AI coding | Tech
Tech news, context and sources.
Latest articles
Claude Code review finds hidden issues in a Codex-built web app
Cross-model review can help, but one user's experience does not establish general quality superiority.
Agent-written systems still require fundamentals and diagnostic depth
Ownership depends less on having written the code than on detecting trouble and investigating deeply enough.
Android Bench separates task success from partial completion
Different agents and reasoning settings limit what a simple model ranking can establish.
You’ve reached the end.