Today in Generative Media

Meta's real-time genAI images; Thaler fights for genAI copyright; AI hype deflation

• Meta is adding real-time AI image generation to WhatsApp (The Verge)

• Thaler, Copyright Office Fight Over Human-Authorship Requirement for AI-Created Artwork Continues (IP Watchdog)

• The AI hype bubble is deflating. Now comes the hard part. (Washington Post)

• A.I. Has a Measurement Problem (New York Times)

• HarperCollins Publishers and ElevenLabs to Bring More Stories to Life Through Audio (ElevenLabs blog)

• Microsoft's new AI tool is a deepfake nightmare machine (Creative Bloq)

• Peter Thiel says AI will be 'worse' for math nerds than for writers (Business Insider)

• Adobe Express brings generative AI to photo and video on Android (9to5Google)

• Netflix Accused of Using AI Photos in True Crime Documentary (PetaPixel) A few ethics-related issues, among them: Fact vs Fiction: This is rewriting history. Minimally, needs mechanisms in place (C2PA and/or watermarking) so future handling of GenAI images can be marked as such. Consent: Likeness edited w/o consent. Disclosure: People need to know this is fake nowish. (X)

• Meet the first AI with emotional intelligence (Hume.ai)

• Imagine Flash: Accelerating Emu Diffusion Models with Backward Distillation (project page) Available at meta.ai.

• Dynamic Gaussian Mesh: Consistent Mesh Reconstruction from Monocular Videos (project page)

• AniClipart: Clipart Animation with Text-to-Video Priors (project page)

• We're excited to release Text-to-Animation for production-grade meshes! 🎉 (X)

• Magic Clothing is a virtual try-on custom node for ComfyUI that works with IPAdapter FaceID & ControlNet Openpose (X)

• I spent a bit of time using Udio, and it turns out a bunch of its output closely resembles copyrighted music. (X)

• she used “delve” in her texts twice jerry. TWICE! (X)