Articles tagged: AI safety

17 articles

Seeing beyond BMI: Estimating cardiometabolic risk with smartphone imagery

Smartphone imagery may offer a more nuanced view of cardiometabolic risk than BMI alone. This article examines the promise, safety conside...

Explore the risks and strategies for executing untrusted AI agent code without sandboxing, including isolation techniques, monitoring, and...

AI research is shifting from scaling generative models to building efficient, reasoning-driven systems. New paradigms like neuro-symbolic...

Explore how to use an LLM as an intelligent arbiter to select the best document from RAG retrieval candidates, enhancing accuracy with con...

AI research is advancing rapidly, from deep learning breakthroughs to foundational models. This article explores key trends like multimoda...

AI research is advancing rapidly, from deep learning to reinforcement learning. This article explores key breakthroughs, current challenge...

MosaicLeaks reveals how AI research agents can inadvertently reconstruct sensitive information from fragmented data. This article explores...

AI research is rapidly advancing, moving beyond narrow tasks toward general intelligence. Key areas include reinforcement learning, natura...

AI research is rapidly advancing, moving beyond pattern recognition to causal reasoning and foundational models. This evolution promises b...

The path to AI safety may lie not in treating AI as extraordinary, but in integrating it as normal technology. This article explores how s...

olmo-eval is an evaluation workbench designed to integrate seamlessly into the model development loop, enabling rapid iteration and system...