AI for Coding
AI is reshaping software development faster than any other profession — from Copilot autocomplete to agents like Claude Code and Cursor that write, fix and refactor whole projects. "Vibe coding" — building software through natural-language dialogue — has entered the mainstream. Here is everything on AI coding tools: releases, benchmarks (SWE-bench) and the debate over quality and productivity.
Microsoft showed how to run VibeVoice for ASR, realtime TTS, and speech-to-speech
Microsoft released a practical Colab guide to VibeVoice: it walks through speaker-aware ASR, realtime TTS, batch audio processing, and building a speech-to-speech…
Why a human working with ChatGPT and Claude performs better than blind trust in AI
KDnuggets explains why real work with AI starts not with a prompt, but with verification, context, and a clear division of roles between the human and the model.
Microsoft veteran ran a transformer on a 6 MHz PDP-11 with 64 KB of memory
Former Microsoft developer Dave Plummer trained a tiny transformer on a 47-year-old PDP-11 and showed that the essence of AI is not magic, but optimized arithmetic.
Habr details an AI framework for Claude with Clean Architecture and a TDD cycle
Alibaba releases Qwen3.6-35B-A3B — a multimodal MoE model focused on agentic coding
Machine Learning Mastery explained how to build long-context RAG without extra tokens
Telegram chats became a stable lead generation channel with a 5,000-ruble AI bot
OpenCode, Ollama, and Qwen3-Coder: how to run a local AI coder without the cloud or limits
The OpenCode, Ollama, and Qwen3-Coder stack makes it possible to deploy a local AI programming assistant right on your own computer — free, offline, and with no…
zymi framework proposes building AI agents like a dbt project with YAML and event sourcing
Habr described zymi, an experimental framework where multi-agent pipelines are defined with YAML configs and all agent actions are recorded to an event-sourced data…
Waydev: the race for tokens reduces returns from AI coding and inflates rework volume
Cursor in talks for a $2B+ round at a $50B valuation amid enterprise growth
How a Claude-based Telegram assistant eased the CEO's load and took on part of the team's work
AI agents instead of vibe coding: how an autonomous pipeline takes a task to PR
KDnuggets named 10 GitHub repositories that help users master Claude Code more deeply
KDnuggets compiled 10 GitHub repositories that help users learn Claude Code through real workflows, templates, subagents, system prompts, and ready-made…
OpenAI explained how to configure Codex: personalization, level of detail, and access permissions
OpenAI released a guide to Codex settings: where to enable personalization, how to choose the level of detail for its work, and why, for long tasks, it is important…