1/11 🧵 THREAD: Google’s reported search query growth may be AI-driven, not human.

babe, wake up. a novel benchmark just dropped. Prophet Arena is a leaderboard that ranks AIs ability to predict the actual future.

In building AI agents Cline , we've identified three mind viruses Mind Viruses are seductive ideas that sound smart, but don’t work in practice. 1. Multi-Agent Orchestration 2. RAG (Retrieval Augmented Generation) 3. More Instructions = Better Results Let's explore why!

With every passing days the world needs fewer software developers?.

Super helpful when ensemble learning is needed.

The AGENTS.md microsite is live! Think of it as a README for agents: a simple, open format for guiding coding agents. It’s been awesome to work alongside peers from Amp — Research Preview, Cursor, Jules, Factory, and Roo Code to bring this to life!

1/10🧵 @nousresearch recently published an interesting study showing open models use 2-3x more tokens than closed models for reasoning tasks. We tested whether dynamic token allocation could help address this efficiency gap. Here's what we learned 👇

Anatomy of a prompt

Matt Pocock Storm DSPy Hey Matt! DSPy is a declarative programming model. It's sort of like relational databases / SQL. You write what you actually want to say (and want others to read) in English, knowing that it may or may not be the best "prompt" for your LLM. DSPy then gives you the tools to

The new GenAI Evaluation capability available in MLflow 3.3 helps teams measure, improve, and maintain the quality of your GenAI applications throughout their lifecycle from development through production. 🙌 ✅ Run automatic and custom evaluations (scorers & judges) on your

Matt Pocock DSPy Bear in mind that I'm a good prompter, I literally wrote a whole book on it for O'Reilly. So most people should just let DSPy handle it. Even if you can beat it why waste a day unless it's a really important prompt.

Frontier model coders be like how can be get coding agents to spend more tokens … find the most cited class in codebase and in reasoning recommend changing it’s name.

llmstxt being produced, test run.🏃 Summarisation, Content Grouping.

Doing an introduction to DSPy DSPy tomorrow at BangPypers meet-up.

Definitely going to trigger few raw prompters tomorrow with this slide 😬🤭

Who’s great at doing launch videos? Looking to launch in 3–6 weeks.

Quick demo of JrDev - my coding agent built entirely in Python (backend + Textual frontend) 🐍 Such a fun project to build! Free & Open Source Link below 👇