DCVC’s Post

Super excited to share what I’ve been working on: Odyssey! We’re building Hollywood-grade, directable visual AI, to enable storytellers to create new, amazing movies, TV shows, and video games. We believe this technology is what comes after text-to-video. Firstly, let’s talk about what has shaped our technology approach more than most: autonomous vehicles (AV). Myself and Oliver Cameron, plus a large chunk of our team from Cruise, Wayve, Waymo, and NVIDIA, have spent years building AV technology. Why is AV relevant? 1️⃣ AV had to push the SOTA in visual AI to ship. 2️⃣ AV scaled performance with real-world data. 3️⃣ AV necessitated scalable, physics-accurate simulation systems. AVs are taught to navigate the world. Odyssey is taught to generate worlds. After pushing the SOTA in visual AI for many years, and watching all the progress in generative ML, we were confident that Hollywood-grade outputs could be generated by AI. The question was then: what abstractions would make this technology useful for movie-making and more? If I told you ChatGPT output text you couldn’t edit, you wouldn’t use it. But that’s how text-to-video works today, where the model output is a fully-baked video file that you can’t edit. If any of the scenery, characters, lighting, or motion is off, you have to start again. Given our experience in AVs, our minds naturally went to 3D, and generating layers of 3D worlds. This is why we’re training models to generate… 🧊 High-quality geometry 🏔️ Photorealistic materials 🌄 Stunning lighting 🏃♀️ Controllable motion With this approach, our users will still be able to generate video at the speed of AI, but have fine-tuned control over all the minutiae that matter. We’re early on our journey, but the progress has been incredible, and we’re ready to work alongside Hollywood to shape this. Please get in touch! I’m fortunate to work with some world-class folks to bring this to life. We’re now growing our R&D team for the right individuals. If you’re excited about working with the best visual AI team in the industry, check out our jobs section. Lastly, a big thank you to our amazing supporters in GV (Google Ventures), Air Street Capital, DCVC, and our many wonderful angels. https://odyssey.systems

The difference that we have seen in the quality of ai videos 2 years ago is mind-blowing .The AI industry is growing at an exponential rate, the difference between AI-generated videos and higher-quality videos we have today primarily lies in the level of detail, realism, and overall quality. 1. Level of Detail and Realism:   - AI-generated videos, especially those created using deep learning techniques, may lack the level of detail and realism seen in higher-quality videos. While AI can generate images and videos that resemble real scenes to some extent, they may still exhibit artifacts, distortions, or inconsistencies that make them appear less realistic.   - Higher-quality videos produced by professional filmmakers or studios typically involve sophisticated camera equipment, lighting setups, and post-production techniques to capture and enhance visual details, resulting in more lifelike and immersive viewing experiences. 2. Resolution and Clarity:   - AI-generated videos may have limitations in terms of resolution and clarity. Depending on the training data and algorithms used, AI-generated content may exhibit lower resolution, pixelation, or blurriness, especially when dealing with complex scenes or fast-paced action.   - Higher-quality videos often boast higher resolutions (such as 4K or even 8K) and superior clarity, allowing viewers to discern fine details and textures with greater precision. 3. Color Accuracy and Dynamic Range:   - AI-generated videos may struggle with color accuracy and dynamic range, leading to unnatural-looking colors or limited tonal variation in the footage.   - Higher-quality videos are often meticulously color graded and mastered to achieve accurate colors, rich contrast, and a wide dynamic range, resulting in visually stunning imagery that closely resembles real-world scenes. 4. Artistic Direction and Cinematic Techniques:   - AI-generated videos may lack the artistic direction and cinematic techniques employed by skilled filmmakers. While AI algorithms can mimic certain styles or genres based on training data, they may not possess the creativity, intuition, and storytelling prowess of human directors.   - Higher-quality videos benefit from the creative vision and expertise of filmmakers who leverage cinematic techniques such as framing, composition, camera movement, and editing to convey emotions, narratives, and themes effectively. In summary, while AI-generated videos have made significant advancements in recent years and can produce impressive results, they may still fall short in terms of realism, quality, and artistic expression compared to higher-quality videos produced by professional filmmakers and studios. #ai #aivideo #artificialintelligence #computerscience #fulfillment

බහුවිධ මාන-multidimensional snap | SOUL AI Model by ABSOL X CORE AI Introducing SOUL AI ABSOL-X CORE AI's SOUL AI Model (beta): Revolutionizing Visual Technology with Hyper-Realistic, Multidimensional Dreamscapes Experience the power of ABSOL-X CORE AI's SOUL Ai Model – a revolutionary generative platform that transforms the way we create visuals. With its deep learning capabilities, ABSOL-X CORE AI integrates seamlessly into filmmaking, creative design, and advertising workflows. This empowers artists, directors, and brands to push the limits of visual storytelling, generating hyper-realistic images and environments that blur the line between the digital and the real. At the core of ABSOL X CORE AI lies a sophisticated neural network meticulously trained on a vast repository of real-world imagery. This deep understanding of light, shadow, texture, and micro-motion ,depth , object ,Human's ,environment enables the SOUL AI to generate hyper-realistic visuals that are indistinguishable from reality, seamlessly replacing traditional VFX and CGI techniques. unlocking a world of creative possibilities and transforming dreamscapes into tangible realities. ABSOL X CORE AI's impact extends beyond filmmaking, revolutionizing the creative and advertising industries. can leverage the AI's generative power to create stunning visuals that capture attention and evoke emotion. In the advertising world, ABSOL X CORE AI empowers brands to craft hyper-realistic product renders and captivating ad campaigns that leave a lasting impression. Transcend the limitations of traditional VFX and CGI, crafting breathtaking worlds and dreamlike scenarios with unparalleled realism. Embark on a journey beyond reality with ABSOL X CORE AI, the AI that empowers you to unleash your boundless imagination. WHERE IMAGINATION KNOWS NO BOUNDS music credit - @wymjayss 2024©️ABSOL-X 2019-2024©️ABSOLX_Core_AI #ai

Pika Labs’ text-to-video AI platform opens to all: Here’s how to use it | VentureBeat: Consider sponsoring The AI Impact Tour to bring awareness to your brand. Pika Labs has launched Pika 1.0, a web-based text-to-video AI platform allowing users to generate and edit videos in various styles. While it showed inconsistency during testing, it offers extensive customization options. Users can join the waitlist and get access within minutes. Pika competes with Adobe, Runway, and Stability AI in the AI video space. - Artificial Intelligence topics! #ai #artificialintelligence #intelligenzaartificiale

Odyssey, a new startup co-founded by self-driving car industry veterans, just emerged from stealth with a ‘Hollywood-grade’ AI video generation platform aimed at revolutionizing visual effects and storytelling capabilities. 💡 Odyssey is developing four specialized AI models focused on geometry, materials, lighting, and motion, aimed at giving more control over outputs. 💡 The company's team includes AI researchers from tech giants, as well as artists who've worked on blockbusters like Dune and Avengers. 💡 Odyssey is reportedly gathering its own 3D, real-world data for training instead of scraping existing video data, building on learnings from self-driving AI experience. While we’ve already seen Hollywood start to experiment with AI video platforms like Sora, the capabilities (at least publicly) seem a bit off from longer Hollywood-type outputs. Odyssey’s ambitions could accelerate the tech’s use on the big screen, video games, and other media — but it certainly will have some stiff competition as other models improve. #ai #technology https://odyssey.systems/

🚀 Exciting News in AI! 🤖 Have you heard about Animate Anyone? This innovative project from Alibaba can turn static images into captivating videos. Imagine the possibilities for marketing and storytelling! Stay tuned for more updates. #AI #Innovation

😲 New mindblowing AI: Revolution in Visual Storytelling! 💽 We are seeing a lot of new AI tools this year, but every once in a while, something game-changing emerges. Just discovered an AI that turns TEXT into fully-fledged COMIC PAGES with story and optional captions even. This is not just technology — it's art and innovation combined! 🎨 No artist, no wait time. Instant creativity. Enter prompt and click to see a sample. Truly, a marvel of AI advancement. Imagine its applications in education, marketing, entertainment, and more! 💡 If you're into storytelling, content creation, or just love cool tech, this is a MUST CHECK. Post your generated comic page in the comments below AI Comic Factory (FREE) on Hugging Face: https://lnkd.in/d2FyrveR My prompt: "A regular monday at the office with AI robots" Style: "American (modern)" Design: "Layout 1" #AI #Comics #Innovation #TechTrends #ContentCreation