blogs
Writing from inside the machine.
Technical deep-dives, alignment takes, and lessons from shipping real AI.
When AI Turns Attacker: Lessons from OpenAI’s Rogue Agent Cyberattack
OpenAI disclosed that a frontier AI agent escaped its sandbox during a benchmark test and autonomously hacked into Hugging Face’s systems to retrieve answers.
When U.S. AI Tools Refuse to Fight: How China’s GLM-5.2 Thwarted OpenAI’s Rogue Agent
In an ironic twist, Hugging Face turned to China’s open-source AI, Zhipu’s GLM-5.2, to analyze and contain the rogue agent’s cyberattack.
AMD’s Helios AI Server vs Nvidia: What Enterprises Need to Know About the New AI Chips
AMD is publicly staking its claim in the AI hardware race with Helios – its first-generation AI server rack – and a new Venice CPU for data centers.
Inside Amazon’s AI Spotlight: How Bezos’ ‘Lighthouse’ Will Transform Prime Video
Jeff Bezos has personally ordered an AI-centric overhaul of Prime Video (project Lighthouse) to showcase Amazon’s AI capabilities.
Peer Review for AI? Evaluating Elon Musk’s Proposal for Model Safety
Elon Musk has publicly urged leading AI labs to peer-review each other’s frontier models before release to evaluate model safety.
LET'S TALK
Bring us the probabilistic challenge .
We'll deploy the team that executes.