Yellow

Check it out—The 2025 Game Changers in Times Square! Thank you to Nasdaq for congratulating the winners on the big screen: Bitmagic, Brilliant Labs, Cartwheel, Cosmic Lounge, Eggscape Entertainment Inc, ego (YC W24), emptyvessel, Giant Skull, k-ID, Layer AI, Ludeo, Look North World, Mudstack, Noodle Cat Games, Pok Pok, Praktika.ai, Ray Browser, Regression Games, Sloyd, Stoke Games, Superbloom, Talofa, Womp, and Yellow. Cc Moritz Baier-Lentz ??, Dean Takahashi, GamesBeat

The countdown is officially over – introducing the 2025 Game Changers! For the second year in a row, we’ve partnered with GamesBeat to recognize and celebrate the 25 most innovative gaming and interactive media startups. Congratulations to these companies, who made the 2025 list and are playing a pivotal role in how we play, work, and connect: Bitmagic, Brilliant Labs, Cartwheel, Cosmic Lounge, Eggscape Entertainment Inc, ego (YC W24), emptyvessel, Giant Skull, k-ID, Layer AI, Look North World, Ludeo, Mudstack, Noodle Cat Games, Pok Pok, Praktika.ai, Ray Browser, Regression Games, Sloyd, Stoke Games, Superbloom, Talofa, Torbie, Womp, and Yellow. Learn more about the 2025 Game Changers here: https://lnkd.in/euhME8fF Cc Moritz Baier-Lentz ??, Dean Takahashi, GamesBeat

new blog! the next Pixar will be an AI company built on interactive video - a blend of film & games where frames are generated entirely by a neural net in real-time, with no code, assets, or game engines. Google (DOOM!), OpenAI, Microsoft, and many others are building here. summary ?? 1/ a recent interactive video example is DOOM simulated by Google GameNGen, where they use an image generation model to predict the next frame of the game without any of the underlying assets or code 2/ why is interactive video exciting? it blends the best of film/games - it's a video, so anyone can watch - you can lean in to play but it doesn't rely on pre-loaded assets / design, so gameplay is theoretically infinite & personalized - you could play a "1000 hour video" 3/ interactive video became feasible recently with advances in model distillation for image generation. in just the last year, we saw generation speeds improve by ~20x & efficiency by >30x. the chief remaining hurdle is reaching 30 FPS - which we estimate to be 1-2 years out 4/ what could all this unlock? interactive video could be a huge new storytelling format on its own. a few precedents: - Telltale's The Walking Dead, a choose-your-own-adventure game with >28M copies sold - Netflix's Black Mirror: Bandersnatch, a critically acclaimed live action film with 5 endings 5/ prior attempts like TWD & Bandersnatch faced an intractable problem - content costs scaled linearly with hours of gameplay. interactive video generated by diffusion models could unlock this storytelling format for scale 6/ interactive video could also be huge for vertical shorts - apps like Reelshort have >30M downloads and $10M+ monthly revenue, delivering 1-2min soap operas for a viewer's particular tastes (ex. werewolf billionaire love story!) 7/ many of our beloved stories from Disney, Marvel, Pixar etc were enabled by new technologies. similarly, we believe the next Pixar will arise through a new uniquely AI-enabled format - interactive video 8/ Pixar's success came from both human creativity & new technology. John Lasseter, Disney, & a team of storytellers created incredible films. similarly, we believe the next Pixar will need to be both a world-class interactive storytelling studio as well as a top AI company 9/ the team will also need to navigate legal / regulatory / ethical challenges around training data & AI use. but for those who get it right, the opportunity is huge. one day, the AI Pixar could create not just 1-off stories but entire virtual worlds - blurring the lines between traditional storytelling formats much thanks to Neel Jain for his research & contributions. blog below: https://lnkd.in/gsRNpkjb

Today A16Z GAMES is thrilled to lead the $5M seed round for Yellow, who's revolutionizing 3D character creation with generative AI and is developed by a team of top academics. Despite rapid progress in 2D image generation, 3D has remained stubbornly difficult and expensive due to challenges like maintaining clean topology (e.g. water-tight meshes) and requiring numerous intermediate representations (meshes, UV maps, textures, rigs). This complexity makes high-quality 3D incredibly time- and skill-intensive, with top-tier character models taking weeks and costing upwards of $10K. Generative AI hopes to democratize this process by reducing the costs and skill needed to create commercial quality characters. When we first met Yellow's founder, CEO Mandeep Waraich, we were astounded by his infectious energy and drive.. As the former Head of CoreML and Large Models at Google, Mandeep's deep learning expertise and passion for the arts made him the ideal fit for this challenge. Yellow is also developed alongside a team of top academics like: Vincent Sitzmann, from MIT's Scene Representation Group, Christian Rupprecht, who's an Associate Professor at Oxford, and Ludwig Schmidt, who's an acting Assistant Professor in Stanford's Computer Vision Group. What sets Yellow apart is their focus on empowering human creativity through AI. Rather than aiming for pure automation, they're building tools that amplify today's 3D artists while making character creation accessible to all. So, we're thrilled to support Yellow on their journey. You can find: - their website here: yellow3d.com - Their partnership with Tafi: https://lnkd.in/eVUxJtgr - Or our blog post: https://lnkd.in/eebqXpUi The future is bright... and Yellow ??