Today in Generative Media

Adobe and Google make AI music; More publishers sue OpenAI; Write a book with AI

• Adobe’s new prototype generative AI tool is the ‘Photoshop’ of music-making and editing (The Verge)

• Is AI coming for your DAW? Google's MusicFX DJ doesn't just generate music - it mixes multiple parts together, too. (MusicRadar)

• More News Publishers Sue OpenAI Over Copyright Claims As Legal Strategy Shifts (Hollywood Reporter)

• AI 'writers' are mostly worthless right now. But that won't last. (Mashable)

• Diffusion transformers are the key behind OpenAI’s Sora — and they’re set to upend GenAI (TechCrunch)

• State TV airs China’s first AI-developed cartoon series, Qianqiu Shisong, as Sora sets off frenzy for text-to-video generation tech (South China Morning Post)

• Lightricks presents LTX Studio: Storytelling Transformed. Control every aspect of your video using AI, from ideation to final edits, on one holistic platform. (ltx.studio)

• Playground v2.5 is the state-of-the-art open-source model in aesthetic quality. (Playground blog)

• EMO: Emote Portrait Alive - Generating Expressive Portrait Videos with Audio2Video Diffusion Model under Weak Conditions (project page)

• Seeing and Hearing: Open-domain Visual-Audio Generation with Diffusion Latent Aligners (project page)

• Sora Generates Videos with Stunning Geometrical Consistency (project page)

• Holoported Characters: Real-time Free-viewpoint Rendering of Humans from Sparse RGB Cameras (project page)

• VastGaussian: Vast 3D Gaussians for Large Scene Reconstruction (project page)

• Outline-Guided Object Inpainting with Diffusion Models (arXiv)

• Between Copyright and Computer Science: The Law and Ethics of Generative AI (SSRN)

• AnimateLCM Stable Video Diffusion now working in ComfyUI (X)

• I uploaded the source code for the text adventure game, Zork, and asked Gemini 1.5 to play it with me. The model walked me through the game play -- but also made changes and modifications based on the questions I asked. (X)

• Controversial opinion: if we end up in a system where you can only train good AI models on $$ licensed data, there's a massive risk of concentration of power. It might not be the users, artists, or content creators who will benefit from this but big companies and Hollywood studios who will trade their rights and not redistribute. (X)