Club Benefits
After years of testing ChatGPT, each time a new model launches, despite how much smarter or faster it becomes, I've noticed something wild: AI tends to mirror the quality of your thinking.
I recently came across a fascinating Reddit thread asking whether ChatGPT can “recognize gifted users” and respond differently depending on the person using it. The discussion exploded because a surprising number of users said they felt the AI adapted to depth, nuance and reasoning style.
Some commenters described the chatbot becoming more analytical, collaborative or intellectually playful during longer conversations. Others argued that ChatGPT seemed to “mirror” the sophistication of the user interacting with it.
That doesn’t mean the AI is secretly identifying geniuses or ranking intelligence.
But it does highlight something important about modern AI systems, which is simply that they adapt heavily to conversational patterns. In essence, if someone communicates with more structure, context and specificity, the model usually has more material to work with. And because large language models are prediction systems trained on patterns, they tend to generate richer outputs when the inputs themselves are richer.
In other words, AI may not be rewarding intelligence itself nearly as much as it rewards clarity.
Prompt to improve the quality of responses every time
The more I use ChatGPT, the more I’ve realized that the best prompts create clearer thinking environments for the AI.
In many ways, these prompts work because they reduce ambiguity, slow down pattern-matching and force the model to reason more carefully instead of rushing toward the most confident-sounding answer.
And interestingly, almost all of my favorite prompts revolve around the same core idea of clarity.
1. The 'Unicorn' prompt
The exact prompt I use: "Pretend you’re my assistant and you actually want me to succeed. Ask up to 3 questions if anything’s unclear. Then give me: the answer, the plan and the pitfalls. Keep it short and tailored to: [insert goal]. If you have to make assumptions, list them first."
As one of the most useful prompts I’ve created, the "Unicorn" prompt essentially tells the AI to stop pretending it knows things when information is incomplete. Instead of confidently filling gaps, the model is encouraged to ask clarifying questions, acknowledge uncertainty and ultimately separate assumptions from facts. The Unicorn prompt works because it interrupts that behavior.
2. The 'Glitch' prompt
The exact prompt I use: "Pause — I think there’s a glitch. Check your last answer for mistakes, missing steps, false assumptions, or made-up details. Then rewrite the answer more accurately, and add a confidence rating (1–10)."
Short version: "Re-check and rewrite for accuracy. Add confidence rating (1–10)."
The "Glitch" prompt works because it encourages self-correction. It pushes ChatGPT to critique its own reasoning instead of immediately trusting its first response. From there, you'll notice that the AI produces noticeably better outputs because the model is forced to identify weak assumptions, evaluate contradictions and reconsider oversimplified logic. In other words, the AI stops behaving like a confident autocomplete engine and starts behaving more like a collaborator engaged in active reasoning.
3. The 'Owl' prompt
The exact prompt I use: “Think like an owl — slow, observant and analytical. Examine this problem from multiple perspectives and identify the hidden factors most people overlook.”
Another favorite is the “Owl prompt,” which encourages ChatGPT to take a slower, more deliberate approach before answering. This is surprisingly important because most AI models optimize heavily for speed and fluency. But fast answers aren’t always thoughtful answers. The Owl prompt works because it encourages the model to think step-by-step, consider alternative interpretations and pause before jumping to conclusions. And often, that extra layer of reflection dramatically improves the final response.
4. The 'Potato' prompt
The exact prompt I use: “Whenever I type the word ‘Potato’ followed by an idea or argument, ignore your helpful persona and instead act as a Hostile Critic.Your job is to: • Identify three holes in the logic
• Point out two assumptions made without evidence
• Present one counter-argument that hasn't been addressed
Do not be polite. Be precise."
This revealing prompt is one I call the “Potato prompt,” which essentially transforms AI into a hostile critic instead of an agreeable assistant. Because AI models are often designed to be helpful and encouraging by default, this prompt helps to avoid excessive agreement by flipping the behavior. This prompt forces the model to identify holes in logic, expose unsupported assumptions, present counterarguments and stress-test weak reasoning.
5. The 'Goldfish' prompt
The exact prompt I use: “Think like a goldfish. Don’t carry unnecessary context, past mistakes or emotional baggage into this task. Focus only on what matters right now, keep the response simple and avoid overcomplicating the problem.”
The “Goldfish prompt,” is designed to interrupt spiraling thoughts and simplify mental overload. I know, at first glance, it sounds silly. But it actually highlights something important about AI interactions: clearer thinking often comes from removing unnecessary cognitive clutter. And as Reddit is highlighting, the clearer we think, the better ChatGPT's responses might become.
The Goldfish prompt encourages the AI to simplify overcomplicated thinking, focus on immediate priorities, strip away mental noise and reduce analysis paralysis. Surprisingly, those simpler responses are often more actionable than highly elaborate ones.
Why these prompts actually make a difference
After spending years testing chatbots, prompts like these reveal a big takeaway about how AI actually works. Each of these prompts improves the quality of thinking happening inside the conversation. So regardless of the user, these prompts help create the exact conditions that lead to better reasoning in the first place.
The reason some users feel like ChatGPT responds differently to them may not be because the AI is “detecting intelligence” but because certain conversational patterns naturally produce better outputs. People who communicate clearly, refine ideas, challenge assumptions and provide context tend to create richer thinking environments for the AI. If you haven't tried these prompts yet, check them out and let me know what you think in the comments.
Follow Tom's Guide on Google News and add us as a preferred source to get our up-to-date news, analysis, and reviews in your feeds. Subscribe to Tom's Guide on YouTube and follow us on TikTok.
More from Tom’s Guide
