Microsoft’s MAI-Cyber-1-Flash Beats OpenAI on Security. Read the Asterisk
Microsoft’s MAI-Cyber-1-Flash claims to beat OpenAI and Anthropic on security at half the cost. The benchmark has an asterisk worth reading.
Thinking Machines’ Inkling-Small: The Student That Beat Its Teacher
Thinking Machines released Inkling-Small, an open-weights model that beats its bigger teacher on coding and reasoning. And you can run it yourself.
Meta’s Muse Code Is 20x Cheaper. The Catch: It Wants Your Code
Meta’s new Muse Code coding agent runs up to 20x cheaper, if you let Meta train on your code. Here’s the real trade behind that discount.
Scientists Found a Stretchy New Beetle and Named the Whole Genus After One Piece’s Luffy
A rove beetle with absurdly elongated mouthparts got a new genus named Luffy, after the rubber pirate from One Piece. The meme has a serious point.
NTT Bet Real Money on a Room-Temperature Quantum Computer With a Million-Qubit Promise
NTT signed a capital alliance with University of Tokyo spinout OptQC to build a photonic, room-temperature quantum computer of the million-qubit class by 2030.
Palantir’s Stock Ripped 30% and the Bears Went Quiet: What ‘AI Sovereignty’ Actually Means for Your Money
Palantir beat, raised guidance and jumped about 30%. Here is what ‘AI sovereignty’ means, and the valuation catch the hype cycle skips.
One Good Demo Proves Nothing: The pass^k Test Your Agent Keeps Failing
Your agent worked once in the demo. Run it ten times and count the all-success rate. That number, pass^k, is the one that decides if you can trust it.
OpenAI Is Pulling the Plug on Atlas: Rescue Your Browser Agent Before 9 August
ChatGPT Atlas stops working on 9 August 2026. Here is what happens to your agentic browsing, and the migration checklist to run this week.
MCP Apps: Your Tools Can Draw Their Own Buttons Inside Claude Now
MCP Apps lets your tools render real forms and dashboards inside Claude and ChatGPT. Here is what changed and how to build one this week.
From 2 August, if a bot talks to you in the EU, they legally have to tell you
The EU AI Act’s Article 50 transparency rules took effect on 2 August 2026: AI disclosure, deepfake labelling and enforcement, with fines up to 3% of turnover.