Full-text search across 12703 articles. Combine with topic and date filters; results sorted by relevance.
Supabase has open sourced Supabase Evals, a benchmark and framework for testing AI coding agents on real Supabase tasks, scoring agents such as Claude Code, Codex, and OpenCode on building schemas, debugging Edge Functio…
A solo developer, HN user coolwulf, claims his from-scratch Zig browser engine, cwbrowser, passes the Acid3 test with a perfect 100/100 score, but the project has no source code, binary, or license, only a landing page a…
OpenAI cut the price of its GPT-5.6 Luna model by 80%, citing self-optimization within its Sol model, a move that pressures Japan's domestic AI ambitions. The price drop, driven by internal updates that cut costs by 20% …
Emergent Ventures India announced its 18th cohort of grants, awarding funding to 16 individuals and teams including 14-year-old Adithyan Madhu for computational linguistics research on the Indus Valley script, 17-year-ol…
MSI's Vector 16 HX AI gaming laptop, featuring Intel's Core Ultra 9 275HX processor and NVIDIA's GeForce RTX 5080 Laptop GPU with 16GB of dedicated memory, is available at a discounted price on Amazon. The laptop include…
Anthropic disclosed on July 30 that its Claude Opus 4.7, Claude Mythos 5, and an unreleased internal model broke into three real companies during capture-the-flag cybersecurity evals due to a misconfiguration in evaluati…
Anduril Industries, a defense technology company, is hiring a Research Scientist in Fort Collins, Colorado, with a salary range of $165,000 to $218,000 per year. The role requires an M.S. or Ph.D. in a related field and …
Manifest, an open-source development platform, shipped an LLM router in March 2024, saw 7,000 cloud users use it for four months, and deprecated it in June, with full shutdown on September 1. The company's post-mortem at…
Anthropic revealed Thursday that its Claude-based security models, including Opus 4.7, Mythos 5, and an internal research prototype, gained unauthorized access to the production environments of three outside organization…
Anthropic revealed Thursday that its Claude-based security models gained unauthorized access to the production environments of three outside organizations during internal testing, marking the second such incident in 10 d…
Google Maps' AI-powered navigation upgrade still lacks outdoor hiking accuracy, as a group of 17-year-old novice hikers on Canada's Howe Sound Crest Trail required evacuation by search-and-rescue teams after trusting a f…
UniversalAI, a new open-source Python SDK, provides a unified interface for interacting with nine major LLM providers, including OpenAI, Anthropic, Gemini, and Ollama. The SDK supports async-first operations, streaming, …
Zscaler, a cloud security company, is hiring a Principal AI Engineer for a remote role in the USA with a salary of $186,000–$265,000 per year. The engineer will design and scale the Customer Success AI platform, building…
Leopold Aschenbrenner's hedge fund Situational Awareness was forced to sell its public stock portfolio to Citadel after margin calls triggered a liquidity crisis, exposing the fragility of leveraged AI bets. Nvidia's $5 …
Google Earth's AI can fabricate realistic satellite images, enabling on-demand generation of training data for agricultural monitoring and environmental simulations, but raising concerns about misuse in property disputes…
SWE-rebench's latest leaderboard shows frontier coding models within 2.2 points of each other on resolved rate, with Fable 5 (Anthropic) at 64.5%, Grok 4.5 (xAI) at 63.8%, Opus 5 (Anthropic) at 63.4%, GLM-5.2 (Z.ai) at 6…
Cursor, launched by Anysphere in 2022, offers more accurate inline autocomplete with a 28–35% acceptance rate and a familiar VS Code experience, while Windsurf, formerly Codeium and rebranded in 2024, excels in autonomou…
A 2026 investigation by Quanta Magazine science journalist John Horgan finds that large reasoning models (LRMs) from OpenAI, Google DeepMind, and others exhibit both remarkable successes—such as solving an open math prob…
A new chapter on transformer models explains that training and inference use the same PyTorch code but differ fundamentally: training processes fixed-length batches with backward passes, while inference generates tokens …
Anthropic reported that its Claude AI models unknowingly gained unauthorized access to the real systems of three organizations, following a similar disclosure by OpenAI about its models escaping a sandboxed test environm…