Distributed backend specialist. Perfectly happy playing second fiddle—it means I get to chase fun ideas, dodge meetings, and break things no one told me to touch, all without anyone questioning it. 😇
I 100% agree that you should never blindly prompt or accept AI output without meeting a few basic requirements:
Every line of code deserves a personal review and test. We call it HITL (human-in-the-loop) review.
AI absolutely repeats what it’s been trained on—but that’s not trivial. Its training spans an enormous breadth of knowledge. It just relies on you to define which parts are relevant right now.
Working effectively with AI takes practice. It’s not just about the prompt:
It’s about the instructions you built ahead of time to guide its behavior.
The constraints you’ve defined and the non-negotiable truths behind every success case.
Your ability to manage context: what the model sees, when it’s relevant, and when it isn’t.
Even model selection changes everything.
And above all, your ability to steer the conversation decides whether AI amplifies your expertise or multiplies your mistakes.
While I agree with most of your points, I think some of your framing expects precision from AI that you never taught it. For example, you said:
“It breaks them because it doesn’t know what systems are.”
“It doesn’t see your deadlines, your users, or your bills — it only sees text.”
That’s only true because you didn’t tell it.
In my environments, AI knows all of that—because I taught it.
It knows what a valid inline comment looks like, when traffic spikes, which tests add value versus noise, what tone belongs in documentation, and which pre-commit checks must pass. It knows these things because they live in the repo instructions.
I’ve also learned how to manage context without throwing Copilot into the deep end. I know when to interrupt, question, or redirect. I configured it to read terminal output and validate results before it closes a turn.
We debate architecture, debug traces, explore alternatives, identify logic gaps, and write code that I can personally verify as secure, maintainable, and production-ready.
AI isn’t magic—but it’s also not a “prompt-and-pray” sort of situation. It’s a tool that rewards structure, feedback, and discipline. You wouldn’t roll out a brand-new teammate straight into prod; why expect any LLM to ship perfect code on day one?
So, on your final points:
“Optimized for fluency, not truth.” True only if you never define what “truth” means in measurable ways.
“Never run code you didn’t verify.” Absolutely. Human review is non-negotiable—AI or not, your name’s on the commit.
“Use ChatGPT for thought, not execution.” Mostly right. ChatGPT isn’t a coding agent—it’s conversational. But GPT-5 can produce accurate, reliable code inside the right environment.
“Keep your backups sacred.” Always. That’s standard AI or not.
“Treat AI like electricity.” Perfect analogy—it’ll burn your house down if you wire it wrong. Wire it right, and it powers everything. ⚡
PS: If you or anyone else ever wants to learn more about how to use AI the right way, reach out! I'm happy to share every single thing I know! A lot of it is already documented in various posts if you pull my profile. 😉
For further actions, you may consider blocking this person and/or reporting abuse
We're a place where coders share, stay up-to-date and grow their careers.
I 100% agree that you should never blindly prompt or accept AI output without meeting a few basic requirements:
While I agree with most of your points, I think some of your framing expects precision from AI that you never taught it. For example, you said:
That’s only true because you didn’t tell it.
In my environments, AI knows all of that—because I taught it.
It knows what a valid inline comment looks like, when traffic spikes, which tests add value versus noise, what tone belongs in documentation, and which pre-commit checks must pass. It knows these things because they live in the repo instructions.
I’ve also learned how to manage context without throwing Copilot into the deep end. I know when to interrupt, question, or redirect. I configured it to read terminal output and validate results before it closes a turn.
We debate architecture, debug traces, explore alternatives, identify logic gaps, and write code that I can personally verify as secure, maintainable, and production-ready.
AI isn’t magic—but it’s also not a “prompt-and-pray” sort of situation. It’s a tool that rewards structure, feedback, and discipline. You wouldn’t roll out a brand-new teammate straight into prod; why expect any LLM to ship perfect code on day one?
So, on your final points: