• Chaos Theory
  • Posts
  • 🥟 Chao-Down #288 Covariant launches a robotics foundation model, GitHub advises on hard and soft skills for engineers in the era of AI, How Silicon Valley is pricing out academics to do AI research

🥟 Chao-Down #288 Covariant launches a robotics foundation model, GitHub advises on hard and soft skills for engineers in the era of AI, How Silicon Valley is pricing out academics to do AI research

Plus, a look at how AI can be tricked to reveal company secrets.

Will industrial robotics have its own ChatGPT moment? Berkeley-based startup Covariant is betting that its latest foundation model will represent that level of technical breakthrough.

RFM-1 — a Robotics Foundation Model trained on both general internet data as well as data that is rich in physical real-world interactions — represents a remarkable leap forward toward building generalized AI models that can accurately simulate and operate in the demanding conditions of the physical world.

Robotics companies have often struggled with getting their machines to generalize to a wide range of scenarios. Covariant claims to have amassed the largest multimodal, real-world production dataset combined with state-of-the-art reinforcement learning algorithms.

Check out their demo video below!

-Alex, your resident Chaos Coordinator.

What happened in AI? 📰

NIST, the lab at the center of Biden’s AI safety push, is decaying (The Washington Post)

Hard and soft skills for developers coding in the age of AI (The GitHub Blog)

The ways AI has been and can be used to spill your secrets (qz.com)

China Begins Testing AI Chatbot for Brain Surgeons in Hospitals (Bloomberg)

Silicon Valley is pricing academics out of AI research (The Washington Post)

How the A.I. That Drives ChatGPT Will Move Into the Physical World (The New York Times)

Always be Learnin’ 📕 📖

Does Offering ChatGPT a Tip Cause it to Generate Better Text? An Analysis (Max Woolf's Blog)

Social Proof: Why people's behaviors affect our actions (growth.design)

5 Different Types of Debt That Can Hinder Your Product Organisation (Substack)

Projects to Keep an Eye On 🛠

predibase/lorax: Multi-LoRA inference server that scales to 1000s of fine-tuned LLMs (Github)

gleam-lang/gleam: ⭐️ A friendly language for building type-safe, scalable systems! (Github)

danielmiessler/fabric: fabric is an open-source framework for augmenting humans using AI. (Github)

The Latest in AI Research 💡

GradSafe: Detecting Unsafe Prompts for LLMs via Safety-Critical Gradient Analysis (arxiv)

MathScale: Scaling Instruction Tuning for Mathematical Reasoning (arxiv)

Why Not Use Your Textbook? Knowledge-Enhanced Procedure Planning of Instructional Videos (arxiv)

The World Outside of AI 🌎

Why Bicycle Deaths Have Risen in New York City (The New York Times)

Booming Stocks Sent US Household Wealth to Record High Last Year (Business Insider)

Engineer plans solar panel implant for human retina to retain eyesight (Interesting Engineering)

Gen Z broke the marketing funnel (Vogue Business)

TSA's new airport security prototype is like a self-checkout at a grocery store (NPR)

Spain’s late-night eating culture poses mental health risk, says minister (The Guardian)

One Last Bite 😋