Cap your agent review loops at 3 rounds Per their writeup, one Claude Code PR ran 50+ rounds of review and fixes overnight, mostly re-flagging old code. After the cap, PRs needing 5+ rounds went from 8.9% to 1.5% for your REVIEW.md After 3 rounds, stop and ask the author to rethink the approach xitter.cf/ClaudeDevs/status/2111…

oh yeah btw guys we can do integer multiplication faster than n log n lol I was definitely very surprised when this one came in lol github.com/openai/math/tree/…

Honestly stunned by result 128 Shortest Common Superstring went from 3x optimal to 7/3 over 35 years of human work. OpenAI's model got it to exactly 2. And it didn't use greedy, the algorithm the 1988 conjecture was about. Lean-checked I animated the proof in 6 minutes:

Meet Mistral Large 4, aka Le Chonk. • 1T parameters, natively multimodal. 49B active. It is the best open weights model from US or Europe on aggregated benchmarks. • State-of-the-art on critical workloads, including cyber defense, manufacturing and finance and it surpasses closed frontier models on visual grounding. • Forged in Europe end-to-end and is deployable from Europe via our own Mistral Cloud infrastructure. • Available to all via API today. Working with cybersecurity partners privately. Open weights release end of October.

MISTRAL 🔥: A new big model from Mistral is about to drop, according to Reuters. > “The model we’re actually announcing today is actually above the Chinese models on certain aspects, including cyber. So the narrative that Europe cannot compete is something that is not true,” Since we expect it to outperform Chinese models, there is a big Chunk of possibilities that it will be open weight. Right? A big Chunk 👀

Auto-review is now free for everyone who uses it via ChatGPT account! I default to “Approve for me” for all my codex tasks and auto-review makes sure that I’m protected from any unintended consequences If you haven’t already, enable it in your settings - Enjoy!

We are making Markdown files work natively across Google Drive and Docs. You can now preview .md files in Drive, open them in Docs, edit them, comment on them, and collaborate without converting the file into a Google Doc!

I released herdr gpui exactly 2 weeks ago, it has become my daily driver for all projects including at work. Looks like it's getting traction and others like it too, let me know what's missing to make it your daily driver too. github.com/penso/herdr-gpui

serious hot take: astra > opus hear me out before you start crying in my comments i'd never pick astra over opus for daily work, but every time i run an adversarial review on a plan that opus created, astra absolutely tears it apart 30% of astra's feedback is over-engineering, yes but the other 70% are genuine holes and improvements that, honestly, opus should have paid attention to TL;DR: - astra still is significantly smarter/more diligent - despite this, opus is the better everyday model once OpenAI gets speed and cost under control, i'd place bets that they will have the better overall model (for the first time ever, i'd argue)