# NVIDIAS NEW Insane PICASSO SHOCKS The Entire Industry! (NEW SOFTWARE ANNOUNCED!) ## Метаданные - **Канал:** TheAIGRID - **YouTube:** https://www.youtube.com/watch?v=dHY5LWfkYLk - **Дата:** 22.03.2023 - **Длительность:** 10:40 - **Просмотры:** 407,791 ## Описание NVIDIA Picasso is a cloud service for building generative AI–powered visual applications. Enterprises, software creators, and service providers can run inference on their models, train NVIDIA Edify foundation models on proprietary data, or start from pretrained models to generate image, video, and 3D content from text prompts. Picasso service is fully optimized for GPUs and streamlines training, optimization, and inference on NVIDIA DGX Cloud. https://www.nvidia.com/en-gb/gpu-cloud/picasso/ contact@theaigrid.com (For Business Enquiries) ## Содержание ### [0:00](https://www.youtube.com/watch?v=dHY5LWfkYLk) Segment 1 (00:00 - 05:00) so Nvidia just announced something that is completely game changing introducing Picasso at their new event at the 2023 keynote they introduced a whole wide range of new AI applications that's going to be powered by nvidia's large GPU service and this is truly going to change the game for AI because there are a large amount of applications so introducing Picasso a cloud service for building and deploying generative AI powered image and 3D applications you can all see that this is definitely going to change the game because this is what the next set of companies are going to be built on now you can see right here it says what is in video Picasso and essentially it's something that other companies can build on and they can use to train different models and essentially you know make their companies off of because these companies can't create these image Generations themselves now the three main things that Nvidia is working on with Picasso that it did just release and you can all see that one of them is one that we are very familiar with but the other two aren't quite too familiar so of course you can see on the left hand side we've got text to image which is something that many different you know many different image generation platforms have such as mid Journey Dolly 2 stable of fusion and even Microsoft recently with their new announcements but now we have something new oh text the video is fairly new because it is something that is very hard to develop but Nvidia is working on this and it says temporal layers and novel video denoiser generates High Fidelity videos with temporal consistency now this is truly interesting because they aren't the only company that have been working on this and if we've seen anything recently by a company called Runway you'll know that they've actually been working on this for quite some time and their advancements are very decent and they look pretty good which brings me to the question how crazy will text a video look in the future are we going to get entire movies basic videos I mean it truly depends on the user's input but right now what we are seeing in the early stages of AI shows us a very promising future especially with text to video and I know that yes what you're seeing behind you is footage from Runway but this is just to show you what is capable with text to video and the current stages that it is at because of course Nvidia haven't really released too much footage on their text to video AI software but this isn't about Runway because of course they're doing incredible things but this is about Nvidia and what is crazy about Nvidia is that with the entire event this small texture video although it's gonna literally revolutionize the creative industry it only seemed to be a very small part of their entire keynote event which honestly I'll leave a link to a short summary of the presentation but I just want to dive in to the very specifics that are Nvidia Picasso which compromises of the three key points that I talked about earlier text to image text a video and text to 3D now the last one text through 3D is definitely very interesting there are a few models and a few you know research papers that are online that you can actually use and I was meant to make a tutorial I should still be in the works of making that because um it is actually possible to do this and it is actually quite accurate but it is something that has just gone under the radar trust me when I say there are so many different research papers out there about AI that have just gone under the radar and texture 3D is something that seems to be already near perfect based on the current technology that we do have but it's going to be interesting to see how Nvidia deploys it and I do think that they're going to deploy it through Adobe because Adobe did release something very recently that I need to show you if you haven't you can see the video on my channel tool is called Firefly this is the video that you should click it goes into detail on all the crazy stuff that they have and we're gonna go be going into some of that later on in this video too but if we move into the actual Partnerships you can see that we are going to be seeing a partnership with Nvidia and Getty Images and this one is going to be super interesting because Getty Images is something that already additionally provides a large image data set for mid-journey which is a very interesting image producing software and they've achieved photo realism within just five iterations of their software you know that this was a pressing issue for them because they were already targeted by these companies saying that you know you already trained your AI on our data and your software is amazing but you didn't give us credit so it shows us now that if the original companies are going to be training their own AIS on these different softwares just shows us now with the actual legal data sets that they're about to use it just shows us honestly how crazy this is about to get because I'm pretty sure Getty Images has a wide variety of images that we maybe have not seen and it's truly going to be very interesting to see how this is going to develop now Getty Images is not the only partnership that Nvidia is also doing it's good they're also doing it with ### [5:00](https://www.youtube.com/watch?v=dHY5LWfkYLk&t=300s) Segment 2 (05:00 - 10:00) Shutterstock now Shutterstock is going to be moving towards the 3D space and essentially a move to gain some of that ground back because of course as you know mid-journey stable diffusion Dali 2 all these platforms have kind of taken the market share because where you needed to previously be able to buy images in order to be able to get something from the internet now all you need to do is simply use an image creation software and of course Shutterstock wants some of their market share back so this will be very interesting to see exactly how this really does work and I do think from the demos it's going to be very interesting now with Adobe they've already managed to secure some crazy software stuff and we're going to get into that into a moment but for Picasso let's take a look at what they demoed and exactly how this is going to work because it truly is interesting to see how this integrates with Adobe Shutterstock and Getty Images to provide us with the next level of AI stuff so you see right here that this is the trailer it says edify generative AI Foundation models create images videos and 3D simply from Tech so this is definitely something that you do want to pay attention to so you can see right here we have this prompt and then we have this image now this is something that of course does look absolutely incredible I mean it looks photo realistic I think you know it's absolutely insane that in just a year we've come from you know not being able to do this and we're now at photo realism and of course let me just take this caveat of course it's not just one year it feels like a year because of course many years of progress lead to that exponential growth that we're currently experiencing right now with that hockey curve growth and let's move on so you can all see right here that this is where the videos come into place so the prompt that they've used is a bioluminescent jellyfish swimming in space with shining stars in the background Ultra detailed trending on Art station and the prompt that they have is truly really cool because it does look somewhat realistic now if we go back to this I just want to you know keep this on for a second because this is only the first iteration of this and you have to understand that we are in the early days of what this is now if you remember technology previously when we had the first you know video games the first mobile phones they're very clunky slow and just not that great to look at but this is already looking and shaping up to be something very interesting because this is already looking smooth it's already looking decent ad I wonder where we'll be in three years time are we going to have complete movies complete 4K footage it's going to be very interesting to see exactly where this goes because the creativity limits are going to be interesting so also what we also want to talk about which is very interesting is the 3D models now 3D modelers take so much time to create the products that they do make now if you don't understand what I'm talking about let me show you this is a video of someone speed modeling in blender and this shows you the traditional kind of methods that people use to create 3D models now of course there are different UV meshes and different textures that this person is using which are going to create a far more intricate and highly detailed model but this just goes to show how much detail there is in this space already and why these kinds of tools are going to shake up the space because this person could simply you know create these assets with a single prompt and create the textures as well because there are software out there that is being developed that will be able to do this and this is all sped up and this is something that does take a very long time to learn but if we do have something where we can simply put a 3D model of a male boss with a forward Brown and deep set eyes wearing a replica ivy leaves highly detailed and we simply get that immediately that is going to save people a lot of time so it's very interesting because of course they also do detail exactly how this software kind of does work and essentially what it runs on and it does run on the Nvidia dgx cloud so Nvidia is going to be one of the most powerful companies coming up in the future I wouldn't you know hesitate that they are going to be potentially a trillion dollar company in the future if the AI continues to develop at the rate that it does so it's definitely very interesting to see what kind of things they have because additionally on this you can all see that they have image video in 3D or they also have the ability to integrate these things into many different apps and services so not only is this base stuff good we're going to be seeing a vast majority of companies be able to actually use all of this stuff because you can see right here and we've got the client app and you're going to see many different companies be able to use their own custom data sets and build on this stuff so companies like Getty Images which we talked about and you can all see right here that this is already being built into the Creative Suite so right here what this hat if you don't know what this is Photoshop this is Premiere Pro and um yeah it's definitely truly insane to see exactly how this is going to be because for example in Premiere Pro where people would download videos you can see right here that there might just be something where you can actually have you know text to video to just quickly create a video and then import that into your project and of course you can see right here in the Omniverse which is something another thing built by Nvidia which essentially it's a realistic 3D landscape which companies build things ### [10:00](https://www.youtube.com/watch?v=dHY5LWfkYLk&t=600s) Segment 3 (10:00 - 10:00) on in order to better predict what happens in the real physical world rather than building out you know certain simulations it's going to be very interesting to see exactly how all this stuff pans out and I think that Picasso was severely underrated and not picked up by enough people so once this is going to be deployed and once come companies are going to be able to use this we're going to see a huge wave many different companies coming in and disrupting the space and if I had to make a prediction I would say that adobe is going to win because they are just already integrated with the products that we use on a day-to-day basis especially for those in a Content creation so if there was something I missed let me know in the comment section below and if there was something that you thought this video was missing --- *Источник: https://ekstraktznaniy.ru/video/14927*