Today in Generative Media

Court finds AI is not fair use; Gemini watches YouTube, LeCun: Don't ban AI sharing

News and Opinion

• Thomson Reuters Wins First Major AI Copyright Case in the US (Wired)

• Some Good News for Hollywood Creators Suing AI Companies (Hollywood Reporter)

• Gemini can now watch YouTube for you - skip the video, get the highlights (ZDNet)

• Don't ban AI researchers from sharing their models or you'll fall behind, Meta's head of AI warns Europe (Business Insider)

• Why Do AI Chatbots Have Such a Hard Time Admitting ‘I Don’t Know’? (Wall Street Journal)

• AI Agents Are Now Trading IP Rights With Each Other—And Earning Crypto for Their Owners (Decrypt) Story Protocol's new framework lets AI agents autonomously buy and sell IP, in what could be the first working example of an AI-to-AI economy.

• US and UK refuse to sign AI safety declaration at summit (Ars Technica)

• Can A.I. Heal Our Souls? (New York Times)

Software

• Napkin AI’s ‘design agency’ of AI agents is changing how professionals create graphics (VentureBeat)

• Image to Video with Ray2 is here (LinkedIn)

• Introducing Intangible Studio (YouTube)

• Lumina 2.0 is a pretty solid base model, it's what we hoped SD3/3.5 would be, plus it's truly open source with Apache 2.0 license. (X)

• Transform your hardware design With Zoo Dev's Text-to-CAD, turn simple text prompts into detailed designs instantly. (X)

• google / imagen-3: Google's highest quality text-to-image model, capable of generating images with detail, rich lighting and beauty (Replicate)

• Zonos-v0.1 is a leading open-weight text-to-speech model trained on more than 200k hours of varied multilingual speech, delivering expressiveness and quality on par with—or even surpassing—top TTS providers. (HuggingFace)

• CUT3R: Continuous 3D Perception Model with Persistent State (GitHub)

Research

• Goku: Flow Based Video Generative Foundation Models (project page)

• History-Guided Video Diffusion (boyuan.space)

• SC-OmniGS: Self-Calibrating Omnidirectional Gaussian Splatting (project page)

• AuraFusion360: Augmented Unseen Region Alignment for Reference-based 360° Unbounded Scene Inpainting (project page)

• Fillerbuster: Multi-View Scene Completion for Casual Captures (project page)

• HumanDiT: Pose-Guided Diffusion Transformer for Long-form Human Motion Video Generation (project page)

• Imagine while Reasoning in Space: Multimodal Visualization-of-Thought (arXiv)

Misc

• Welcome to the 🤗 AI Agents Course (HuggingFace)

• Our Creative Lab team is looking for the next group of Fivers in NYC + London! Writers, visual + motion designers, creative coders, filmmakers, especially people who are excited to prompt and explore the latest AI capabilities (LinkedIn)

• VEO2 → Ridiculous. Nailed this abstract idea. (X)

• If you’re the CEO of the alphabet you should make sure to include all 26 letters (X)