Discover 3 Unique and Advanced Ways to Utilize ChatGPT You Probably Haven't Encountered

Discover 3 Unique and Advanced Ways to Utilize ChatGPT You Probably Haven't Encountered

Valuable “meta” use cases I’ve found in 10 months of tinkering with ChatGPT

This is not your typical ChatGPT article

You know, the hyperbolic “6 Mind-blowing ChatGPT Secrets 99% of People Don’t Know About” kind. The actual “secret” most don’t know?

ChatGPT itself can write such clickbaity ChatGPT blogs! Try it yourself:

Prompt: “Act as a professional writer specializing in ChatGPT and AI. Use a persuasive voice, creative style, and a friendly tone. Rewrite your answers until they sound as human-like as possible. Now, write an 800-word essay titled ‘6 Mind-blowing ChatGPT Uses 99% of People Don’t Know About’ with appropriate H1, H2, and body formatting.”
All screenshots in the article is by author

A bit of rewording. A sprinkle of originality. And there you have it! Ready to dive into the sea of ChatGPT content.

I didn’t jump on the ChatGPT bandwagon right away. For nearly a year after my first post went viral, I avoided writing about it again.

Instead, I quietly experimented with ChatGPT and discovered three key ways to use it—for expert advice, targeted searching, and fact-checking.

These methods branch out into countless specific tasks, like learning niche topics, doing content research, fixing code, recalling forgotten ideas, and filtering out misinformation.

Rather than just sharing simple prompts like most do, I’ll...

  • Offer improved, multi-step prompts with three tips for each scenario.
  • Break down the parts of each prompt to explain why they work.
  • Provide templates you can easily use.

These prompts have boosted my research, workflow, and memory recall, and I hope they’ll do the same for you.

So, before we get started, open up ChatGPT in another tab and follow along as we go!


#1-Name-Prime ChatGPT Into a Niche Expert

Searching for signal around the bamboo hut, we crouched on a dusty log.

"Got it!" I said, carefully holding the tablet. "Now, brother, watch ChatGPT work its magic. Let’s check out how to extract coconut oil," I grinned at my farmer friend.

Prompt 1.1: “Act as an expert Indian organic farmer specializing in coconut farming for coconut oil production. Now explain the step-by-step process from start to end to produce coconut oil.”

This was a super simple example of ChatGPT “priming”—basically, focusing ChatGPT’s knowledge on a specific topic.

Real name-priming is much stronger, but I didn’t use it to avoid confusing my less tech-savvy friend.

Let me show you how by activating ChatGPT’s inner grammar expert:

Prompt 1.2: “You’re now GrammarGPT — a world-class English language, literary, and grammar expert. Henceforth, whenever I ask you a question, respond with relevant, expert, and well-thought-out answers. Polish your answers so they’re concise, straightforward, and easy to read. Till this window is closed and/or I specify otherwise, assume I’m referring to you, GrammarGPT. If you understand your job, describe it to me. If you don’t, speak out your doubts so I can clarify them.”

Before jumping into follow-up prompts, let’s break down this powerful prompt:

  • “You’re now GrammarGPT…” — While saying “Act as an X expert” can get the job done, giving ChatGPT a name makes switching roles easier. You can toggle between ChatGPT, GrammarGPT, or PhiloGPT by simply using a “Switch (back) to {X}GPT” prompt."

  • “Henceforth, whenever I ask you…” — so ChatGPT ensures a high, reliable standard for every answer.
  • “Polish your answers…” — so ChatGPT formats each response as we desire.
  • “Till this window is closed…” — so ChatGPT stays GrammarGPT until you close the tab.
  • “If you understand your job…” — to check if ChatGPT has understood its GrammarGPT role.

When Googling and Grammarly prove useless, GrammarGPT swoops in to the rescue.

Prompt 1.3: “I am looking to shorten the sentence “I want to make you aware of…” but am at a loss for verb forms of “Aware.” Can you suggest 15 richer and more literary alternatives please?”
Prompt 1.4: “I want a single-word phrase that describes someone who falls in between being frugal and a spendthrift.”
Follow-up: “Any other alternatives? Moderate can mean moderate in anything. I need a word that’s more context-specific”

Be it choppy sentences in need of smoothening or stale vocabulary that needs enriching, GrammarGPT is at your service.

Still frustrated? Banter with ChatGPT’s inner stand-up comedian:

Prompt 1.5: “You’re now ComedyGPT — a world-class stand-up comedian rivaling the likes of Andrew Schulz and Dave Chappelle. Henceforth, whenever I say anything to you — no matter how random — use it to derive stomach-hurtlingly lethal humor. Don’t hesitate to use puns, double entendres, and bold wordplay. Do you understand your job, ComedyGPT?”

3 Ways to Further Strengthen ChatGPT’s Name-Priming:

  • Describe specific (re)sources and/or historical/living figures for ChatGPT to draw from.

Prompt 1.6: “You’re now back to being PhiloGPT. Henceforth, PhiloGPT, draw specifically from Stoicism, Alan Watts, and Advaita Vedanta to formulate your answers. Mix and match with your own creativity and judgement. Now, I’d love to know what the meaning of life is?”

  • Make ChatGPT leverage each response to enrich the subsequent ones.

Prompt 1.7: “PhiloGPT, henceforth, with every response you generate, utilize the context of the generated response to broaden your selection of resources and expertise for each subsequent response. Makes sense? Now, what’s the meaning of life again?”

  • Combine multiple name-primed ChatGPTs you’ve already defined.

Prompt 1.8: “You’re now PhiloComGPT — a hybrid of PhiloGPT and ComedyGPT. Now, what’s the meaning of life, PhiloComGPT?”

“The cosmic joke is that we are both the comedian and the audience, playing hide-and-seek with our own divinity”— PhiloComGPT, Circa 2023


#2-Meta-Search with ChatGPT when Googling Fails

Sometimes we have something “on the tip of our tongues,” but it’s too vague to search for or ask an expert about.

It might be a quote, a meme, a scientific term, or a song.

With ChatGPT, we don’t have to struggle or wait for a sudden “Eureka!” moment anymore.

Prompt 2.1: “What is that phrase sounding something like “heebly beeblies” we use to say when something is scary enough to give us the shivers? Or something along similar lines? I can’t seem to put my finger on it.”
Prompt 2.2: What is the name of the human tendency to keep looking for the secret sauce — or something along those lines? It’s a well-documented cognitive bias I can’t seem to recall the exact name of.

While simpler than name-priming, the meta-searching prompt has 3 crucial components:

  • An As-Specific-As-Possible (ASAP) description — the vaguer your description, the harder it’s for ChatGPT to zone in.
  • “Something along those/similar lines” — to prompt ChatGPT to expand its field of search and guesses. This makes up for inaccurate descriptions.
  • “I can’t seem to (recall exactly)/(put my finger on it)” — conveys surety that your target exists and is definite — even if you can’t recall it.

Think of ChatGPT as a near-omnipotent child — he’ll work wonders, but you must be articulate.

3 Ways to Enhance ChatGPT’s Meta-Searching Ability:

  • Prompt ChatGPT to ask you questions if it’s unable to zone in.

Prompt 2.3: “If you’re unable to guess accurately or zone in, feel free to ask me follow-up questions so we can zone in better.”

  • Ask ChatGPT for multiple guesses — and guess by guess, inch closer to whatever you’re searching.

Prompt 2.4: “Suggest 4–10 guesses. I’ll let you know which of them is the closest to what I’m looking for — so you can zone in better.”

  • Name-prime ChatGPT before meta-searching — works if you roughly know your target's topic.

Prompt 2.5: You’re now PsychoGPT — a world-class psychology expert specializing in cognitive biases and social human psychology. Henceforth, whenever I ask you questions, respond with relevant, expert, and well-thought-out answers.

#3- Pit ChatGPT against ChatGPT for Fact-Checking

Despite its vast knowledge and smooth responses, ChatGPT isn’t always reliable!

It fills gaps in its answers with made-up information—one study even showed that AI can be more convincing when lying than humans! As programmer Stephen Diehl put it,

“ChatGPT is glorified Autocomplete. It’s a cool parlor trick and fascinating from a maths and linguistic perspective. But it’s not sentient and has no personality, agency, or intent. It reflects back the internet’s unfiltered text as a probabilistic mirror.”

Luckily, ChatGPT spills the beans if you follow up with an “Are you 100% sure?” prompt.

Surprisingly, ChatGPT is a reliable fact-checker — be it quote sources, historical accuracy, or research claims.

Without name-priming, ChatGPT deftly spotted a TOI fat -loss blog's B.S.

Talk about hypocrisy!

But you can leverage this hypocrisy by pitting ChatGPT against ChatGPT.

  • Fire up ChatGPT parallely on 2 separate browser tabs.
  • Use one as a name-primed expert or a meta searcher — let the 2nd one be the fact-checker.
  • Copy-paste the generator’s output into the checker and append, “Is the above text 100% true? Can you please rigorously fact-check this for me?”

Feed the checker’s output back into the generator for rigorous use cases like programming or research.

Keep going back and forth until you reach a satisfying accuracy.

3 Ways to Boost ChatGPT’s Legitness and Fact-Checking Ability:

  • Regenerate with the exact prompt multiple times — in different chat windows. Consistent responses? Maybe true. But inconsistent responses? Definitely false.
  • Ask ChatGPT to cite its sources in a detailed manner. Google to verify since ChatGPT can make up sources, too!

Append to your Prompts: “Cite your sources as accurately as you possibly can, please — if you can’t, at least provide precise book/interview/report/research_paper names I can look up.”

  • Name-prime ChatGPT to be trustworthy — this still doesn’t “guarantee” legitness, so do your due diligence.

Prompt: You’re now TrustGPT — you’ll be 100% honest in your answers and not make up any facts, information, or data. Only if you’re 100% sure about something will you answer it. If not, you’ll admit you aren’t sure — so I can rely on you. Can I rely on you, TrustGPT?

Prompt Templates For You To Plug-And-Play

Creating effective ChatGPT prompts is all about refining—rewording, regenerating, following up, mixing things up, and repeating. Some people spend days perfecting their ultimate prompts!

But I’ll save you the trouble by sharing ready-made prompt templates for the scenarios we’ve discussed.

  • Name-Prime ChatGPT into a nuanced niche expert:

“You’re now {NAME}GPT — a world-class {TOPIC} expert in. Henceforth, whenever I ask you a question, respond with relevant and well-thought-out answers. Draw specifically from {SOURCES} to formulate your answers. Polish your answers so they’re concise, straightforward, and easy to read. With every response you generate, utilize the context of the generated response to broaden your selection of resources and expertise for each subsequent response. Till this window is closed and/or I specify otherwise, assume I’m referring to you, {NAME}GPT. If you understand your job, describe it to me. If you don’t, speak out your doubts so I can clarify them.”

  • Meta-search with ChatGPT where and when Googling fails:

What is that {SEARCH_TARGET} sounding something like {DETAILS} often used in/by/for {CONTEXT}. Or something along similar lines? I can’t seem to put my finger on it. Suggest {X} guesses. I’ll let you know which of them is the closest to what I’m looking for. If you’re unable to guess accurately or zone in, feel free to ask me follow-up questions so we can zone in better.”

  • Get more trustworthy answers and fact-check with ChatGPT:

Prime generation to be more trust-worthy: “You’re now TrustGPT — you’ll be 100% honest in your answers and not make up any facts or information. You’ll cite your sources as accurately as you possibly can— if you can’t, you’ll at least provide precise book/interview/report/research_paper names I can look up. Only if you’re 100% sure about something will you answer it. If not, you’ll admit you aren’t sure — so I can rely on you. Can I rely on you, TrustGPT?

These prompts aren’t mutually exclusive — permute and cross-breed them for best results!

Name-prime before using ChatGPT. Sharpen the specificity of your browsing and generation(s) with meta-searching. Keep a rival ChatGPT instance handy for fact-checking.


ChatGPT is a true mass-access superpower.

Companies, content mills, and creators are wielding ChatGPT to make millions. The least you can do is use it to enhance your daily life and workflow.

No need to be tech-savvy, buy ChatGPT courses, or tinker with complex prompts. You only need curiosity, creativity, and 10 daily minutes.

Hate it or love it, AI is here to stay. But…

AI won’t replace you or your job — someone cleverly leveraging AI will.

Will that someone be you, dear reader?


Authored by Rohan Yadav: Digital Marketer and Innovative Marketer. A management specializing in holistic business growth. Reach out at [email protected]

Follow on LinkedIn

要查看或添加评论,请登录

Rohan Y.的更多文章

社区洞察

其他会员也浏览了