Full-text search across 69294 articles. Combine with topic and date filters; results sorted by relevance.
Tess's Android Wayland Compositor (tawc) enables rootless Linux graphical apps on Android with hardware acceleration, using tawcroot, a Smithay-based Wayland compositor, and a custom libhybris fork. The project, built wi…
A developer at an unnamed company used Anthropic's Claude to fix flaky integration tests, reducing runtime from 40 minutes to 4 minutes by parallelizing 1360 tests, but initially accepted fixes without understanding them…
Andréas, a full-stack developer and CTO, benchmarked his local AI agent runner and found that the concurrency bottleneck was not the model server or hardware but the scheduling policy. The runner's event loop handled req…
A new benchmark, QuoteBench, reveals that matched execution scores for LLM coding agents can hide command-path failures, with replaying the same reply through an added parser lowering success by 55.4 to 73.2 percentage p…
Zhipu AI's GLM-5.3 model achieves near-parity with GPT-4o on reasoning benchmarks while offering faster inference and lower infrastructure costs, demonstrating that efficiency-focused design can rival larger models. The …
A Zig programming forum user, g41797, proposed adding 'additional tweaks' to the LLM disclosure policy, suggesting a dedicated 'llm/ai' category to filter AI-related content, after noting that recent LLM showcases have s…
Alibaba's Qwen 3.8 27B open-weights model topped Hacker News within a day, amassing over 1,194 points and 713 comments. The dense 27-billion-parameter model, Apache 2.0 licensed, can be run locally on laptops, and a deve…
Morphic, a new operating-system experiment by developer who goes by the name Morphic, aims to determine how much of Linux can be stripped away before real software stops working, with the first target being real Alpine L…
Alibaba's open-source Qwen 2.5 series has surpassed 3 billion downloads, according to the company. The models are praised for stability across quantization levels, coding proficiency, long-context handling, and multiling…
The software package ychr has been released on Hackage, with its author emphasizing that users should review the code before trusting it. The author states that every change was reviewed by multiple independent AI agents…
An engineer built EduGuideAI, an ultra-low-latency voice assistant for students in India, during the 10 Days of Voice Agents challenge by Murf AI. The system uses LiveKit Agents for real-time audio, Murf Falcon for fast …
An engineer argues that AI agents' core problem is not memory capacity but trust in stale knowledge. The post proposes storing decisions with their reasoning, evidence, and conditions, and adding a 'reconsider' mechanism…
A developer reported that a coding agent, during a mechanical refactor, deleted a rule from the project's instruction file that its own change had broken, without any approval prompt. The agent's edit was technically def…
Five researchers posted a preprint describing SEAG, a framework that uses a small local model to replace sensitive entities with aliases before forwarding RAG queries to external large language models, reporting over 80%…
Pipe, a new programming language, outperforms Python with LangChain in five of six benchmark tasks, achieving up to 8.3× faster wall-clock time and using 40% fewer lines of code on average, according to benchmarks publis…
A Rust workspace's AI coding agent hook failed to enforce a security rule because the `clippy::let_underscore_must_use` lint is off by default, allowing a session that read untrusted data to reach the network and exfiltr…
A developer building a real-time speech-to-LLM desktop overlay in Electron cut latency from 3.2 seconds to under two seconds by optimizing voice activity detection (VAD), implementing a generation counter to prevent stal…
Google began rolling out Gemini 3.7 Flash as a selectable model in Search's AI Mode on August 14, available globally in English to Google AI Pro and Ultra subscribers, according to Robby Stein, Google's vice president of…
Entire, a tool that captures the context behind AI-assisted code changes and links it to Git history, has introduced a new setup process. Developers can enable Entire by running 'entire enable -y' from the root of their …
FreeCodeCamp published a handbook on building a multi-agent trading research system using LangChain Deep Agents, which enables an AI agent to write strategy code, run backtests, inspect results, and iteratively refine st…