This weekend in Generative Media

Japan's AI copyright rules; Midjourney evades copyright cops; Lego regrets AI images

• Japan’s New Draft Guidelines on AI and Copyright: Is It Really OK to Train AI Using Pirated Materials? (National Law Review)

• This Week in AI: Midjourney bets it can beat the copyright police (TechCrunch)

• Lego says its use of AI-generated images was a mistake (Axios)

• Google brains plumb depths of the uncanny valley with latest image-to-video tool (The Register)

• Kids are watching brain melting AI-generated videos on YouTube without parents realizing. (Futurism)

• I, Cyborg: Using Co-Intelligence (Ethan Mollick on Substack)

• The exponential enshittification of science (Gary Marcus on Substack)

• Video Editing via Factorized Diffusion Distillation (project page)

• StereoDiffusion: Training-Free Stereo Image Generation Using Latent Diffusion Models (arXiv)

• What if you could easily translate dialogue into different languages? Check out this sneak preview from Adobe Research. (X)

• Introducing: Transparent Generations! Real transparency, not just background removal. Now possible with Leonardo AI. (X)

• We are excited to preview a breakthrough in 3D generative AI. Video models are incredible but do not directly allow for controllability and simulation. We are closing this gap by predicting 4D meshes from video inputs. (X)

• There's a new lip-syncing feature coming to RunwayML. (X)

• don't fall victim to the "Lump of labor fallacy". There is not a fixed amount of engineering work. (X)

• Working with AI making films makes you a better director 10 fold almost guaranteed because of how many roles you juggle doing it yourself, which is awesome. (X)

• SXSW audiences were booing and screaming at pro-AI videos (X)