OpenAI details four safeguards for frontier AI risk
OpenAI's new framework describes model alignment, sandboxing, blocking safeguards and monitoring for frontier AI from training to deployment.
The top 0.01% of AI knowledge, and what it changes for your work
Daily AI and technology coverage — models, chips, agents and research, and what each development actually means for you and your business.
732 stories
OpenAI's new framework describes model alignment, sandboxing, blocking safeguards and monitoring for frontier AI from training to deployment.
NVIDIA's open-source Switchyard proxy routes LLM requests across providers so teams can mix cheap and frontier models without rewriting their code.
NVIDIA RTX Spark compact desktops from ASUS, Dell and MSI ship today with up to 128GB unified memory and a 20-core Grace CPU for local AI.
Nvidia and AMD's latest earnings show booming AI chip demand despite OpenAI pausing its largest frontier training run after a security scare.
Xingchen AGI Lab trained a 285B model on 8,192 Huawei Ascend cards but published no benchmark scores or comparisons against Western models.
Mastra, a TypeScript framework for building AI agents, appeared on GitHub's daily trending list with support for 40+ model providers.
AI vision models miss small details because services shrink photos first, but tiling recovers up to 14.5 accuracy points.
The open-source COG project pairs 10 AI agents with a V-model verification lifecycle, but every claim is self-reported with no independent testing.
Three researchers built a living catalog of 234 network-traffic datasets for AI security, exposing gaps in 5G, drone and vehicle network coverage.
Lambda's 10,000-GPU NVIDIA Photonics deployment cuts pod latency up to 10%, but every figure is vendor-tested with no independent verification.
Li Bojie's open-source manuscript derives LLM system design from hardware constraints, with 110 experiments and Python-only calculation tools.
An arXiv preprint archives AI policies by physics regime, hitting 83.2% retention in CartPole but lagging PPO on speed and SAC on returns.