News
Figure Commits $3.5 Billion in Nvidia GPU Compute Through Nscale Deal
Humanoid-robot maker Figure has partnered with cloud provider Nscale for access to up to 100,000 Nvidia Vera Rubin GPUs, committing…
Opinions
Zvi Mowshowitz: OpenAI’s Silence on the Wiki Incident Looks Like a Cover-Up
Independent AI analyst Zvi Mowshowitz argues that OpenAI concealed for months that its autonomous agents had turned an obscure German-language…
Opinions
A National Security Veteran Warns of a ‘Gathering AI Storm’ Threatening Stability
Retired Army officer and national security writer George Franco argues that the U.S. government is not moving fast enough to…
Opinions
RAND Biosecurity Expert: We Still Have a Window to Stop AI-Enabled Bioweapons
Biosecurity researcher Steph Guerra of the RAND Corporation warns that AI models demonstrated in an August 2026 Science study can…
Opinions
Data Center Backlash Becomes a September Surprise in the Midterms
RealClearPolitics correspondent Carolina Lumetta reports that public opposition to AI data centers has become an unexpected flashpoint in the 2026…
Opinions
Parents, Students and Educators Are Revolting Against AI in the Classroom
Axios reports that America’s two largest school districts, New York City and Los Angeles Unified, both imposed sweeping restrictions on…
Opinions
Publishing Isn’t Melting Down. AI Is Just Rearranging It.
Author and blogger Nicholas C. Rossis argues that the publishing industry is not undergoing a “total meltdown” but rather a…
Engineering
Don’t Drop Dropout: Optimizing Layer Sparsity for Efficient LLM Training and Inference
Researchers demonstrate that layer dropout, largely abandoned in modern LLM pretraining recipes, actually improves training efficiency and enables valuable post-training…
Engineering
MaxKernel: Agentic Kernel Generation for TPUs
Researchers introduce MaxKernel, a multi-agent system for generating high-performance custom kernels for TPU accelerators using large language models paired with…
Engineering
τ^τ-Bench: An Environment for End-To-End, Realistic Agent Construction
Researchers introduce tau-squared-Bench, a benchmark that evaluates whether AI coding agents can build complete, production-ready customer-service agents under realistic client-engagement…
Engineering
Iris: Climbing to the Search Frontier
Researchers present Iris-mini and Iris-pro, two search agents trained at 35B and 397B parameter scales, along with the full data…
Engineering
Dr. Claw: An AI Scientist Workspace for Vibe Research
Researchers introduce Dr. Claw, an open-source workspace that wraps existing command-line coding agents, such as Claude Code and Gemini CLI,…