Today in Generative Media

AI pioneers win Nobel; Publishers open up to AI scrapers; Cara: the anti-AI platform

News

• Nobel Physics Prize Awarded for Pioneering A.I. Research by 2 Scientists (New York Times)

• The Race to Block OpenAI’s Scraping Bots Is Slowing Down (Wired)

• The rise of Cara: the anti-AI social media platform for artists (Creative Bloq)

• The “Whisperverse”: The future of mobile computing is an AI voice inside your head (Big Think)

• AI expert Gary Marcus thinks OpenAI will be the 'most Orwellian company of all time' (Business Insider)

• An AI chatbot of a girl murdered in 2006 was created on a hugely popular service — and her family had no idea (Business Insider)

Software

• DepthCrafter for Nuke allows you to generate temporally consistent Depth sequences inside Nuke (GitHub)

• openai-gradio is a Python package that makes it very easy for developers to create machine learning apps that are powered by OpenAI's API. (GitHub)

• FacePoke (HuggingFace)

  • This AI demo is the most mind-blowing thing you'll experience this week! (X)

• Depth Pro with Depth Flow now on tost.ai , runpod.io and ComfyUI (X)

• LoRA Lab [beta] Mix and match any FLUX[dev] LoRAs (HuggingFace)

Research

• Inverse Painting: Reconstructing The Painting Process (SIGGRAPH Asia 2024, project page)

• MonST3R: A Simple Approach for Estimating Geometry in the Presence of Motion (project page)

• EdgeRunner: Auto-regressive Auto-encoder for Artistic Mesh Generation (project page)

• Posterior-Mean Rectified Flow: Towards Minimum MSE Photo-Realistic Image Restoration (project page)

• VideoGuide: Improving Video Diffusion Models without Training Through a Teacher's Guide (project page)

• GS-VTON: Controllable 3D Virtual Try-on with Gaussian Splatting (project page)

• PhotoReg: Photometrically Registering 3D Gaussian Splatting Models (project page)

• ACE: All-round Creator and Editor Following Instructions via Diffusion Transformer (project page)

Tutorials

• AI Video Hackathon: Automating a Faceless YouTube Channel with SwarmZero (YouTube)

• Flow Matching: Simplifying and Generalizing Diffusion Models | Yaron Lipman (YouTube)

Misc

• I think the “you won’t be replaced by Ai, you will be replaced by a human using Ai” line is wildly short sighted. What will end up mattering is if you built public human connection to distinguish you from automation. (X)

• google ai LOVES to step in when it sees “vs” and will try its hardest to compare whatever you throw at it (X)

• The Internet, RIP. Well done guys. (X)

• Using 'Facepoke' as input for Runway's video-to-video, it's clear we're entering a new era of creation. (X)

• Our GPU organ plays music acoustically by controlling the RPM of each fan. The Call is now open to the public until February at Serpentine in Hyde Park London (X) It plays along with music generated from a diffusion model that creates new songs based on the (volunteered) training data of choirs from across the UK.