Today in Generative Media

Authors sue Anthropic; AI comes for kids' TV; AI photo editing; 'Deepfake Elon Musk'

News

• Authors sue Claude AI chatbot creator Anthropic for copyright infringement (AP News)

• AI Is Coming for Our Kids (or at Least Their TV Shows) (Hollywood Reporter)

• The AI photo editing era is here, and it’s every person for themselves (The Verge)

• How ‘Deepfake Elon Musk’ Became the Internet’s Biggest Scammer (New York Times)

• Can You Be Emotionally Reliant on an A.I. Voice? OpenAI Says Yes. (New York Times)

• The makers of ‘Alien: Romulus’ defend their AI-resurrected Ian Holm: ‘We did it all with a lot of respect’ (LA Times)

• Under the influence (Australia ABC). As people spend more time online, bonds forged with influencers generated by artificial intelligence are rising.

• Procreate’s anti-AI pledge attracts praise from digital creatives (The Verge)

Software

• Google’s upgraded AI image generator is now available (The Verge)

• Dream Machine 1.5 is here Now with higher-quality text-to-video, smarter understanding of your prompts, custom text rendering, and improved image-to-video! Level up. (X)

• Hotshot launches new text-to-video AI generator (VentureBeat)

• New quality of life updates to the Replicate Flux lora trainer are now live. (X)

• Midjourney releases new unified AI image editor on the web (VentureBeat)

• Flux vs Midjourney: which AI image generator wins? (Tom’s Guide)

• Exists launches GenAI platform to create 3D games from text prompts (VentureBeat)

• birefnet for background removal (HuggingFace)

Research

• Photorealistic Object Insertion with Diffusion-Guided Inverse Rendering (ECCV 2024, project page)

• ControlNeXt: Powerful and Efficient Control for Image and Video Generation (project page)

• Puppet-Master: Scaling Interactive Video Generation as a Motion Prior for Part-Level Dynamics (project page)

• Generative Photomontage (project page)

• JPEG-LM: LLMs as Image Generators with Canonical Codec Representations (arXiv)

• Factorized-Dreamer: Training A High-Quality Video Generator with Limited and Low-Quality Data (arXiv)

Misc

• You can take outputs from my consistent-character model on Replicate and use them to train a flux lora, and it just works (X)

• rickrolling is deeply embedded in the training set for every LLM (X)

• Guy Who Sucks At Being A Person Sees Huge Potential In AI (The Onion)

• McDonald’s Japan just tested an AI ad in collaboration with AI artist KakuDrop and Luma Labs (X)