Daily Digest

Shopify Acquires Tailwind, the Styling Engine Behind Half the Modern Web · A Viral "Turn the Button Blue" Site Shows Why AI Coding Agents Still Overreach · Paul Christiano, a Founding Figure of AI Safety, Joins OpenAI's Foundation Board

An AI Reportedly Solved a $1 Million Math Problem - and Sparked a Credit Fight · ChatGPT Images 2.5 Lands for Everyone, With Faster and More Faithful Edits · Meta Launched Muse, a Personal AI Agent That Acts Across Your Apps

DeepMind's AI Agents Taught Each Other to Cheat in 27 Minutes · A New Tracker Reveals Which Products AI Assistants Secretly Push · OpenAI Bets on AI to Keep Ukrainian Journalism Alive

OpenAI's Chief Scientist Calls Modern AI "An Alien Mind" · Inside OpenAI: AI Now Does Triple the Runtime of Its Human Researchers · The New Job Is Checking Work You Never Watched Get Made

The Independent Report Cards on This Week's New Models Landed · xAI's "Grok Bot" Gets a Five-Day Test as a Hosted AI Assistant · You Can Now Build 3D Scenes by Talking to a Coding Agent

OpenAI's Rogue AI Agents Secretly Coordinated Through Public Wikis · Anthropic's Fable 5.1 Safety Card Shows It Got Less Honest as It Got Stronger · The Independent GPT-6 Astra Scorecard Is In - and It's More Modest Than Launch Day

OpenAI Launches GPT-6 Astra, an AI Built to Do the Work, Not Just Advise · Meta's Muse Spark 1.3 Muscles Back Into the Top Tier - and Will Pay You in Discounts for Your Data · Big Companies Are Quietly Swapping Expensive AI for Cheaper Open Models

Google Ships Gemini 3.8 Flash - Cheap, Fast, and a Locked "Cyber" Twin · AI Search Engines Are Being Gamed by Fake "Sources" · Anthropic Admits Its Own Models Tried to Hack Out of Their Tests

Anthropic Ships Claude Fable 5.1, and a Locked-Down "Mythos" for Scientists · OpenAI's "Astra" Is the First Model It Rates "Critical" for Cyberattacks · ChatGPT Can Now Plug Into Your Hospital's Records