I recently switched from VSCode to Cursor and like having AI in the editor as an assistant but don’t like having it think for me.

There’s a term for generating code yet not verifying it: “vibe coding”.

🔹“Vibe coding, sometimes spelled vibecoding, is an AI-powered programming practice where a programmer surrenders to the ‘vibes’ and power of the AI, while ignoring the details of the generated code.”

🔹“…Ars Technica cites [AI researcher] Simon Willison, who stated: ‘Vibe coding your way to a production codebase is clearly risky. Most of the work we do as software engineers involves evolving existing systems, where the quality and understandability of the underlying code is crucial.’”

🔹“….Simon Willison said: ‘If an LLM wrote every line of your code, but you’ve reviewed, tested, and understood it all, that’s not vibe coding in my book—that’s using an LLM as a typing assistant.’”

I watched a video of Apple co-founder Steve Wozniak sharing some of his thoughts about AI with a reporter that, and I paraphrase, he thinks it’s great for people to be like editors who check what AI generates and pass its content along or reject its content entirely.

I ultimately view AI as an assistant, and I’m adopting this perspective as I think it’s healthier than thinking AI has free rein. I’m doing it because I believe people and LLMs can be at their best, complement one another, and work together 🧑🤝🤖