🥟 Chao-Down #342 A new $1 million prize for AGI progress, Latest study shows AI training could soon run out of human-written text, Apple reclaims throne as most valuable company after AI-fueled WWDC

Plus, how game theory can make AI more reliable.

Jun 12, 2024

How do the top AI chatbots perform against one another?

The Wall Street Journal tested the top 5 chatbots on a range of useful everyday skills and found that it’s not a clear-cut answer, with a different model coming out on top for different tasks. Overall, Perplexity scores more consistently in the highest while Microsoft’s Copilot comes out in last (except for one unintuitive area: creative writing).

Do you agree with the WSJ’s ranking?

-Alex, your resident Chaos Coordinator.

What happened in AI? 📰

AI-powered Apple overtakes Microsoft as world's most valuable company (MSN)

How Game Theory Can Make AI More Reliable (WIRED)

ARC Prize – a $1M+ competition towards open AGI progress (Hacker News)

AI 'gold rush' for chatbot training data could run out of human-written text (AP News)

California Proposes 30 AI Regulation Laws Amid Federal Standstill (The New York Times)

Elon Musk drops suit against OpenAI and Sam Altman (CNBC)

Always be Learnin’ 📕 📖

What are AI agents? Types, Benefits, & Use Cases (ampcome.com)

Generative AI Is Not Going To Build Your Engineering Team For You (Stack Overflow)

Say Hello to My New AI Marketer: How Gen AI-Based Software Is Advancing Marketing and Sales (a16z.com)

Projects to Keep an Eye On 🛠

apple/axlearn: An Extensible Deep Learning Library (Github)

huggingface/lerobot: 🤗 LeRobot: End-to-end Learning for Real-World Robotics in Pytorch (Github)

fchollet/ARC-AGI: The Abstraction and Reasoning Corpus (github.com)

The Latest in AI Research 💡

CRAG -- Comprehensive RAG Benchmark (arxiv)

CARES: A Comprehensive Benchmark of Trustworthiness in Medical Vision Language Models (arxiv)

Learning to Route Among Specialized Experts for Zero-Shot Generalization (arxiv)

The World Outside of AI 🌎

X is about to start hiding all likes - The Verge

Can Apple Rescue the Vision Pro? (The New York Times)

Workers over 50 most likely to be fully in-office or remote (qz.com)

Elite researchers in China say they had ‘no choice’ but to commit misconduct (Nature)

Women may be more resilient than men to stresses of spaceflight, says study (The Guardian)

Why Seasonal Allergies Are Getting Worse (Bloomberg)