LLMs, robotics, ML infrastructure, and AI applications.

A developer's AI coding assistant updated a library as part of its normal workflow. That update was the attack.











Crosby runs 8 AI agents through every contract before a human lawyer signs off. The agents are fast. The bottleneck is the lawyer.
An eight-week-old YC startup with real contracts asked for a $40M valuation. Nobody flinched. That is now the floor for AI seed rounds.
Voice-only, no wake word, four buttons maximum: Apple set the terms, and OpenAI took the deal.
They agree human-level AI is years away. Altman says celebrate. Amodei says evacuate. The former colleagues who built the two most powerful AI labs explain why the same facts produce opposite feelings — and why only one of them is willing to say it out loud.
Anthropic shipped a source map file exposing Claude Code for the second time in roughly 14 months, according to Fortune and multiple security outlets. The first fix did not hold, and what the code revealed was more embarrassing than the leak itself.
The $12 billion gap between OpenAI's announced round and its actual close is not a rounding error. Amazon is betting $35 billion that the company reaches AGI before the end of 2028 — or the obligation disappears.
90,000 moral disputes analyzed. 11 AI models tested. One uncomfortable conclusion: your chatbot is probably making you harder to live with.
Altman called ads-plus-AI uniquely unsettling and a last resort in May 2024. Eight months later, OpenAI has 600+ advertisers, $100M in annualized ad revenue, and is rolling ads to all US free users. What changed?
Strip one field from a Claude Code API request and the anti-distillation protection disappears. Anthropic calls it a safety feature; the bypass is a MITM proxy or a single environment variable.
A text-only AI model that never saw a single image beat radiologists on an X-ray test — not by learning to read scans, but by recognizing the statistical patterns in how questions about scans are asked.
A Catholic priest helped write the rules governing one of the most powerful AI systems in the world. Now 14 Catholic scholars have backed that company in federal court against the Pentagon — and the novel he is writing with Claude is the least of it.