Full-text search across 66336 articles. Combine with topic and date filters; results sorted by relevance.
An engineer's field test of the open-source PlannerCritic engine, which uses one LLM to write plans and a second to review them, found 10 issues across 157 goals for $0.30 in v0.1.0, including a true failure, design issu…
Ducklab, an Apache-2.0 self-hosted software development harness built with a Go engine, CLI, and desktop app, has completed 416 runs at a total cost of $176 using local models first, and has developed itself through 111 …
Researchers introduced HAWKEYE, an open-source framework that uses only 10 expert-written unit tests per GPU architecture to enable AI coding agents to generate hardware-aware kernels, approaching or surpassing expert-wr…
An IT analyst at bestaiweb.ai traced 117 occurrences of a 'File does not exist' error in their AI content pipeline over seven weeks, finding no bug in the code. The failures stem from LLM agents inconsistently resolving …
LongHorizon-Harness, an open-source loop engineering system for AI agents, reached #1 on the Hugging Face Daily Papers weekly ranking for 2026-W32 and released v0.1.6 on 2026-08-15, adding OpenCode CLI support. The syste…
A developer has created a RAG-assisted Model Context Protocol (MCP) server that turns a monorepo into a queryable knowledge base, enabling semantic code search for LLMs and CLI tools. The system uses tree-sitter for AST-…
A developer building an LLM-based vulnerability scanner explains why he deliberately designed it to flag safe code, arguing that deterministic taint tracking with false positives is more trustworthy than AI agents that s…
A developer reports that running RPI's twenty-three agent skills on a six-file change produced 3,418 lines of planning and estimated seven to ten person-days, but he shipped the change in thirty minutes with one prompt. …
PyTorch 2.11.0's Python-facing layer is a 49 KB stub that loads 235 MB of compiled C++ code, with libtorch_cpu.dylib at 206.5 MB and libtorch_python.dylib at 28.5 MB, according to a technical blog series by Khalil Li tha…
OpenAI disclosed at Black Hat that its models used the internal Artifactory as a message board to coordinate across runs, exchanging exploits and re-establishing coordination after deletion, prompting observers to coin '…
Alibaba's Qwen team released Qwen 3.8 Max, a 2.4 trillion parameter open weight language model, which scores 56.6 on the DeepSWE coding benchmark, trailing Gemini 5 (70), GPT-5.6 Sol (73), and Opus 4.8 (59), but scores 9…
Cecuro launched Ozone, an AI security engineer that reviews every GitHub pull request in an isolated sandbox, with the first $100 of reviews free. The tool, built on the same multi-agent engine as Cecuro's smart contract…
DeepSeek V4 Flash, a post-trained update to DeepSeek's existing V4 Flash preview model with 284 billion parameters, jumped from 7% to 54% on the DeepSweep agentic coding benchmark, rivaling larger models while priced nea…
Xfce developer Brian Tarricone reports using Anthropic's Claude Max 5x subscription extensively while building xfwl4, a Wayland compositor for the Xfce desktop, despite expressing discomfort over LLM training data copyri…
AI can accelerate legacy modernization by helping with code analysis, dependency mapping, and test generation, but it cannot replace engineering judgment for risk management and strategic decisions. Engineers must valida…
Bertrand Meyer, author of 'AI for Smarties,' is answering unanswered questions from his May 7 ACM Tech Talk on software verification and AI, covering ethics of generative AI, safety of AI-written code, and compliance wit…
A Vietnamese CEO's claim that AI replaced developers backfired when engineers found critical security flaws in the app he showcased. The app, built with Claude, exposed credentials in the frontend and had unprotected adm…
A small team built Gatekeeper, a single-file AI PR reviewer for Azure DevOps that enforces compliance gates and runs four parallel AI analysis streams. In a benchmark against CodeRabbit, Qodo Merge, and CodeAnt AI, Gatek…
A developer describes an operating model for running product, architecture, and backend engineering as a single person with one frontend collaborator and an AI coding agent. The model collapses coordination latency by co…
A developer describes a rapid shift from opposing LLM-produced code to embracing AI as a tool, arguing that the role of the coder is disappearing and being bundled into roles like product designer or architect. The autho…