AI Just Wrote My Code While I Slept (Here's How!)

Video Description

AI coding tools are evolving fast—and they’re starting to feel like true teammates. In this video, I dive into the next big shift: coding agents that don’t just suggest code but actually take on full tasks while you step away. I walk through a real example using a small NFL scheduling app, showing how I can hand an issue to AI, and it returns with a complete pull request—tests included. From Claude Code to Gemini CLI to OpenAI’s Codex, we explore how these tools handle code changes asynchronously, and what that means for scaling your own work. Along the way, I cover: How GitHub issues, branches, and pull requests fit into this workflow Setting up AI integrations that actually do the coding for you The hidden costs (and free options) you should know about Why having tests in place makes or breaks this workflow What worked, what didn’t, and where this is all heading This isn’t hype—it’s a peek into how developers will scale themselves in the near future. Whether you’re curious about Claude, Codex, Gemini, or just want to see what “bots coding while you sleep” really looks like, you’ll find something useful here. #AI #Coding #ClaudeCode #OpenAI #Gemini #Codex #GitHub #CodingBots #CodeWhileYouSleep #AgenticAI 00:00 - Intro 01:11 - Intro App 01:22 - Intro GitHub 03:06 - The Change 03:53 - What is all of this 05:57 - Creating the Issue 07:00 - Installing our Agents 08:20 - Adding to GeminiCLI 09:26 - Tell the bots to work 11:25 - ChatGPT Codex time 12:54 - So, why? 16:58 - Claude's work 18:17 - Claude Results 18:50 - But what about Codex? 19:34 - Codex results 20:26 - Conclusion