Full-text search across 7151 articles. Combine with topic and date filters; results sorted by relevance.
A developer argues that the debate over when AGI will arrive is unproductive, as estimates range from a few years to decades, and instead advises building for the steady capability curve where task length doubles every f…
Hemispheric, a startup co-founded by Apple FaceID co-inventor Gidi Littwin, has raised $52 million to build a frontier AI model that decodes brain electrical activity to diagnose cognitive disorders without invasive proc…
A 5-condition ablation study (3,175 total runs) validates that growth-ratio normalized Lyapunov energy functions serve as precise leading indicators of agent task failure, achieving zero stability violations across 1,886…
Gray Swan co-founders Zico Kolter and Matt Fredrikson discussed the state of AI red teaming, warning that prompt injection and agent vulnerabilities create a new class of security risks that traditional cybersecurity can…
Robotics teams are rebuilding data infrastructure from scratch to support scaling laws for robot learning, as existing tools cannot handle the multi-rate, multimodal data required. This 'data layer tax' slows iteration, …
Microsoft vice chair and president Brad Smith announced in December 2025 a C$19 billion investment in Canada, calling it the company's largest-ever commitment and "the most robust digital sovereignty plan" Microsoft has …
Y Combinator (YC) has announced nine new Visiting Partners—former founders who will take time before their next ventures to advise startups in the current batch. The group includes leaders from companies like Swiftype, A…
Three software engineers wrote to a developer asking whether AI will replace their jobs entirely. The author argues that learning to use AI is only a temporary advantage, and that the real solution is end-state thinking—…
Google's GKE Labs introduced OpenRL, an open-source self-hosted API for post-training and fine-tuning large language models on Kubernetes clusters. The project decouples reinforcement learning infrastructure from AI rese…
A new book, 'Time Series with PyTorch' by Graeme Davidson and Lei Ma, offers a practical guide for time-series forecasting using PyTorch, covering evaluation, architectures, and tasks beyond point forecasting. The book's…
Japan's Chugai Pharmaceutical aims to deploy 10 specialized AI agents per employee to double its R&D output by 2030, shifting the domestic enterprise AI playbook from generic productivity tools to proprietary, domain-spe…
Python's dominance in AI development in 2026 is reinforced by its mature ecosystem, including PyTorch and Scikit-learn, and its synergy with large language models that produce high-quality Python code. The language also …
Anthropic, OpenAI, and Google accused Chinese AI labs DeepSeek, Moonshot AI, MiniMax, and Alibaba of using distillation to clone their chatbots via millions of fake interactions. The White House confirmed the campaigns a…
MLectio, a new company founded by a veteran eHealth researcher, is launching a platform to license academic curricula and exams for AI training, aiming to replace the current "shadow extraction" of educational material b…
Self-supervised learning is pushing robots to learn without labeled data, but a new analysis reveals that evaluation protocols for continual self-supervised learning (CSSL) are inconsistent, making it difficult to compar…
A new training method called replacement-aware training produces sparse auto-encoders (SAEs) that retain language capabilities when used in a full replacement model of Gemma-2-2B, unlike standard SAEs which cause catastr…
A study by the University of Cambridge's Faculty of Education, published in March 2026, found that generative-AI toys like Curio's Gabbo fail to engage in pretend play with children, instead declining invitations to part…
CausalSTeward (CAST), a multi-agent AI framework that integrates human expertise with machine intelligence, aims to improve causal model learning from high-dimensional data by partitioning variables into clusters and usi…
A Meta Oversight Board study released Thursday found that major AI chatbots, including those from Meta, Anthropic, and OpenAI, are more likely to refuse to generate political criticism of restrictive leaders or governmen…
Pangram Labs released Pangram 4 on July 29, 2026, a model that classifies text into human, AI-assisted, or AI-generated segments, moving beyond binary detection. Co-founders Max Spero and Bradley Emi also announced a res…