🥟 Chao-Down #221 $10 million AI Math Olympiad Prize Launches, The Pentagon's AI initiatives accelerate decisions towards autonomous weapons, McKinsey leader says 70% of jobs can be automated by AI

A significant limitation of large language models is their deficiency in doing complex reasoning and long-term planning, which is a crucial element for true intelligence. One organization is looking to help catalyze this space by putting up a $10 million prize.

XTX Markets, an algorithmic trading firm, is launching The AI Mathematical Olympiad — a new challenge to accelerate the development of AI models that can reason mathematically and solve problems at a level comparable to high-performing humans.

The grand prize will go to the team that releases a public model that can “perform at a standard equivalent to a gold medal in the International Mathematical Olympiad (IMO).”

It’s clear the motivation behind why XTX Markets is sponsoring the prize. Their firm is built on the work of math PhDs and geniuses who have likely competed in similar competitions.

Will we see a such a model emerge from the competition? I think of 10 million reasons why we will.

-Alex, your resident Chaos Coordinator.

What happened in AI? 📰

$10mn AI Mathematical Olympiad Prize Launches (aimoprize)

Pentagon's AI initiatives accelerate hard decisions on lethal autonomous weapons. (qz.com)

Sports Illustrated Published Articles by Fake, AI-Generated Writers (Futurism)

70% of jobs can be automated via AI, McKinsey thought leader says (Fortune)

AI’s Spicy-Mayo Problem: A chatbot that can’t say anything controversial isn’t worth much. Bring on the uncensored models. (The Atlantic)

The Sam Altman drama points to a deeper split in the tech world (The Economist)

Always be Learnin’ 📕 📖

Securing generative AI across the technology stack (TechCrunch)

The Roots of Today's Modern Backend Engineering Practices (The Pragmatic Engineer)

Navigating the Present and Future of Consumer-Led Growth (BCG)

Projects to Keep an Eye On 🛠

argilla-io/distilabel: ⚗️ AI Feedback framework for scalable LLM alignment (Github)

BuilderIO/gpt-crawler: Crawl a site to generate knowledge files to create your own custom GPT from a URL (Github)

run-llama/rags: Build ChatGPT over your data, all with natural language (Github)

The Latest in AI Research 💡

The GAIA benchmark: Next-gen AI faces off against real-world challenges (VentureBeat)

An Embodied Generalist Agent in 3D World (arxiv)

Comparing Humans, GPT-4, and GPT-4V On Abstraction and Reasoning Tasks (arxiv)

The World Outside of AI 🌎

70% of jobs can be automated via AI, McKinsey thought leader says (Fortune)

Tesla open sources all design and engineering of original Roadster (teslarati.com)

Zoom Fatigue is Real, According to Brain Scans (IEEE Spectrum)

Mirrors, Mazes and Meatballs: The Unbelievable Strategies IKEA Uses to Keep You Spending! - The Art of Want (Substack)

Why We’re Still Breathing Dirty Indoor Air (The New York Times)

Americans, Especially Millennials, Regret Moving to New Cities (Business Insider)

One Last Bite 😋

Watch my colleague Vic Perdana share an example reference architecture for building an AI Copilot using Semantic Kernel!