Large language model (LLM) news — GPT-4, Claude, Gemini, Llama, Mistral and the latest research on training, fine-tuning, RLHF, and deployment of LLMs.
A Hacker News user reports that Google's AI Mode and AI summaries have disappeared for users in the EU, but reappear when using a US VPN. The user speculates this is due to a recent EU decision holding Google accountable…
A study funded with $120 in API credits found that five AI coding agents, including GPT-5.4 and Claude models, refused to fact-check documentation when given a stale doc, leading to task failure rates of 68-100%. The age…
A developer built CarbonWise, a carbon footprint awareness platform using Flask, SQLite, and the Groq LLM API (Llama 3.1). The platform combines deterministic carbon calculations with personalized AI reduction strategies…
Independent tests show the viral 'Caveman' AI tool, which claims to cut Claude Code token usage by 75%, actually reduces overall costs by only 4-10% for most coding tasks, with savings limited to conversational output. C…
A production incident involving an AI agent went undetected because traditional web service monitoring failed to catch a prompt cache invalidation that doubled costs and degraded answer quality while returning 200 OK sta…
The creator of Claude Code declares that prompt engineering is no longer the primary skill for AI work, introducing 'loop engineering' as the new essential competency in June 2026.
A user's request to access a Reddit post about a suitcase robot with a gas sensor wired into an LLM sampler was blocked due to a network policy. Reddit prompted the user to log in or register with developer credentials t…
A 3-billion-parameter model from Weibo scored 94.3 on the AIME 2026 math Olympiad, outperforming Google's Gemini 3 Pro which scored 91.7. The model is MIT-licensed and its success challenges assumptions about model size …
Anthropic released five agent design patterns that serve as a vocabulary for design decisions, and the Claude Agent SDK provides implementations for each pattern. The patterns help developers build any agent by mapping d…
A field guide details the ten most common errors developers encounter when using the Claude Agent SDK in production, organized by symptom rather than feature. The guide aims to help engineers quickly diagnose failures wi…
MLOps Data Engineer traces unreliable AI agent failures to poor knowledge retrieval, arguing that Retrieval-Augmented Generation (RAG) is the foundation for reliable agents. The article explains why chunking documents in…
An economist with no software engineering background built a Retrieval-Augmented Generation (RAG) research tool to help analysts quickly extract insights from dense financial and policy documents. The tool breaks PDFs in…
Researchers are advancing AI alignment by using Reinforcement Learning from AI Feedback (RLAIF) with Proximal Policy Optimization (PPO) to train language models, replacing expensive human annotations with AI-generated fe…
Spring AI, the official Spring framework for building AI-powered Java applications, offers Java developers a unified interface to integrate with multiple AI providers like OpenAI, Google Gemini, and Anthropic Claude. The…
Spring AI 1.0 reached general availability in May 2025, becoming a practical framework for Java developers building AI-powered applications like RAG pipelines and tool-calling agents. Five open-source GitHub repositories…
A developer outlines the case for running local private AI models, citing the recent shutdown of Anthropic's Fable 5 by the US government as a wake-up call. The post details four hardware options ranging from $3,000 to $…
A developer at Riversea Lab accidentally ran Claude Code with --dangerously-skip-permissions on a production database after misconfiguring an environment file. The multi-agent loop hit 12K KV writes per minute due to a c…
A developer built NoFlattery, a single-file React SPA that hosts a council of AI agents with different reasoning biases to debate user decisions. The app generates a Decision Record with a verdict, reasoning, risks, and …
Subrata Kumar Das published a lesson on Generative AI as part of a new course, explaining the evolution from traditional rule-based software to modern AI agents. The lesson covers the shift from predictive AI to generati…
A developer tracked writing output over 90 days and found that switching from Obsidian-first drafting to Claude-first drafting nearly doubled usable words per 90-minute block, from 420 to 780. However, the AI drafting in…