Full-text search across 67361 articles. Combine with topic and date filters; results sorted by relevance.
A developer released a self-contained C++ testbed on GitHub that compresses RGB textures as a shared low-resolution latent texture plus a tiny MLP decoder, trained entirely with Evolution Strategies rather than backpropa…
Docker Captain and AI Engineer Yhary from Colombia built a Docker Sandbox Kit mixin that automatically starts an MLflow tracking server for Claude Code, enabling out-of-the-box experiment logging and model versioning. Th…
Google released Gemini 3.8 Flash on Wednesday, its fourth Flash model in 106 days, claiming it matches larger rivals' performance on some coding benchmarks at lower cost, but the flagship Gemini 3.5 Pro, promised by CEO …
Rohit Edathil released Dagic, a typed DAG language and async execution engine that lets LLM agents compose registered tools into a typed DAG instead of calling them one at a time, avoiding the need to sandbox arbitrary g…
Amazon Web Services announced that Australian teams can now access OpenAI's GPT-5.6 Sol, Terra, and Luna models through Amazon Bedrock with global cross-Region inference from the Asia Pacific (Sydney) and Asia Pacific (M…
Anthropic's documentation for Claude Fable 5.1 and Claude Mythos 5.1 details behavioral differences and prompting patterns, advising users to test all effort levels (low, medium, high, xhigh, max) because effort level na…
A developer explains harness engineering, a methodology for building reliable AI agent systems by designing deterministic guardrails around large language models. The approach, highlighted by an OpenAI experiment where t…
Meta released Muse Spark 1.3 on September 2, 2026, into Muse Code and the Meta Model API, claiming it uses ~20% fewer tool calls and ~25% fewer tokens than Muse Spark 1.2, while Artificial Analysis found it consumes abou…
A developer's hands-on comparison of Claude Sonnet 4.5, GPT-5, and DeepSeek V3.2 finds Sonnet 4.5 best for refactors and agent workflows, GPT-5 fastest at 0.9s p50 latency on 200K context, and DeepSeek cheapest at $0.27 …
A new report on machine learning systems warns that secrets such as API keys and credentials are often mishandled, exposing applications to security risks. The report emphasizes the need for environment isolation, where …
GitHub retired six AI models from GitHub Copilot on September 1, 2026, including Claude Opus 4.5/4.6, Claude Sonnet 4.5/4.6, Gemini 3.1 Pro, and Raptor mini, with MAI-Code-1-Flash following on September 10, breaking inte…
A developer's hands-on comparison found that Grok (xAI) outperformed Cursor paired with Claude 3.5 Sonnet in catching a subtle race condition in asynchronous database calls, though it showed a slightly higher hallucinati…
Software engineers should be evaluated on "value over replacement" as coding costs drop to roughly $100 a month for tools like GPT-5.6-Sol and Claude Opus 5, according to a blog post by an unnamed author. The post argues…
OpenChamber, a free, MIT-licensed graphical front-end for the OpenCode AI agent, reached the Hacker News front page with 103 points and 54 comments, offering a multi-run feature that runs up to five models in parallel an…
An engineer behind the InversifyJS dependency injection framework built a new tool called rflct in a single night after an architectural proposal was dismissed as 'AI slop' on Reddit. The tool generates ahead-of-time ref…
A developer detailed how losing a Confluence documentation space led to a new architectural philosophy called 'Documentation as a Contract' for their startup. The approach uses git-backed markdown files as the source of …
The OpenSearch project has introduced Agent Skills, executable workflows packaged as SKILL.md files that give AI coding agents the domain expertise to build, configure, and deploy OpenSearch applications from natural lan…
Z.ai confirmed it built Ox Alpha, the anonymous model that topped OpenCode's usage charts, and will release its weights. The model, traced to Zhipu's GLM line, became the most-used on the open-source coding agent, nearly…
A developer built a regression test harness to evaluate AI code reviewers by seeding known bugs into pull requests. The harness, which uses the OpenAI chat-completions API, checks whether the model flags the planted bug,…
Clojurists Together's Q2 2026 funded project iLLaManati, led by Dragan Djuric, delivered a high-performance local LLM library in 100% Clojure, achieving 16 tokens/second on a 7-year-old CPU and 80+ tokens/second on an Nv…