Today in Generative Media
OpenAI's GPT-4o; Web braces for Google's AI answers; TikTok will label 3P AI content
OpenAI releases GPT-4o, a faster model that’s free for all ChatGPT users (The Verge), OpenAI Unveils New ChatGPT That Listens, Looks and Talks (The New York Times), Hot take on OpenAI’s new GPT-4o (Gary Marcus on Substack)
Web publishers brace for carnage as Google adds AI answers (Washington Post)
TikTok will automatically label AI-generated content created on platforms like DALL·E 3 (TechCrunch)
Illness took away her voice. AI created a replica she carries in her phone (AP News)
Cinema Synthetica Contest Challenges Filmmakers to Make AI-Created Shorts in 48 Hours (Variety)
Why Adobe CEO Shantanu Narayen is confident we’ll all adapt to AI (The Verge)
DeviantArt and Midjourney deny wrongdoing in copyright infringement lawsuit over in AI image generators (The Art Newspaper)
Hyper-SD and Hyper-SDXL fast models [Tutorial] (Stable Diffusion Art blog)
Gaussian Splatting SLAM [accelerated code] (GitHub)
HOLD: Category-agnostic 3D Reconstruction of Interacting Hands and Objects from Video (CVPR 2024, project page) Code on GitHub.
Rip-NeRF: Anti-aliasing Radiance Fields with Ripmap-Encoded Platonic Solids (SIGGRAPH 2024, project page)
RPBG: Towards Robust Neural Point-based Graphics in the Wild (arXiv)
Direct Learning of Mesh and Appearance via 3D Gaussian Splatting (arXiv)
This is hands down the best LLM refusal I've ever gotten and it was from a base model. (X)