This weekend in Generative Media

Alexa will use Claude AI; LAION removes CSAM; How AI is changing the 2024 election

News,

• Amazon’s new Alexa voice assistant will use Claude AI (The Verge)

• The org behind the dataset used to train Stable Diffusion claims it has removed CSAM (TechCrunch)

• How AI-generated memes are changing the 2024 election (NPR)

• The first presidential election of the generative AI era is only weeks away — and we’re not ready (Yahoo)

• Google extends election policies to most of its AI products (Axios)

• Big-name publishers are refusing to let Apple Intelligence train on data (Apple Insider)

• OpenAI called ChatGPT as a witness in a major copyright lawsuit (Business Insider)

• Character.AI lays off at least 5% of its staff, The Information reports (Reuters)

• Voice Actor Jennifer Hale on Video Game Strike and AI Fears: ‘You’re Using Technology to Take Away Our Ability to Feed Our Kids (Variety)

• Former Riot Games employees leverage generative AI to power NPCs in new video game (TechCrunch)

• Cher admits his smash hit pizza was invented by ChatGPT (Futurism)

Opinion

• In a new Weekend Essay, Ted Chiang argues that artificial intelligence can’t make real art. (X) Reactions:

  • The New Yorker piece on AI not being able to make art has some interesting bits but is otherwise full of arbitrary claims, like an elongated op-ed. (X)

  • I don't think "is it art or not" is a very interesting question (X)

  • He neglects the possibility of artists using AI to create works of unprecedented scale or scope. Instead of making an image or novel or song, an artist could make an entire universe. (X)

  • it's futile to argue that a paintbrush, because it does not experience humanity, cannot create art. (X)

  • When something generates enough steam to warrant people writing long essays about how it’s not art, it’s probably art (X)

• No more AI movie trailers, please (Mike Gioia on Substack)

Events

• Corner Conversations: The impact of AI on the creative economy (Tate Modern) 8 October 2024 at 19.00–22.30. Join Marcus du Sautoy, Massimo Sterpi, and Dr Rebecca Swift to discuss AI’s impact on art, artists, and the creative industries.

• Czech International AI Film Festival. Regular deadline: November 1, 2024. Event date: December 14, 2024.

Software

• Another Chinese 'Sora': A new AI video tool launched today by Minimax, backed by major investors Alibaba Group and Tencent. Check out their official AI film Magic Coin, created entirely with text-to-video . (X)

• Playground may be the simplest AI image generator yet (Creative Bloq)

• How Mozilla’s AI website creator, Solo, is shaking up a $2.1 billion industry (Mozilla)

• Sapiens @Gradio demo is live! (X) Sapiens: Foundation models for human tasks (HuggingFace)

Research

• Meta’s Transfusion model handles text and images in a single architecture (VentureBeat)

• MagicMan: Generative Novel View Synthesis of Humans with 3D-Aware Diffusion and Iterative Refinement (project page)

• Neural Assets: 3D-Aware Multi-Object Scene Synthesis with Image Diffusion Models (project page)

• DynaSurfGS: Dynamic Surface Reconstruction with Planar-based Gaussian Splatting (project page)

• BiGS: Bidirectional Gaussian Primitives for Relightable 3D Gaussian Splatting (arXiv)

• OctFusion: Octree-based Diffusion Models for 3D Shape Generation (arXiv)

Misc

• Nearly half of Nvidia's revenue comes from just four customers ($13.8 bn out of $30 bn). Presumably, that's Amazon, Google, Microsoft, and Meta. Or in other words: these companies own almost half of all modern GPUs, forming an oligopoly of #AI compute. (X)

• Image 1: Generated with my Influencer Lora. Image 2: Downloaded from ALO website. Image 3: Virtual try on + upscale. Conclusion? I can very likely make any ad for any clothing company in less than 5 minutes (X)

• Vegetable man (X)