This week I planned to publish only two videos on YouTube channel, as I was busy with non-AI work on other channels, but suddenly the new Grok model was released, so hurried up to test it! This video, and other news from the community, are in the links below.
My YouTube Videos
What I published over the last week:
AI Coding News, Tools, and Opinions
A usual, a week of small mini-releases of new features, and people reporting various experiences with different AI tools.
Claude Code default recommendation is now Sonnet 4 with 1M context. Great for bigger/longer prompts.
Also, Claude Code will soon have two new features: a Web UI and more “formal” track of progress (so does it mean no more “Flabbergabbering…” and other weird words?..)
In Cursor, you can now launch agents directly from Linear.
Also, Cursor v1.5 is out for everyone, including native OS notifications (finally!).
Lee Robinson from Cursor has published 1-hour long YouTube video Refactoring a Next.js & Tailwind app with Cursor.
While we’re talking about Claude vs OpenAI models… why not BOTH? Here’s a Reddit post: GPT-5 has been surprisingly good at reviewing Claude Code’s work
Another Reddit post, this time for Cursor users: Using Cursor Pro like a Pro
Andrej Karpathy (same guy who coined the term “vibe coding”) has shared how he actually uses AI at the moment. ~75% of his LLM assistance continues to be just (Cursor) tab complete.
Claude Code PM: Project management system for Claude Code using GitHub Issues and Git worktrees for parallel agent execution.
Finally, two articles on the similar topic: quality of Claude Code. What makes Claude Code so damn good and Peeking Under the Hood of Claude Code.