This weekend in Generative Media

OpenAI courts Hollywood; Stability AI CEO resigns; Photography is no longer evidence

• OpenAI Courts Hollywood in Meetings With Film Studios, Directors (Bloomberg)

• Stability AI CEO resigns to ‘pursue decentralized AI’ (The Verge) Stability AI, Makers of Stable Diffusion, Could Be in Huge Trouble (PetaPixel) Stability AI Announcement (Stability)

• Photography Is No Longer Evidence of Anything (Wired)

• Arizona newsletter makes Kari Lake deepfake (The Hill)

• Financial Times tests an AI chatbot trained on decades of its own articles (The Verge)

• The Top 100 Gen AI Consumer Apps (a16z)

• Fully AI-generated games by 2030? That’s what Nvidia’s CEO believes – but what exactly will that mean for PC gamers? (TechRadar)

• How AI Can Help Us End Design Education Anachronisms (Arch Daily)

• Create Visual Stories with Consistent Characters and Scenes using Generative AI (Katalyst)

• Turn your words into videos. Easily create videos in minutes with your own on demand digital actors. You do the typing, they’ll do the talking. (Pipio)

• AI Music Generator for Commercial Use with Rightsify's Hydra. Perfect for Businesses, Creators and Artists – Generate Unique, Copyright-Clear Music Instantly (Rightsify)

• Create breath-taking videos with AI (PixVerse)

• AI for Dreamers: Make Games, Make Money Using AI (BuildBox)

• Controllable Video Generation: Powered by JST-1, the first video-3d foundation model with actual physics understanding, starting from making any character move as you want. (Viggle.AI) Made this Viggle dialogue test in about 20 minutes on my mobile in the cab ride home from work. (X)

• StableSwarmUI 0.6.2 Beta: A Modular Stable Diffusion Web-User-Interface, with an emphasis on making powertools easily accessible, high performance, and extensibility. (GitHub)

• Bilateral Reference Network achieves SOTA result in multi Salient Object Segmentation dataset, this repo pack BiRefNet as ComfyUI nodes, and make this SOTA model easier use for everyone. (GitHub)

• DynamiCrafter: Animating Open-domain Images with Video Diffusion Priors (HuggingFace)

• LightIt: Illumination Modeling and Control for Diffusion Models (CVPR 2024, project page)

• CMD: Efficient Video Diffusion Models via Content-Frame Motion-Latent Decomposition (ICLR 2024, project page)

• Explorative Inbetweening of Time and Space (project page)

• OMG: Occlusion-friendly Personalized Multi-concept Generation In Diffusion Models (project page)

• Diffusion Models are Geometry Critics: Single Image 3D Editing Using Pre-Trained Diffusion Priors (project page)

• You Only Sample Once: Taming One-Step Text-To-Image Synthesis by Self-Cooperative Diffusion GANs (arXiv)

• If this is true OpenAI really are closed AI, peddling their model to the highest bidder. Preaching an egalitarian future while ensuring the technology is in the hands of big business alone. (X)

• Sam Altman told Lex Friedman that "people have got to get paid", but there's a lot more to unpack in the full quote. (X)

• AI girlfriends are a new normal (X)