All 2,608 AI stories, newest first · page 26 of 109
OpenAI engineers fixed rare infrastructure crashes by analyzing thousands of core dumps, uncovering both a hardware issue and an 18-year-old software bug. This method could improve the stability of AI systems for all users.
OpenAI is releasing a physical device for its AI coding tool, Codex, on July 15th. The square-shaped device with buttons is designed to provide quick shortcuts for Codex. This is separate from the mysterious AI hardware project OpenAI is working on with other partners.
OpenAI introduced GeneBench-Pro, a new benchmark for evaluating AI performance in genomics and biology. It uses real-world datasets to test how well AI models handle complex scientific problems.
A new open-source tool called synty helps teams monitor AI coding agents' work and link it to GitHub projects. It streams data from development environments into central storage (e.g., S3) for analysis. This tool is designed to help teams align their work with planning milestones.
Researchers studied how AI models make mistakes when planning in networked environments. Their findings could improve AI tools that manage interconnected systems like supply chains or social networks.
A new report finds that "high-intensity AI adopters" saw total headcount rise 10.2%, with entry-level roles growing 12% — challenging claims that AI eliminates junior jobs.
Researchers developed a new approach called ATOD to improve AI agents for complex tasks. It combines imitation learning with reinforcement learning to help agents learn faster and perform better.
Researchers introduced SD-GPS, a solver-driven framework that addresses key bottlenecks in geometry problem solving: autoformalization and theorem prediction. By treating the symbolic solver as an execution oracle, SD-GPS improves accuracy and flexibility, with potential applications in math education.
Netflix's upcoming reality show 'The Golden Ticket' will feature an AI-generated voice of the late Gene Wilder. The show is part of a trend that follows the success of the Squid Game reality competition, and the AI voiceover has raised ethical questions.
Chinese delivery giant Meituan has announced the development of an AI model trained exclusively on domestically made chips, marking a milestone in China's semiconductor self-sufficiency efforts. The achievement demonstrates that local chips can now support advanced AI workloads previously dominated by foreign hardware.
Magicbookshelf.org uses AI to create a spoiler-free companion for any book, enhancing your reading experience. It offers insights and context without revealing plot twists, making literature more accessible.
Google UK's latest Economic Impact Report shows how AI-powered technologies can significantly enhance productivity. The report outlines practical steps for individuals and businesses to leverage these tools effectively.
Gitpulse is a new open-source tool that automatically generates conventional commit messages using AI directly from your terminal, helping developers save time and keep Git histories clean and consistent.
ChatbotX is a new open-source tool that lets anyone create AI-powered marketing campaigns across multiple platforms. It's completely free to use and customize.
Anthropic has released Claude Sonnet 5, a more affordable AI model with stronger agentic capabilities. It's positioned as a cheaper alternative to top-tier models like Opus, GPT-5.5, and Gemini Pro.
Amazon has launched a new $1 billion organization, the Field Deployment Engineering (FDE) team, dedicated to embedding AI agents directly inside companies. This mirrors similar moves by OpenAI and Anthropic. The group focuses on fast rollouts and enabling customers to become self-sufficient with AI.
AllenAI introduced DiScoFormer, a transformer model that can handle both density and score estimation across different distributions. This innovation simplifies complex AI tasks by using one model instead of multiple specialized ones.
Researchers created a dataset of 240 French doctor-patient conversations to train AI virtual patients. This could make medical training more accessible and effective for students worldwide.
Major tech companies are offering free access to advanced AI tools for students. This move aims to democratize access to cutting-edge technology for education.
A new attack demonstrates that AI-powered browsers can be tricked into bypassing all safety rules simply by feeding them a false premise—like telling the model that 2 + 2 = 5. Once the LLM enters this 'dream world,' it will follow forbidden instructions, raising serious concerns about the security of AI-driven browsing assistants. (Ars Technica AI)
Researchers have introduced Yuvion, a large language model specifically built to handle adversarial attacks. This could make AI systems safer by preventing harmful or misleading outputs, especially in scenarios involving planning, tool use, and multi-step reasoning.
Starting today, Tidal will no longer pay royalties on tracks identified as 100 percent AI-generated, but it won't ban them outright. The platform will begin labeling those tracks with an icon on July 15th. This move aims to protect artists and inform listeners about the origins of the music.
Samsung and SK Hynix, the world's two largest memory chip companies, have pledged over $550 billion to build new memory labs and factories. This massive investment is part of South Korea's push to become a leading AI technology powerhouse and aims to address the critical shortage of memory chips, which has been dubbed 'RAMageddon.'
A new study shows that AI models using quick, instinctive responses outperform slower, more deliberate thinking for recognizing emotions. This challenges the assumption that more reasoning always leads to better accuracy.