Today in Generative Media

AI to earn Meta $1.4T by '35; SAG-AFTRA seeks AI protections; genAI music playlists

News and Opinion

• Meta forecasted it would make $1.4T in revenue from generative AI by 2035 (TechCrunch)

• SAG-AFTRA Chief Lays Out What AI Protections It Will Be Looking For In Next Studio Contract (Deadline)

• “No one noticed that half the playlist was AI-created, and a few even preferred the AI tracks”: Should today’s music-makers and producers fear AI - or embrace its inevitability? (MusicRadar)

• U.S. Copyright Office has registered 1,000+ works enhanced by AI (Mashable)

• Americans largely foresee AI having negative effects on news, journalists (Pew Research Center)

• Instagram Is Blocking Minors from Accessing Chatbot Platform AI Studio (404 Media)

• Is Duolingo the face of an AI jobs crisis? (TechCrunch)

• Progressive Insurance tests limits of AI-generated ads – and learns when to pull back (The Drum)

• Google has launched new film and TV production wing, Business Insider reports (Reuters)

• This data set helps researchers spot harmful stereotypes in LLMs (MIT Technology Review)

Software

• Google has tuned up its AI Music Sandbox for musicians and producers (TechRadar)

• Generate a rough cut of your video in seconds. (Clova) Clova is your AI assistant, it understands your footage and can generate rough cuts with just a prompt.

• 👋 Wan 2.1GP v4.5: FantasySpeaking model… animate a talking head using a voice track… 👋 Wan 2.1GP v4.4: Phantom model support… transfer people or objects into video… 👋 Wan 2.1GP v4.3: Added … support for Sky Reels v2 Diffusion Forcing for high quality "infinite length videos"… (GitHub)

Research

• Taming Video Diffusion Models for Panoramic 4D Scene Generation (project page)

• Predict-Optimize-Distill: Self-Improving Cycle for 4D Object Understanding (project page)

• Exploiting Deblurring Networks for Radiance Fields (project page)

• LIRM: Large Inverse Rendering Model for Progressive Reconstruction of Shape, Materials and View-dependent Radiance Fields (arXiv)

• Radiance Surfaces: Optimizing Surface Representations with a 5D Radiance Field Loss (arXiv)

• Generative AI for Character Animation: A Comprehensive Survey of Techniques, Applications, and Future Directions (arXiv)