• Chaos Theory
  • Posts
  • 🥟 Chao-Down #104 DeSantis campaign may have used AI images, Researchers find ChatGPT repeats 25 jokes over and over, Congress considers bills on transparency and US competitiveness in AI

🥟 Chao-Down #104 DeSantis campaign may have used AI images, Researchers find ChatGPT repeats 25 jokes over and over, Congress considers bills on transparency and US competitiveness in AI

Plus, Bloomberg investigates the stereotypes and biases of generative AI

Boston Consulting Group just conducted a survey of 13,000 workers about their thoughts on AI in the workplace and they found that people’s perceptions of the technology depended on where they sat in the organization’s hierarchy.

Is AI creating a divide between those at the top and those at the bottom? For the sake of organizations, leaders should look to fill the gap.

-Alex, your resident Chaos Coordinator.

What happened in AI? 📰

DeSantis Campaign Uses ‘Deepfake’ Images to Attack Trump, Experts Suggest (The New York Times)

U.S. Congress to consider two new bills on artificial intelligence (Reuters)

Researchers discover that ChatGPT prefers repeating 25 jokes over and over (Ars Technica)

Hey, Alexa, What Should Students Learn About A.I.? (The New York Times)

Generative AI Takes Stereotypes and Bias From Bad to Worse (Bloomberg)

Always be Learnin’ 📕 📖

Understanding GPT tokenizers (simonwillison.net)

Your guide to SaaS metrics 2.0 (Substack)

The Asymmetry of Open Source — The Asymmetry of Open Source (link)

Projects to Keep an Eye On 🛠

Introducing Snorkel’s Foundation Model Data Platform (link)

Otter - A Multi-Modal Model with In-Context Instruction Tuning (otter-ntu.github.io)

Tracking Everything Everywhere All at Once (omnimotion.github.io)

hustvl/Matte-Anything: Matte Anything: Interactive Natural Image Matting with Segment Anything Models (Github)

web_ui

The Latest in AI Research 💡

SpQR: A Sparse-Quantized Representation for Near-Lossless LLM Weight Compression (arxiv)

Inference-Time Intervention: Eliciting Truthful Answers from a Language Model (arxiv)

Can Large Language Models Infer Causation from Correlation? (arxiv)

The World Outside of AI 🌎

Popular baby names: The psychology of growing up with an extremely common name (Quartz)

Is Marathon Running Healthy or Unhealthy? (Bloomberg)

NASA spends $45M to seed tech from hundreds of space businesses (TechCrunch)

Morality is declining, right? Scientists say that idea is an illusion (Nature)

Takeaway apps could drive healthier eating by listing salads higher (thetimes.co.uk)

One Last Bite 😋