Microsoft and OpenAI Deepen Ties for Responsible AI Progress
New agreement between Microsoft and OpenAI aims to reshape AI safety and innovation, impacting industry dynamics.
Latest News
Tech Titans Forge the Future: Neuralink, Meta, and Blue Origin's 2026 Vision
Neuralink's sight-restoring chip, Meta's AI supercluster, and Blue Origin's lunar innovations could redefine technology by 2026.
OpenAI Buys Neptune to Boost AI Monitoring
OpenAI's acquisition of Neptune aims to sharpen AI transparency with better tracking and accountability tools.
OpenAI’s GDPval: A New Benchmark for AI's Economic Impact
OpenAI introduces GDPval, evaluating AI models on economically valuable tasks across 44 occupations, shifting focus to real-world impact.
Samsung and SK Team Up with OpenAI to Boost AI Infrastructure
Samsung and SK collaborate with OpenAI's Stargate to enhance AI infrastructure via advanced memory chips and data centers in Korea.
Oracle and OpenAI Expand Stargate: 4.5 GW Boost to AI Infrastructure
Oracle and OpenAI's partnership to expand Stargate data centers by 4.5 GW aims to strengthen U.S. AI leadership and create jobs.
OpenAI Challenges Court Order on User Data Retention Policies
OpenAI contests a court order demanding indefinite data retention, raising privacy and governance concerns in AI.
AI Framework Precisely Targets Subtle Online Sexism
Researchers unveil a novel two-stage AI system that significantly boosts the detection of nuanced sexist content, overcoming data limitations to set new performance benchmarks.
OpenAI's Gradient Noise Scale: AI Training Gets a Science Upgrade
New OpenAI research introduces a metric to predict neural network training efficiency, shifting AI development from guesswork to predictable science.
AI Safety Rethink: New Framework Tackles LLM Agent Uncertainty
A new paper argues current uncertainty quantification methods fail interactive LLM agents, proposing a 'conditional uncertainty reduction process' for safer AI.
OpenAI's Robot Hand Masters Rubik's Cube with Smart Training
A robotic hand, trained with reinforcement learning and a new randomization technique, now solves a Rubik's Cube. This shows AI's growing ability to handle complex physical tasks.
Research
See all →Tech Titans Forge the Future: Neuralink, Meta, and Blue Origin's 2026 Vision
Neuralink's sight-restoring chip, Meta's AI supercluster, and Blue Origin's lunar innovations could redefine technology by 2026.
AI Framework Precisely Targets Subtle Online Sexism
Researchers unveil a novel two-stage AI system that significantly boosts the detection of nuanced sexist content, overcoming data limitations to set new performance benchmarks.
OpenAI's Gradient Noise Scale: AI Training Gets a Science Upgrade
New OpenAI research introduces a metric to predict neural network training efficiency, shifting AI development from guesswork to predictable science.
OpenAI's Robot Hand Masters Rubik's Cube with Smart Training
A robotic hand, trained with reinforcement learning and a new randomization technique, now solves a Rubik's Cube. This shows AI's growing ability to handle complex physical tasks.
Hilbert-VLM Uses Fractal Curves to Map 3D Medical Images
By combining Meta’s SAM2 with space-filling curves, researchers aim to pinpoint complex pathology in 3D scans.
Uni4D Framework Boosts 3D Search and 4D Content Creation
Uni4D links text, 3D models, and images to sharpen AI for gaming and VR.