beat · 163 stories
Judge Hill called a warrantless license-plate-reader search "indiscriminate mass surveillance" and, though the ruling isn't binding precedent, gave state and federal offices a working warrant template from the 2018 Carpenter v.
Prompt injection hides commands in plain text to hijack AI assistants, and the fix sits with whoever deployed the model, not with the model itself.
About 80% of Wikipedia's budget comes from small donations prompted by human readers. AI is reading the same content without sending any, and two of the largest labs won't say if they pay.
Robinson's exit turns the safety debate from a vibe into a buildable question: who runs the next review, and which labs adopt third-party evaluation before the next launch.
Satoshi Ishida outlined the REX Project — a new AI-development phase of Capcom's in-house RE Engine technology, which powers Resident Evil — with auto-generated code and AI-driven test plays as targets.
A Maine lawsuit alleges DHS agents entered ICE observers into Palantir's case system as 'Threat to Law Enforcement, Professional Protestor.' The case now asks who audits the audit log.
Aleph Alpha's new open-weight language model Kolibri — 78B total / 3B active per inference pass via a Mixture-of-Experts design — is what decides whether a regulated European buyer can self-host a frontier-class model on its own hardware.
Musk says he is in talks with TSMC about a dedicated chip fab, his 'Terafab' initiative, to supply Tesla, SpaceX, and xAI. Intel, the only other named partner contributing its most advanced announced 14A chipmaking process, may not be in the room.
The 24GB M4 Pro MacBook Pro runs the lower 40 layers (stacked model stages) of the 64-layer Qwen3.8-27B; the iPhone 17 Pro Max's A19 Pro GPU runs the upper 24.
A new measurement study shows why one Adam run (the default neural-network training optimizer), in a neural network under ill-conditioning (a sharply uneven loss surface), can drift a thousandfold from the theoretical ideal and still reach the
An arXiv preprint proposes sampling tailored modifications to a model's own settings, rather than just longer streams of answers, as a way to buy more capability per fixed compute budget.
The 85% cache share — re-reads of text the model has already processed in the same session to keep its working memory coherent — points to high-bandwidth memory, not GPUs, as the next build constraint.