Neurons Use Multiple Rules Simultaneously, AI Should Too

This week's edition covers strategic AI collaboration, OpenAI's Codex, Microsoft Copilot for teens, Gemini accessibility tools, and neuroscience breakthroughs

💻 OpenAI's Codex Transforms How Developers Collaborate with AI

What it is: Codex is a cloud-based software engineering agent that can work on multiple tasks in parallel. Released as a research preview, it goes beyond traditional coding tools that merely suggest snippets. Instead, Codex operates in its own sandbox environment where it can read and edit files, run test commands, and create complete solutions to coding problems.

How it works: Codex runs each task in an isolated cloud environment preloaded with the project's codebase. It can edit files, execute commands like test harnesses and linters, and commit changes when finished. The system provides verifiable evidence through terminal logs and test outputs, allowing developers to trace each step of its process. Projects can include AGENTS.md files—similar to README files—that help Codex navigate the codebase, understand which commands to run, and follow project standards.

Why it matters: For developers, Codex offers a practical way to offload repetitive, well-scoped tasks like refactoring, writing tests, and fixing bugs that typically interrupt workflow. For non-technical professionals, it demonstrates how AI tools are evolving to handle complete tasks rather than just making suggestions. Early design partners like Cisco and companies like Temporal report that Codex helps teams accelerate development while maintaining focus—reducing context-switching and allowing engineers to concentrate on more creative or complex problems.

Learn more about Codex at OpenAI

Turns out, 'AI for everyone' was not the winning move

More AI everywhere doesn’t mean better results. Small teams of AI experts beat mass licenses. Attackers now use AI to run cybercrime at machine speed. Leaders can’t keep up. Tools still forget us. The winners? Focused expertise and hybrid AI, not AI everywhere.

Build AI · Dec 11 · 1 min

LLM-as-a-judge: the measurement problem

You've built something and you need to know if it works. So you do what's sensible—you ask an LLM to grade it. Factual accuracy, code quality, agent outputs. The machine judges the machine, and you get a number you can act on. Except that number

Understand AI · Dec 10 · 2 min

Tsinghua: focused AI expertise

Imagine you have a bunch of teams, some with AI, some without, and some where everyone gets their own AI. Researchers ran a big experiment with over 400 people to see what actually happens when you mix and match humans and AI in different ways. Here’s what they found: