Most people use ChatGPT like a smarter Google. These 5 prompt cheats unlock the 90% of the model no one's actually touching. Paste any of them in front of your normal prompt. That's it ๐
cheat code oneWhat it does: Tells ChatGPT to project forward based on everything it already knows about you. Works best when you've been talking to it for a while and memory is on. Game changer for money and career calls.
PREDICT: Based on everything you know about me from our chat history and memory, project forward 12 months on [insert decision or situation]. Be specific about which variables I'm probably underweighting and which of my assumptions are most likely to break. No hedging. Give me your actual read.
Example use: "Should I take the corporate offer or keep building my brand full-time?"
What it does: Snaps ChatGPT out of its default agreement loop. No more "great question," no more "you're absolutely right." Actual pushback.
KILL CRITIC: For this entire conversation, stop agreeing with me by default. Push back hard on the weakest part of what I just said. If I'm wrong, say I'm wrong. Skip the "great question" intros, the flattery, and the "on the other hand" both-sides framing. Be the friend who tells me the truth, not the assistant who keeps me happy.
Example use: Paste your business idea, your offer, or a draft email and tell it to tear in.
What it does: Forces plain-language breakdowns of any topic. No jargon, no fluff, no terms you have to Google mid-sentence.
EXPLAIN LIKE I'M 10: Explain [insert topic] like I'm 10 years old. No jargon. No buzzwords. Use a real-world example a kid would actually relate to. If you have to use a technical word, define it in one short sentence right after.
Example use: "Explain MCP servers and how they connect to Claude."
What it does: ChatGPT's version of Claude's Ultra Think. Forces it to slow down, reason deeper, and stop giving you the surface-level answer.
10X THINK: Take 10x longer to think through this before you answer. Stress-test your first instinct. Consider angles I didn't ask about. Show me the reasoning, then give me the conclusion.
Example use: Any decision where the first answer isn't going to cut it. e.g. "Should I niche down my brand or stay broad?"
What it does: Strips the AI tells out of any writing. No em dashes, no "it's not just X, it's Y" patterns, no perfectly balanced 3-item lists.
HUMAN: Rewrite this so it reads like a real person wrote it. No em dashes. No "it's not just X, it's Y" patterns. No perfectly balanced 3-item lists. No "in conclusion." Use contractions. Vary sentence length. Sound like a friend texting, not a press release.
Example use: Paste any AI-written email, caption, or LinkedIn post.