Full-text search across 67173 articles. Combine with topic and date filters; results sorted by relevance.
OpenAI retracted its recommendation for Scale AI's SWE-Bench Pro coding benchmark after an audit found roughly 30% of its 731 tasks were broken due to overly strict tests, underspecified prompts, low-coverage tests, and …
Context rot degrades AI agent output quality when a model's context window reaches 70-80% capacity, causing attention dilution and signal-to-noise loss. Anthropic's Claude Code addresses this with an auto-compact feature…
ProductSpec, an open standard for documenting software intent before implementation, has been released. It provides a portable Markdown format for product decisions that precede tickets, engineering plans, and code, enab…
Microsoft clarified in a September 2025 .NET Blog post that GitHub Copilot's repository-wide copilot-instructions.md files apply automatically to every request, while *.prompt.md files are reusable, task-specific templat…
The tensor4all team, led by Hiroshi Shinaoka of Saitama University, has released tenferro-rs, a Rust-native dense tensor stack for scientific computing that supports linear algebra, autodiff, einsum, FFT, and CPU/CUDA ba…
A staff engineer quit their job at a startup after 2.5 years, citing dissatisfaction with AI's impact on software development. The engineer, who turns 40 next year, argues that large language models replace 90% of the en…
Cursor acquired open-source AI coding assistant Continue.dev in a quiet acqui-hire on June 16 and is shutting the product down. All user data will be permanently deleted on July 15, with no migration path provided. Users…
Google AI Studio's Build mode enables developers to create full-stack web and native Android apps from natural language prompts, with one-click deployment to Google Cloud. A developer demonstrated building a Task Manager…
OpenAI launched GPT-5.6 Sol, Terra, and Luna as a restricted preview limited to trusted partners at the request of the U.S. government, marking a shift toward government-mediated frontier AI releases. The flagship Sol mo…
A developer compared Bolt.new and Velra for building web apps with billing. Bolt.new excels at rapid prototyping with live editing, while Velra generates a production-ready SaaS with Stripe subscriptions wired in and ful…
A developer warns that using AI to write code without designing for the system around it leads to a 'Tactical AI Trap' of accumulating technical debt. They advocate shifting from tactical programming—solving immediate ta…
AWS released the Agent Toolkit for AWS, an open-source suite of MCP servers, plugins, and skills that gives AI agents governed access to live AWS resources, addressing the problem of LLMs hallucinating outdated cloud API…
Microsoft Threat Intelligence reported that North Korean state-sponsored actor Sapphire Sleet compromised over 140 packages in the @mastra npm scope, targeting developers building AI coding tools. The attack used a posti…
A developer built a private AI-readable work memory by wiring email, meetings, documents, and notes into an Obsidian vault that Claude Code can access, solving the problem of scattered work context that prevents AI from …
A developer built an end-to-end AI-SRE system using HolmesGPT, mirrord, and a Claude wrapper to autonomously verify fixes on a real GKE cluster. The system investigated a checkout error, generated a code patch, and teste…
Anthropic developed a method to read a model's internal state as plain English. Groq's founder argued that cheaper AI only increases compute demand. GLM-5.2 released an MIT-licensed model with near-frontier coding capabi…
An engineer tested how much of a blog post Anthropic's Claude actually retrieves during AI search, finding that the search stage only returns the title and URL—no body content. The WebFetch stage can retrieve the full pa…
The VS Code Eval Team ran a five-line evaluation task 50,974 times across 30 models over six months, revealing significant differences in how models handle a simple file-writing request. Model-A always used a single dire…
OpenMontage, the first open-source agentic video production system, has been released, enabling AI coding assistants to handle research, scripting, asset generation, editing, and final composition from plain-language des…
A developer released DOS, an open-source tool that verifies AI agents' completion claims by checking git history and other artifacts instead of trusting the agents' own statements. The tool, available as a Python package…