Today in Generative Media

Autodesk buys Wonder Dynamics; Lightroom AI eraser; Russian deepfakes love China

• Autodesk acquires AI-powered VFX startup Wonder Dynamics (TechCrunch) Autodesk buys Wonder Dynamics [interview] (FXGuide)

• Adobe Lightroom gets a magic eraser, and it’s impressive (The Verge)

• This ‘Russian Woman’ Loves China. Too Bad She’s a Deepfake. (New York Times)

• New Windows AI feature records everything you’ve done on your PC (Ars Technica)

• What We Lose When ChatGPT Sounds Like Scarlett Johansson (New York Times)

• A.I.’s Black Boxes Just Got a Little Less Mysterious (New York Times) some highlights 🧵 (X)

• Amazon plans to give Alexa an AI overhaul — and a monthly subscription price (CNBC)

• Suno has raised $125 million to build a future where anyone can make music (Suno blog)

• New AI tools to help merchants market brands and products (Google Keyword blog)

• TextureDreamer transfers textures from a few images to arbitrary 3D shapes. Excited about democratizing 3D content creation! (YouTube)

• Green World is a get-together for people who are unhealthily obsessed with or at least intensely curious about Animatediff and the open source AI art scene. [RSVP for 5/30/24, Los Angeles] (Partiful)

• Personalized Residuals for Concept-Driven Text-to-Image Generation (CVPR 2024, project page)

• Slicedit: Zero-Shot Video Editing With Text-to-Image Diffusion Models Using Spatio-Temporal Slices (ICML 2024, project page)

• MirrorGaussian: Reflecting 3D Gaussians for Reconstructing Mirror Reflections (project page)

• Gaussian Head & Shoulders: High Fidelity Neural Upper Body Avatars with Anchor Gaussian Guided Texture Warping (project page)

• FIFO-Diffusion: Generating Infinite Videos from Text without Training (project page)

• Face Adapter for Pre-Trained Diffusion Models with Fine-Grained ID and Attribute Control (project page)

• This is VastGaussian: Vast 3D Gaussians for Large Scene Reconstruction unofficial implementation (GitHub)

• Diffusion for World Modeling: Visual Details Matter in Atari (arXiv)

• Dreamer XL: Towards High-Resolution Text-to-3D Generation via Trajectory Score Matching (arXiv)

• Why did OpenAI use Scarlett Johansson's voice? As Jessa Lingel & I discuss in our journal article on AI agents, there's a long history of using white women's voices to “personalize” a technology to make it feel safe and non-threatening while it is capturing maximum data. 🧵 (X)

• We spent the last months building a companion to make sure no kid ever feels lonely again. (X)

• 🔥Phi-3 Vision Model is out! (X)

• To my best knowledge this is the only gaussian splatting based experience on standalone headsets! (X)

• "AI Generated 3D is not production ready." The AI Generated 3D: 👇🙃 (X)

• Scarlett Johansson Is All You Need (X)

• Made this cool AI Vid with AnimateDiff and Inpainting with Stable Diffusion! (X)