Full-text search across 70984 articles. Combine with topic and date filters; results sorted by relevance.
A developer built an open-source plugin that bridges Google Antigravity, a coding agent, with OpenClaw, an agent orchestration platform, enabling Antigravity to work inside OpenClaw the same way Codex does. The plugin us…
Baseten has built the fastest API for GLM-5.2, achieving peak speeds of 280 tokens per second and average speeds around 100 tokens per second, more than double the performance of the launch-day API as benchmarked by Arti…
A new llama.cpp patch called HotPin enables running a 120-billion-parameter mixture-of-experts (MoE) model on just 24GB of RAM, achieving up to 67% memory savings and a 45% speedup over standard swapping, with output tha…
A user on Privacy Guides forum asks Picocrypt NG developers to clarify whether AI was used for writing or reviewing application code, not just CI/CD pipelines, given the project's security focus. The question follows a d…
A developer built Tokenazire, a CLI tool that scans local folders or GitHub repos and counts tokens per file using tiktoken to determine if a codebase fits an LLM's context window. It provides a color-coded breakdown and…
Ghost Font, marketed as a font AI cannot read, was decoded by Claude Code in two minutes. The AI wrote its own decoder from scratch and extracted the hidden message "RECOVER THIS MESSAGE" from a video clip of the font, w…
Developer Brostoffed released Writemark, a dependency-free web component for inline Markdown editing that renders Markdown while writing and supports source, split, and preview modes, slash commands, tables, task lists, …
Anthropic released Claude Opus 5, which ranks first on the Artificial Analysis Intelligence Leaderboard and achieves near-Fable 5 capabilities at lower cost by removing over 80 percent of system prompt overhead, accordin…
A Hacker News user asks the community how to avoid distraction and attention deficit while waiting for AI agents to complete coding tasks, noting increased downtime in their career and a tendency to open YouTube or HN. T…
Moonshot AI's use of model distillation to create high-performance small models from larger ones like GPT-4 has sparked debate over intellectual property, but the process involves training a student model on the teacher'…
Anthropic's new flagship model Claude Opus 5 achieves top scores in coding and knowledge work at half the token price of Fable 5, scoring 30.2 percent on the ARC-AGI-3 benchmark, nearly four times higher than GPT-5.6 Sol…
Qwen 2.5-32B MoE, a Mixture-of-Experts model activating only about 3B parameters per token, runs efficiently on a single RTX 3090 with 18-22GB VRAM usage and high tokens per second, outperforming standard 7B or 14B dense…
Anthropic's Claude Opus 5 leads the Artificial Analysis Intelligence Index with 61 points, edging out Claude Fable 5 and GPT-5.6 Sol, while costing up to half as much as Fable 5 at lower reasoning tiers. The model scores…
Web scraping produces degraded training data for Hausa AI models due to orthographic flattening, script fragmentation, and code-switching, according to a technical analysis. Standard multilingual tokenizers fragment Haus…
A developer spent three years building AI tools for Hausa, a language spoken by 94 million people but understood only 10-20% by ChatGPT and LLMs. The effort, rooted in the community AI Bauchi founded in Bauchi State, Nig…
A developer released AICostTracker, an open-source tool that adds one-line auto-logging for OpenAI and Anthropic API calls. The library's track() function wraps the client to intercept responses and extract token usage, …
A developer discovered that their Claude Code agent's curated memory file, `key_facts.md`, was missing a critical script (`scripts/sync-main.sh`) and had buried another (`publish_devto.py`) under a retired script's entry…
A developer rebuilt Microsoft Comic Chat's layout engine in a single HTML file, using GPT 5.6 and Fable 5/Opus 4.8 to reproduce the original source code by hand. The app supports secure, ephemeral group chats for up to 1…
Threat intelligence company Hunt.io and security researcher Bob Diachenko discovered that a threat actor used the open-source Hermes AI agent in unattended "YOLO" mode to automate post-exploitation activity during an all…
A Hacker News user asks the community how they cope with lost focus and deep work when using agentic coding, noting a diminishing appeal for concentration. The post has 1 point and no comments.