If ChatGPT only praises you

@startuppakistanofficial

581.9K followers · usually 27.6K views a post

Views

361.7K

Likes

15.2K

Comments

86

Shares

1.2K

Selling · SaaS › AI Tool

ChatGPTseen in 540 videos →

Why it worked

The video leverages curiosity by presenting "secret codes" for a popular tool, promising to unlock advanced functionality. It uses relatable analogies (ex-partner, best friend) and a clear, numbered list format to deliver actionable tips in a concise and engaging manner.

Summary

The video introduces five "secret codes" or prompting techniques to enhance ChatGPT's responses. These techniques aim to make the AI provide more natural, detailed, and critical feedback, acting more like a helpful friend than just a praise-giver.

Structure

1Introduction to ChatGPT's limitations
2Introducing 5 prompting techniques
3Explaining /human for natural language
4Explaining 10xthink for detailed answers
5Explaining EL10 for simple explanations
6Explaining ALT3 for multiple perspectives
7

Original caption

ChatGPT Secret Codes 😱 #ChatGPT #ArtificialIntelligence #OpenAI #Tech #Technology #AITips #Productivity #Innovation #TechTrends

More from @startuppakistanofficial

Their other posts in the index, biggest breakout first.

Explaining Killcritic for critical feedback

8Concluding with a call to action

Product placement

ChatGPT: An AI language model that generates human-like text in response to prompts. It is used here to provide more natural and detailed answers. It acts by generating text and is the subject of the video.

Call to action

Try these smart prompting techniques and share with a friend who uses ChatGPT.

On-screen text

CP ChatGPT #1 /HUMAN MORE NATURAL ANSWERS #2 10XTHINK #3 EL10 #4 ALT3 #5 KILLCRITIC WEAKNESSES RISKS & FLAWS

Transcript

If ChatGPT only praises you, it keeps on doing it, right? Or like your ex, it only gives boring answers. So maybe the problem isn't with ChatGPT. I mean, it's with your prompt. I'm going to tell you 5 AI codes that will make ChatGPT act like a best friend, not just praise you, but also point out your mistakes. So save this video because later you'll say, "I wasted a moment, I wasted a moment." Oh, codes, I get lost. The first one is /human. Just write this before your prompt. Then ChatGPT will give answers in a more natural and human language. If you're writing emails, captions, or messages, the answer will feel a little less robotic. Next, 10xthink. Instead of basic or surface-level answers, ChatGPT will give answers with more detail, reasoning, and analysis. Third, EL10. With this, AI will explain even complex topics in simple language, like explaining to a 10-year-old. Fourth, ALT3. If you want multiple perspectives on the same question, use this, and you'll get three different solutions. And last is Killcritic. This is my favorite. Adding this will not make AI agree with your point, but it will honestly point out the weaknesses, risks, and flaws in your idea. These are not official ChatGPT hidden codes, but these are smart prompting techniques. Try them and also share with that friend who gives you fake advice like ChatGPT.