🥟 Chao-Down #341 The AI behind Apple Intelligence, Apple makes a privacy-first commitment towards user data, Google increasingly uses AI to answer user health questions
Plus, Meta seeks data from users in privacy-conscious Europe to train its AI model
What is the AI behind “Apple Intelligence?”
Apple had its big developer day at its Worldwide Developer’s Conference and introduced Apple Intelligence, an integrated personal intelligence system available in iOS 18, iPadOS 18, and macOS Sequoia. The platform leverages generative models, including a ~3 billion parameter on-device language model and a more extensive server-based model, to enhance user experiences across writing, summarization, image creation, and app interactions.
Developed using the AXLearn framework, Apple Intelligence prioritizes privacy and responsible AI practices by relying on high-quality, licensed data. The system also optimizes for efficiency and speed and according to Apple has been rigorously evaluated for performance and safety.
Read their full report here.
-Alex, your resident Chaos Coordinator.
What happened in AI? 📰
Here’s everything Apple announced at the WWDC 2024 keynote, including Apple Intelligence, Siri makeover (TechCrunch)
Apple’s AI promise: “Your data is never stored or made accessible to Apple” | (Ars Technica)
Facebook owner Meta seeks to train AI model on European data as it faces privacy concerns (ABC News)
Google Is Using A.I. to Answer Your Health Questions. Should You Trust It? (The New York Times)
Euro banks worry AI will increase dependence on US big tech (The Register)
Who will make AlphaFold3 open source? Scientists race to crack AI model (Nature)
Always be Learnin’ 📕 📖
Cognitive Biases You Don’t Think About - by Deb Liu (Substack)
A Grand Unified Theory of the AI Hype Cycle (Deciphering Glyph)
Securing Research Infrastructure for Advanced AI (OpenAI)
Projects to Keep an Eye On 🛠
andyk/ht: headless terminal - wrap any binary with a terminal interface for easy programmatic access. (Github)
xenova/transformers.js: State-of-the-art Machine Learning for the web. Run 🤗 Transformers directly in your browser, with no need for a server (Github)
abgulati/LARS: An application for running LLMs locally on your device, with your documents, facilitating detailed citations in generated responses. (Github)
The Latest in AI Research 💡
Introducing Apple’s On-Device and Server Foundation Models (Apple Machine Learning Research)
Scalable Extraction of Training Data from (Production) Language Models (arxiv)
Automatic Data Curation for Self-Supervised Learning: A Clustering-Based Approach (arxiv)
The World Outside of AI 🌎
As China’s Internet Disappears, ‘We Lose Parts of Our Collective Memory’ (The New York Times)
Study finds 1/4 of bosses hoped RTO would make staff quit (The Register)
A look at what could be the future for postpartum care in America (NPR)
Why alarm clock design speaks about humanity (UX Collective)
New York Set to Restrict Social-Media Algorithms for Teens (MSN)
Entry-Level Workers Struggle to Get Jobs, Senior Staff Misses Promotions (Bloomberg)
One Last Bite 😋
