How to Get Better AI Responses
4 Easy Prompting Strategies That Work
Ever asked a large language model (LLM) for help and gotten a response that’s technically correct but... not actually helpful? Maybe it’s too vague, too generic, or just misses the point entirely.
That’s not just an AI problem—it’s a communication problem.
Ever had someone hear what you’re saying and still completely misunderstand you? AI does that too. There are so many LLM providers to use, each with different models and different model types. I’ll address these options in a future article, along with some helpful prompt adjustments for the newer reasoning models. For now, though, we’re going to focus on effective prompting with the traditional models, which is what you’ll see most if you’re using the free or first tier paid option of the most popular providers.
Regardless of the model you’re using, it’s generally the case that you’ll get better responses when your prompt is clear and structured. That’s where prompt patterns come in. These are simple ways to phrase your requests so AI can give you answers that actually fit your needs. No coding required—just a few tweaks to how you ask.
After a lot of reading, experimenting, and a few AI-generated disasters, I’ve learned four simple patterns that consistently make AI more useful in everyday life. Let’s break them down.
The Template Pattern: Tell AI Exactly How You Want the Response
Ever gotten a weirdly unhelpful AI response?
You ask for a summary, and it gives you a long-winded essay. You want a simple list, and it spits out a paragraph with no clear structure. That’s because AI doesn’t automatically know how you want your information—it just takes its best guess. And sometimes, that guess is wrong.
The fix? The Template Pattern. Instead of letting AI guess the format, you tell it exactly what you need.
Why It Works
Think of it like ordering at a restaurant. If you just say, “Bring me some food,” who knows what you’ll get? But if you say, “I’d like a burger with no pickles, extra cheese, and a side of fries,” you’re much more likely to get exactly what you want. AI works the same way.
By giving it a format to follow, you:
Keep things consistent – Handy for tasks you repeat often (like summarizing meetings).
Make sure nothing gets left out – AI fills in the blanks instead of making stuff up.
Get easy-to-read results – No more digging through a wall of text to find what you need.
How to Use It
Instead of: "Summarize this meeting."
This might get you: "The team discussed the upcoming product launch, marketing strategies, and budget concerns. Everyone agreed to finalize the promotional materials by Friday."
Try This: "Summarize this meeting using this format:
Key decisions:
Action items:
Follow-ups needed:
Next steps:"
This gives you a structured response like:
Key decisions: Product launch date set for March 15; Green marketing campaign approved
Action items: Sarah to finalize promotional materials; Tom to contact media partners
Follow-ups needed: Budget approval from finance department
Next steps: Reconvene next Tuesday to review progress
See the difference? The first one leaves AI guessing. The second one tells it exactly what to do—and makes the response much easier to use.
Real-Life Uses
Emails – “Write a polite but firm complaint email using this structure: [Greeting], [Issue], [Desired Resolution], [Closing].”
Reports – “Summarize my workweek like this: [Tasks Completed], [Challenges], [Next Steps].”
Lists & Plans – “Plan my trip using this format: [Day], [Activities], [Notes].”
The Chain-of-Thought Pattern: Helping AI Think Through Problems
Ever get an AI answer that feels... shallow?
Now that you know how to structure your AI's responses with the Template Pattern, let's tackle another common problem: shallow answers that lack depth. Ever get an AI answer that sounds smart but doesn’t actually help? That’s because, as mentioned above, many AI models don’t include deeper reasoning. And for those that do, you may or may not have access to see their reasoning, The fix? The Chain-of-Thought Pattern, where you make the LLM show its work.
Why It Works
Many non-reasoning AI models are good at predicting words, but not always at connecting ideas. They can give quick answers but often miss the deeper thinking. When you make AI show its work step by step, you get:
More thoughtful answers – Instead of jumping to conclusions, AI works through the logic.
Better decision-making – AI considers pros and cons instead of giving a one-size-fits-all answer.
Clearer explanations – Useful when you’re learning something new and need details, not just a summary.
How to Use It
Rather than just asking AI for a simple answer, guide it through the reasoning process.
Instead of: "Is it better to buy or lease a car?"
This might get you: "Buying a car means you own it, while leasing is like renting. Buying is better for long-term use, leasing is better for people who want a new car every few years."
Try This: "Compare the pros and cons of buying vs. leasing a car. First, list the financial factors, then the convenience factors, and finally, the long-term impact. Conclude with a recommendation based on different lifestyles."
See the difference? The first prompt gets you a generic, high-level answer. The second one forces AI to think like a person weighing their options.
Real-Life Uses
Big financial decisions – “Explain whether I should refinance my mortgage. First, analyze interest rates, then break down monthly payment changes, and finally, assess the long-term savings.”
Planning projects – “Help me organize my book outline. First, list the main topics, then suggest logical subtopics, and finally, propose a chapter order.”
Learning new skills – “Explain how to improve my public speaking. Start with structuring a speech, then cover voice and tone, and finally, discuss body language and stage presence.”
The Fact-Check List Pattern: Helping You Catch AI’s Confident Mistakes
AI sounds so sure of itself… even when it’s wrong.
While the Chain-of-Thought Pattern helps AI reason better, we still need to be careful about factual accuracy. After all, even thoughtful reasoning can be based on incorrect information. Ever had an LLM tell you something that felt right, only to find out later it was completely off? That’s because AI doesn’t actually “know” things—it's like someone who's read a ton of books but never experienced life firsthand. It can talk about facts, but doesn't truly understand them. And as you know, it confidently throws out incorrect information like it’s an undisputed fact.
Check out this article to learn more about why this happens: https://www.theayiguy.com/p/why-ai-trust-safety-and-privacy-matter
That’s where the Fact-Check List Pattern comes in. Instead of blindly trusting AI’s answers, you make it show its work so you can verify the details.
Why It Works
AI isn’t perfect. It can mix up facts, pull outdated info, or even straight-up invent things - AI can make stuff up out of nowhere (what many call 'hallucinations')!. By asking AI to list the key facts behind its answer, you:
Get transparency – AI has to spell out what it’s basing its response on.
Catch errors – If something looks off, you know exactly what to double-check.
Make AI self-correct – Sometimes, AI will spot its own mistakes just by going through this extra step.
How to Use It
Instead of taking AI’s response at face value, ask it to back up its claims.
Instead of: "Summarize the history of electric cars."
This might get you: "Electric cars date back to the 1800s and were popular until gasoline cars took over. They've made a comeback in recent years with Tesla leading the way."
Try This: "Summarize the history of electric cars. At the end, list the key facts you used and their sources. Highlight any areas where information might be uncertain."
This gives you a more verifiable response like: "Electric cars date back to the 1800s, with the first practical electric vehicle built by Thomas Parker in 1884... [summary continues]
Key facts:
First practical electric vehicle: 1884 by Thomas Parker
Peak of early electric cars: Early 1900s (28% of cars produced in the US were electric in 1900)
Decline period: 1920-1970s due to cheap gasoline and mass production of combustion engines
Modern revival: Began with GM EV1 (1996), accelerated with Tesla Roadster (2008)
Areas of uncertainty: The exact sales figures from the early 1900s vary by source and may not be fully accurate."
The first prompt might lead AI to generate an interesting but potentially inaccurate summary. The second prompt forces AI to be more careful and transparent, helping you verify its claims.
Real-Life Uses
Fact-checking AI’s recommendations – “Recommend the best laptop for under $1,000. List the specific models, their key features, and the sources of this information.”
Checking historical claims – “Summarize the key events of the Cold War. Include a list of major dates and where this information comes from.”
Making AI review its own response – “Give me five reasons why intermittent fasting is beneficial. Afterward, list any claims that might be controversial or need more research.”
The Flipped Interaction Pattern: Let AI Ask You the Right Questions
Ever get a generic AI response that totally misses the point?
You ask for help with productivity, and it throws out the same tired advice: "Make a to-do list!" "Use the Pomodoro technique!" That’s because AI is working with limited info—it doesn’t know your situation unless you tell it.
But what if, instead of you struggling to phrase the perfect question, AI helped you figure out what you actually need? That’s where the Flipped Interaction Pattern comes in. Instead of making a request and hoping for the best, you have AI ask you clarifying questions first.
Why It Works
Sometimes, we don’t even know the best way to ask a question. AI can help by guiding the conversation and narrowing down exactly what we need. This approach:
Makes vague requests more useful – AI won’t just guess; it will get the right details first.
Personalizes responses – Instead of a generic answer, AI tailors its advice based on you.
Improves decision-making – AI helps you think through factors you might not have considered.
How to Use It
Instead of just asking AI for an answer, make it ask you questions first.
Instead of: "Help me improve my productivity."
This might get you: "To improve productivity, try using the Pomodoro technique, create a to-do list, minimize distractions, take regular breaks, and organize your workspace."
Try This: "Before giving me advice on productivity, ask me up to five clarifying questions about my work habits, biggest challenges, and goals. Then, use my answers to give tailored suggestions."
This starts a more helpful conversation like: "I'd be happy to help you improve your productivity. First, let me ask you a few questions:
What type of work do you typically do?
When during the day do you feel most focused?
What are your biggest distractions?
Have you tried any productivity methods before?
What's your main goal for improving productivity?"
The first prompt may get you a one-size-fits-all response. The second one turns AI into a guided coach, adapting its advice to your specific situation.
Real-Life Uses
Career planning – “Help me figure out my next career move. First, ask me questions about my skills, interests, and job preferences, then suggest options.”
Personalized AI learning – “Before teaching me about AI, ask me about my current knowledge level and what I want to use it for. Then tailor the explanation accordingly.”
Travel recommendations – “Before suggesting a travel itinerary, ask me about my budget, interests, and preferred pace of travel. Then provide recommendations.”
Bonus Tip: Combine Patterns for Even Better AI Responses
Want to take things a step further? Try mixing and matching patterns to get even more useful responses. One of the most powerful combos is Flipped Interaction + Persona Pattern.
The Persona Pattern tells AI to take on a specific role (like a financial advisor, travel planner, or writing coach). When paired with Flipped Interaction, AI doesn’t just ask random questions—it asks smarter, role-specific questions that lead to better, more tailored answers.
Click here to learn more about the persona pattern:
How to Use It
Instead of just telling AI to ask clarifying questions, have it take on a role first.
Instead of: "Before suggesting a workout plan, ask me questions about my fitness level and goals."
Try This: "Act as a certified personal trainer. Before suggesting a workout plan, ask me up to five questions about my fitness level, goals, and exercise history. Then use my answers to create a structured plan."
By combining these two patterns, you focus AI’s responses even further, ensuring it asks the right questions for the right reasons.
Conclusion: Try These Patterns and See the Difference
AI isn’t just about getting answers—it’s about having better conversations. A few simple tweaks to the way you phrase your prompts can turn AI from a generic chatbot into a genuinely useful assistant that actually fits your needs.
We’ve covered four powerful prompt patterns:
Template Pattern – Get structured responses exactly how you need them.
Chain-of-Thought Pattern – Help AI think through problems step by step.
Fact-Check List Pattern – Keep AI accountable by making it list its facts.
Flipped Interaction Pattern – Let AI guide the conversation by asking you the right questions.
And if you really want to level up your AI interactions, try combining patterns—like using the Persona Pattern with Flipped Interaction for even more tailored and thoughtful responses.
Now it’s your turn - put these patterns to the test!
Whether you're brainstorming, planning a trip, or just trying to make AI more useful, a few tweaks can make all the difference.
Got a favorite pattern or a cool success story? Share it in the comments—I’d love to hear how you're using them!