there are tons of examples already of what this is capable of. Greg Brockman also took to Twitter to announce Sora announcing Sora, our model which creates minute long videos from a text prompt. And you can see his AI generated video of a woman walking down the streets of what looks like maybe Tokyo at night after a rainy day or something. But the realism on this thing is just amazing. We haven't seen anything like this with AI video generation yet. Now if you head on over to OpenAI.com Sora you can find out all about it and see a whole bunch of examples of the types of videos it can create. Now unfortunately we don't have access to this model ourselves yet. You can see here in Sam Altman's tweet about it, you can see he says today we are starting red teaming and offering access to a limited number of creators. I don't know how they're choosing those creators and who gets access, but as far as I know right now, not many people. If anyone has access outside of the OpenAI team, let's just take a moment to look at some of the examples that they shared on their website of the videos that this can generate. Like these woolly mammoths walking towards us in a 10 second video here, or this movie trailer featuring the adventures of a 30 year old spaceman wearing a red wool knitted Motorcycle helmet. I mean the realism on this is just insane. And you can see what I mean by the quality from Twitter where it looked a little more compressed. Because these videos that we're seeing on the demo site are just mind blowing. I mean these are absolutely insane. How well this is working. Here's another one with an animated scene featuring a close up of a short fluffy monster kneeling beside a melted red candle. The art style is 3D and realistic with focus on lighting and texture. A gorgeously rendered papercraft world of a coral reef rife with colorful fish and sea creatures. Look at that. That is so good. It looks like origami almost. And this video is 20 seconds long. Here we got a vertical video. This close up shot of a Victoria crowned pigeon showcases its striking blue plumage and red chest. Etc, that's a long prompt, but this is a 40 second video. Look how realistic that is. You saw this video online, you would probably assume it's a real video. It doesn't look like AI art at all. Photorealistic close up video of two pirate ships battling each other as they sail inside of a coffee cup. Look at that. You can tell it's supposed to be coffee in a coffee cup and two pirate ships. I mean it's. And it's a 15 second generation. Now because all of these videos that we're looking at right now are on the OpenAI website, there's a fairly good chance that these are cherry picked videos. These are some of the best results that they've gotten. But man, if this is the best results you get, that just tells you that maybe through enough prompting you can generate anything you want from these. Here's another one. A close up view of a glass sphere that has a Zen garden within it. There's a small dwarf in the sphere who's raking the Zen garden and creating patterns in the sand. This is a 10 second video. Now most of these videos I'm not going to show you the whole clip, otherwise this video will be 30 minutes long. But definitely check out OpenAI.comsora and we're going to look at, you know, fractions of each of these clips just to get an idea of the quality. Like this eyeball here or this kangaroo disco dancing. That's a 10 second long clip. You haven't seen Pika or Runway or any of these AI art video generators generating something this good. Anything else that we would get that generates like this would look like the kangaroo was moving in slow motion to do his dance. This actually looks like a kangaroo dancing. We've got a 20 second clip of a beautiful homemade video showing the people of Lagos, Nigeria in the year 2056, shot with a mobile phone camera. Look at that. Look at the realism. I mean, there's not enough, you know, adjectives for me to use to talk about how excited and blown away I am by some of these videos and how well these are coming out. The camera rotates around a large stack of vintage televisions, all showing different programs. 1950 sci fi movies, horror movies, news static 1970s sitcom. And you can see all of these various types of things in these videos. I mean, the prompt adherence alone is mind blowing. 3D animation of a small round, fluffy creature with big expressive eyes explores a vibrant enchanted forest. Says here, Sora is able to generate complex scenes with multiple characters, specific types of motion and accurate details on the subject and background. The model understands not only what the user has asked for in the prompt, but also how those things exist in the physical world. The camera falls behind a white vintage SUV with a black roof rack as it speeds up a steep dirt road surrounded by pine trees. A on a steep mountain slope. Dust kicks up from its tires. The sunlight shines on the SUV as it speeds along the dirt road, casting a warm glow over the sea. The dirt road curves gently into the distance with no other cars or vehicles in sight. The trees on either side of the road are redwoods with patches of greenery scattered throughout. The car is seen from the rear, following the curve with ease, making it seem as if on a rugged drive through the rugged terrain. The dirt road itself is surrounded by steep hills and mountains with clear blue sky above the wispy clouds. Who? All right, that's a 20 second video that used that entire prompt. And if you read the prompt, most of it made it into the video. It is just crazy. You can't tell I'm excited. Check this out. Reflections in the window of a train traveling through the Tokyo suburbs. You could see the reflections of the other people in the train as it passes by. Dark scenes. There is no way you would see this on social media and assume this was AI generated. A drone camera circles around a beautiful historic church built on a rocky outcropping a along amalfi coast. A 20 second video. Look at that. That looks like a real drone shot. It just does a large octopus in a scene resting on the bottom ocean floor. I mean, these long prompts are working really well. Look at that. 17 seconds long. Flock of paper airplanes 20 seconds long. Look at this cat. That cat is a realistic cat. Look at as it hits the girl's face, she like squints and her eyes move to the cat hitting her. Look at that. Oh, smash my face. A cat waking up its sleeping owner demanding breakfast. That cat, that cat is real. That is not an AI generated cat. That just blows my mind. Borneo wildlife on the Kinabatagon River. I probably butchered that name. A Chinese Lunar New Year celebration video with Chinese dragon. Look at this clip in full screen and tell me this doesn't look like an actual video clip from China or something. Maybe it's not shot on the most powerful camera in the world, but it looks real. This site is just full of examples. A tour of an art gallery. A stop motion animation of a flower growing. A robot's life in a cyberpunk city. Look at that. An extreme close up of a gray haired man with a beard in his 60s. 20 second clip. And it is just so realistic. A beautiful silhouette animation shows a wolf howling at the moon, feeling lonely. You know, I love this clip. I love my wolves. New York City submerged like Atlantis. Fish, whales, sea turtles and sharks swim throughout the streets of New York. A litter of golden retriever puppies playing in the snow. Their heads pop out of the snow covered in. And then the prompt actually cuts off. But that's a 20 second video of golden retriever puppies playing in the snow. No doubt about that in my mind. Someone running backwards on a treadmill. Five gray wolves frolicking basketball through hoop that explodes. All right, I got to see it explode. Boom. Nice. Archeologists discover generic plastic chair in the desert. Grandmother with neatly combed gray hair stands behind a colorful birthday cake. Here's a dalmatian crossing between two windows. An otter on a surfboard. 17 second video. A close up shot of a chameleon. I mean crazy realistic. A corgi vlogging in a tropical Maui. A white and orange tabby cat is seen happily darting through a dense garden as if chasing something. 27 second clip. Aerial view of Santorini during the blue hour. Showcasing the stunning architecture. Etc. That's an eight second clip. Tilt shift of a construction site filled with workers, equipment and heavy machinery. That's a really good tilt shift effect as well. That works really well. A giant towering cloud of the shape of a man looms over the earth. Holy crap, look at that. That is amazing. Some dogs, this one, they're running kind of sideways. This one looks a little more AI to me. This is like one of the first ones I've come across from. Like, all right, that's clearly AI like something's Going on with the legs. Still really, really good, though. The Glynfinen viaduct is a historic railway bridge in Scotland, uk, that crosses over the West Highland. That looks real video. You put this online and told me this was a real video. I would not second guess this one at all. Again, all of these examples in their entirety are available to watch over@OpenAI.com Sora and then of course, Sam Altman went to X to say, hey, what do you want to see me prompt? And he's actually taking requests and generating videos from requests here over on X. Here's one. A wizard wearing a pointed hat and a blue robe with white stars, casting a spell that shoots lightning from his hand and holding an old tome. And in his other hand, that pretty much nailed it. Half duck, half dragon flies through a beautiful sunset with a hamster dressed in adventure gear on its back. That one's looking a little AI to me. The dragon sort of going backwards instead of forwards. The little hamster dude on his back is kind of sitting on him funky. But I got the half dragon, half duck. I mean, that's crazy that it managed to make that sort of morph a street level tour through a futuristic city, which in harmony with nature and also simultaneously cyberpunk high tech. And there's a lot more to that prompt here. But you can see the video here looks like a cyberpunk high tech futuristic city. I mean, it does what it says it's supposed to do. Here's another one of a half duck, half dragon flies through a beautiful sunset with a hamster dressed in adventure gear on its back. That druck, drag, duck, duck, Dagon, whatever you want to call it also kind of looks like it's flying backwards, but that looks much more like a hamster riding it. But then Sam had to take a break, and so those were the only demos that he gave us. But oh my gosh, this Sora app is absolutely blowing my mind right now. If you can't tell from my excitement, this is the best I've ever seen with AI Text to video blows every other model out of the water. Unfortunately, we just don't have access to it yet. When you think about, like most movies and most TV shows, if you watch closely, they never really have a shot stay on the screen for more than like three to five seconds. It's constantly changing camera angles and changing shots. Well, with a tool like this where you can generate up to a minute long clip, you can pretty much generate every shot from a film and just keep jumping between them, knowing that most Videos aren't going to have more than five seconds on one scene at a time. So I mean we're getting really, really, really close to AI video being able to generate anything you can imagine. And the amount of time is going to take to prompt this stuff is only going to get quicker and quicker. Our home computers are going to get more powerful, but these models are also going to get more and more efficient. So you'll be able to generate them with less and less compute power. And with all of this, we're likely going to get longer and longer generations. If we can already do 60 second generations, it's not too far of a stretch to believe that we can double that, triple that, quadruple that within the next year. I'm just so blown away. I can't imagine what AI video is going to look like this time next year. Like just for an example, here, here's something that Garrett Scott posted on X. Here's where we were one year ago with Will Smith eating spaghetti. And you'll notice none of these clips are longer than three or four seconds long. It keeps on changing scene because you're only able to generate 3 or 4 second long videos back then. That's what we were generating one year ago today. And here's what we get today. This is the kind of quality it's generating now. Somebody needs to go and prompt Will Smith eating spaghetti so we can compare those two side by side directly. But holy crap, the leap we made in just one year from February 2023 to February 2024, mind boggling, absolutely insane leap. Now I know there's a lot of people that are watching this video that are on every single one of my videos in the comments saying that's not that good. I'm not that impressed. You know, movie creators have nothing to worry about. I'm not saying that you can use this right now to make a full length video. What I am saying is if you look at where we were a year ago, remember a Will Smith eating spaghetti here to where we are today, one year later, four second video that's grainy and really unrealistic to up to 60 second videos that look really realistic. Where are we going to be a year from now? Maybe got nothing to worry about today. But these videos are getting so much better. I also still believe that there's a role for people creating videos. I think people are still going to want to see human made videos. I think AI is going to give them superpowers, make it even easier for them to make videos, make it even easier for them to get the shots that they couldn't otherwise get. I don't buy into a lot of the narratives of game over for movie directors. If you're a videographer, you're in trouble. I don't buy into any of those narratives if you're a film director. If you're a videographer, you're about to get superpowers, not you're about to get eliminated. That's my thoughts on it anyway. Sora from OpenAI super excited about it. Had to make this quick video tell you about it just in case you hadn't seen it yet. This is mind blowing. It's the biggest leap we've seen and AI video generation hands down ever. Mind blowing. I can't wait to get my hands on it. I bet you're the same way. Thanks so much for tuning in. If you want to see more videos like this and really cool AI advancements news tutorials as they happen like this video, subscribe to this channel and I'll make sure more pop up in your YouTube feed. If you're not familiar, I also run the site Future Tools. Lots and lots of cool AI tools to be discovered on this site. The AI news is updated constantly and there's a free newsletter that you can join. I'll also share really cool ways to make money with AI. If you join that free newsletter, you can find it all over at FutureTools IO. Thank you so much for tuning in. Nerding out with me. What an exciting day. I am jacked up over this. I am so pumped to see where we've come. I want to get my hands on it. I'm sure you do too and I will keep us all in the loop. Let you know as soon as it becomes available for commoners like us to be able to play with this. So make sure you're locked in on this channel and I will let you know when it's ready. Thanks again for tuning in. Really appreciate you. See you in the next one. Bye bye.