A developer's Accenture BA exit exposes a structural flaw in India's IT services model — and why splitting focus between backend and data science is the wrong fix.
NanoEuler implements a complete GPT-2 scale transformer in pure C and CUDA — every attention kernel and gradient hand-written, no ML framework required.
DeepSpec from deepseek-ai is the first open framework to tackle speculative decoding's hard part: training draft models and benchmarking them consistently.
An anonymous actor used AI-assisted fuzzing to find 20 open source zero-days and published them without disclosure. Here's what that means for your stack.
OpenKnowledge launches as an open-source, AI-native knowledge base from Inkeep — 347 HN upvotes, real production AI infra behind it, and a business model worth understanding.
Alibaba's Qwen team open-sources a framework that replaces hand-engineered simulation environments with LLMs, and the trade-offs are non-trivial.
Recall adds fully local, session-persistent memory to Claude Code. 82 HN upvotes on launch confirm real developer pain — but stale memory may be worse than none.
Vercel's eve hit 2,009 GitHub stars on launch day. The real story isn't another agent framework — it's who controls the deployment layer.
Z.ai's GLM-5.2 scores near GPT-5.5 on benchmarks while hallucinating 3× less—new data shows why brute-force AI scaling breaks production reliability.
Vercel open-sourced 'eve', a TypeScript-first AI agent framework that hit 1,503 GitHub stars on day one. Here's what the hype obscures.