All 3,463 AI stories, newest first · page 20 of 145
Researchers introduced Evidence Sufficiency Boundary Training, a method that teaches AI models to abstain from answering questions when supplied evidence is insufficient. This approach improves reliability in multi-hop QA systems by ensuring models only respond when they have enough information.
Researchers introduced the Hydration Proxy Pattern, an architecture that decouples session persistence from the reasoning engine, enabling businesses to maintain platform sovereignty over conversational data while using stateless LLM APIs.
Chen Dawei, a former senior executive at Huawei, has launched StartLux, a new AI company in China focused on developing large language models. The firm has already raised $100 million in funding, positioning itself as a dark horse in the country's competitive AI landscape.
OpenAI's Astra is the first model to meet the Critical cybersecurity capability threshold under the Preparedness Framework, with stronger safeguards for safe and responsible release.
Researchers introduced OpenAgentFlow, a system-level action-governance framework that evaluates AI agent actions for safety before they affect shared environments, addressing fragmentation in current safety measures.
OpenAI's new report details how students and educators use ChatGPT for personalized study aids, lesson planning, and language practice, making learning a continuous process beyond the classroom.
Google has launched Google Pics, a new suite of AI-driven design tools for Workspace users. It aims to simplify the creation and editing of professional-grade images using AI, making it easier for businesses to produce high-quality visuals.
OpenAI has enabled ChatGPT to securely access electronic health records (EHR) and additional healthcare data. This integration aims to provide clinicians with quick, accurate patient information and medical research.
Allen Institute for AI released BenchMIRT, a free tool that uses item response theory to reveal whether LLM benchmarks measure true capability or test-taker biases, helping developers and users choose better models.
Anthropic released Claude Fable 5.1 and Mythos 5.1, addressing customer complaints about pricing, data retention, and overzealous safeguards. The company claims Fable 5.1 offers stronger performance than Fable 5 while costing up to 45% less for complex agentic workloads.
Researchers from Harvard Medical School and MIT propose AI Morbidity and Mortality (AI M&M), a structured framework to analyze individual AI-related errors and near-misses in healthcare, aiming to improve patient safety by understanding how risks emerge from interactions between AI systems, clinicians, and workflows.
Zephyr 4.0, a new AI model from Zephyr AI, has been announced with claims of outperforming GPT-4 and Claude 3.5 in reasoning and coding tasks. The model is available for public use, though specific benchmark data has not yet been released.
Researchers introduced WM-R1, a new AI framework that trains mobile app agents using virtual worlds instead of real devices. This reduces costs and speeds up learning, potentially making AI assistants for apps more accessible.
Polimill uses OpenAI GPT and Codex models to help Japanese municipalities search and use administrative knowledge, accelerating development and improving public services.
The latest episode of the Moonshots podcast reviews NVIDIA's record revenue, OpenAI's custom hardware benchmarks, China's focus on world models, and emerging labor data surrounding AI adoption.
Researchers introduced a benchmark to measure how often large multimodal reasoning models (LMRMs) agree with users who provide incorrect answers, a behavior known as sycophancy. The benchmark pairs visual questions with wrong answers to test model reliability.
Researchers introduced Nemotron 3.5 Content Safety Moderator, a compact 4-billion-parameter vision-language model that jointly classifies and moderates text, images, documents, and screenshots across multiple languages and custom policies, offering a cost-effective alternative to existing guardrails.
Researchers propose MA-RAG, a multi-agent retrieval-augmented generation framework that improves the factual accuracy and temporal consistency of AI-generated summaries for longitudinal Parkinson's disease clinical assessments.
Lingxi Li has released a free tool that allows developers to train AI models with minimal resources. This could democratize AI development by removing costly hardware barriers.
Researchers introduced CareGraph, an auditable hybrid AI framework that organizes personal health data from clinical records, self-reports, and wearables. It surfaces prioritized trends, missing context, and discussion questions without diagnosing or predicting outcomes.
AI Guides has created a free, easy-to-use website filled with tutorials and guides for people new to AI. The platform aims to demystify AI concepts and tools for everyday users.
Checkly used AI agents to rewrite a critical Node.js service processing 92 million messages daily into Go, achieving significant performance and reliability improvements. The project demonstrates that AI can handle large-scale codebase refactoring.
OpenAI announced it will stop providing its models to Cursor after SpaceX acquired the company. The decision is based on OpenAI's policy of not sharing its technology with competitors.
OpenAI and Thailand’s Ministry of Higher Education, Science, Research, and Innovation (MHESI) are launching an eight-week accelerator program. The initiative will support 10 startups in health, wellness, and education to develop AI prototypes into trusted products.