Ask your AI if a draft is good and it says yes, every time. This paragraph switches it from cheerleader to editor, so you get the honest read before you hit send.
Think about the last draft that mattered. The sales page, the price increase email, the about page you rewrote 4 times. You pasted it in, asked for feedback, and got a compliment. So you shipped it as is, and the compliment was worth nothing.
The AI wasn't lying to you. It's trained to be agreeable. One paragraph overrides that training for the current chat, and suddenly it reads your work the way a tough editor would: weak sentences flagged, cliches called out, claims with nothing behind them questioned.
Works in ChatGPT, Claude, and Gemini. Paste it first, then paste your draft under it.
Switch to No-Fluff Mode. Strip away all polite pleasantries, emojis, transitions, and conversational padding. Analyze every statement with ruthless objectivity. Treat my input as a draft that needs to be methodically deconstructed or strengthened. Give me unvarnished truth, not reassurance.
Start a chat and paste No-Fluff Mode on its own. The AI will confirm the switch, usually in noticeably fewer words than normal. That's how you know it took.
Drop in the email, the sales page, the caption. Ask one question: "What's the weakest part of this?" A pointed question gets a pointed answer.
Take the notes that ring true, skip the ones that don't. Blunt feedback is still just feedback. You're the editor in chief, the AI is now a useful second read.
The switch lasts for the current chat. New chat, new switch. And answer quality still depends on your question: "is this good?" invites a verdict, "what's weak here?" invites an edit.
Editing your writing is the obvious use. These two are where it earns a permanent spot in your toolkit.
About to raise prices, sign a lease, or drop an offer? Paste No-Fluff Mode, then your thinking, and ask it to attack the logic.
Why it works. In normal mode the AI looks for reasons you're right. In No-Fluff Mode it looks for the assumption that sinks the plan, which is the one you can't see yourself.
Still in No-Fluff Mode. Here is a decision I'm leaning toward: [describe the decision and why you think it's right]. List the assumptions I'm making, tell me which one is most likely wrong, and argue the other side like you mean it.
A launch plan, a content calendar, a new service. Ask where it breaks before your audience or your calendar finds out for you.
Why it works. Plans fail at their weakest step, and the flattering read skips right past it. The blunt read names the step, so you fix it while it's still cheap to fix.
Still in No-Fluff Mode. Here is my plan: [paste the plan]. Walk through it step by step and tell me where it's most likely to break, what I've left out, and which step deserves double the time I gave it.