- Chaos Theory
- Posts
- 🥟 Chao-Down #165 Apple spending millions each day to improve Siri, Colleges wrestle with AI-generated admission essays, HuggingFace launches AI training cluster as a service
🥟 Chao-Down #165 Apple spending millions each day to improve Siri, Colleges wrestle with AI-generated admission essays, HuggingFace launches AI training cluster as a service
Plus, Microsoft suspects Chinese operatives are using AI-generated images to spread disinformation among US voters.
In the battle of AI and copyright law, OpenAI has thrown down their initial arguments for why ChatGPT should be exempt from copyright infringement.
Several authors like Sarah Silverman Paul Tremblay, and others have alleged that ChatGPT was illegally trained on pirated copies of their books. OpenAI states that its false to say that every ChatGPT output represents a “derivative work” regardless of any similarities between the responses and the original texts.
And citing a previous copyright case against Google Books, OpenAI points out that “while an author may register a copyright in her book, the 'statistical information' pertaining to 'word frequencies, syntactic patterns, and thematic markers' in that book are beyond the scope of copyright protection.”
It’ll be interesting to see where all this lands between authors and AI providers. What do you think?
-Alex, your resident Chaos Coordinator.
What happened in AI? 📰
Artificial intelligence: Suspected Chinese operatives using AI generated images to spread disinformation among US voters, Microsoft says (CNN Politics)
Ban or Embrace? Colleges Wrestle With A.I.-Generated Admissions Essays. - The New York Times (nytimes.com)
Apple spending "millions of dollars a day" on developing conversational AI as it seeks to massively improve Siri (superchargednews.com)
OpenAI disputes authors’ claims that every ChatGPT response is a derivative work (Ars Technica)
Training Cluster as a service: Train your LLM at scale on our infrastructure (huggingface.co)
Always be Learnin’ 📕 📖
How Google Stores 10,000 Petabytes of Data Efficiently (devmoh.co)
Understanding Dynamic Programming So You Can Use It Effectively (HackerNoon)
How to build an enterprise LLM application: Lessons from GitHub Copilot - (The GitHub Blog)
Projects to Keep an Eye On 🛠
cofactoryai/textbase: ✨ Textbase is a simple framework for building AI chatbots. ✨ (Github)
kuafuai/DevOpsGPT: Multi agent system for AI-driven software development. Combine LLM with DevOps tools to convert natural language requirements into working software. (Github)
haoheliu/AudioLDM2: Text-to-Audio/Music Generation (Github)
The Latest in AI Research 💡
Improving Few-shot Image Generation by Structural Discrimination and Textural Modulation (arxiv)
Dynamic Low-Rank Instance Adaptation for Universal Neural Image Compression (arxiv)
Peering Through Preferences: Unraveling Feedback Acquisition for Aligning Large Language Models (arxiv)
The World Outside of AI 🌎
The room-temperature superconductor that wasn’t | (Ars Technica)
Lab-grown diamonds are pushing down prices (qz.com)
The End of Airbnb in New York (WIRED)
Disney CEO mess: Inside story of Iger and Chapek (CNBC)
Reddit faces content quality concerns after its Great Mod Purge (Ars Technica)
US lawmakers are pushing Google to restructure amid monopoly concerns (androidpolice.com)