(Cover Photo Source: The Council of Europe)
This article was co-authored by Nas John Lumapas
Introduction
When two different intelligences collide, communication between them won’t be smooth. Humans communicate linguistically, while AI, like all computers, communicates through binary and code.
To understand how we bridge this gap, we must look at the history of Human–AI Interaction (HAX).
According to the Interaction Design Foundation (2025):
- 1950s–1960s – AI systems were built using symbolic logic and rule-based reasoning.
- 1970s–1980s – Neural networks, fuzzy logic, genetic algorithms, and expert systems emerged.
- 1990s–2000s – Machine learning and deep learning led to conversational agents.
- Early 2020s – The “AI Boom” brought Large Language Models (LLMs) into everyday life.
We moved from rigid, command-line computing to graphical interfaces and now to conversational AI systems like ChatGPT and Google Gemini.
We are living in what many call “The Age of AI.”
After reading this article, you will:
- Learn about what prompt engineering is
- Know how AI interprets prompts
- Find out the common mistakes when making prompts
- Learn effective prompt engineering with core prompting techniques
- Know how to use AI responsibly
- Be a part of the future of human-AI interaction
What Is Prompt Engineering?
(Photo Source: Fortune)
According to Amazon Web Services (AWS), prompt engineering is:
The process of guiding generative AI to generate desired outputs.
Google Cloud defines it as:
The art and science of designing and organizing prompts for AI.
When the printing press was invented in the 15th century, literacy increased dramatically. Similarly, generative AI is the innovative tool of our generation. Learning prompt engineering increases our digital literacy in the AI era.
In simple terms, prompt engineering is the key to unlocking the treasure chest that is AI and finding the treasures that we want.
How AI Interprets Prompts
(Photo Source: Forbes)
When you enter a prompt:
- The AI processes it using Natural Language Processing (NLP).
- It breaks your input into smaller units called tokens (tokenization).
- It predicts the most likely next word based on learned patterns.
- It generates a response.
Machine learning (ML), as described by IBM (2025), trains algorithms to recognize patterns and make predictions.
At its core:
AI does not “understand” like humans, it predicts based on patterns.
The quality of output depends heavily on:
- Clarity
- Specificity
- Context
- Constraints
Poor prompts can lead to:
- Generic answers
- Hallucinations
- Logical errors
Prompt engineering is essentially optimizing inputs to improve outputs.
Common Mistakes in Prompt Engineering
1. Vague and General Questions
Ambiguous prompts force AI to guess, which may result in hallucinations.
AI hallucination = confidently generating false or misleading information.
Example:
Prompt:
Tell me about learning coding.
Output:
In 2026, learning to code has shifted from "memorizing syntax" to "**systems thinking**." With AI now able to generate perfect boilerplate in seconds, the value of a developer lies in their ability to debug, architect, and understand the "why" behind the code.
Here is a guide to getting started in the current landscape.
---
## 1. Choose Your "Launchpad" Language
Don't stress over picking the "perfect" language—most principles translate across all of them. Pick based on what you want to build:
| Goal | Recommended Language | Why? |
| --- | --- | --- |
| **AI & Data Science** | **Python** | The industry standard for machine learning and automation. |
| **Web Development** | **JavaScript / TypeScript** | Essential for building anything that runs in a browser. |
| **Mobile Apps** | **Swift (iOS) or Kotlin (Android)** | Modern, safe, and officially backed by Apple/Google. |
| **Enterprise/Systems** | **Java or Go** | Used for massive, scalable backend systems and cloud infrastructure. |
---
## 2. The 2026 Learning Strategy: "Problem-First"
The old way was watching 40-hour video courses. The new way is **Contextual Learning**:
* **Build Before You're Ready:** Start a tiny project (like a "To-Do" list or a tip calculator) on day one.
* **The "Break It" Rule:** When AI gives you a working snippet, intentionally change a line to see how it breaks. This teaches you how the code actually functions.
* **Learn "On Demand":** Instead of learning every type of loop, only learn the one you need to solve the specific problem in front of you.
---
## 3. Use AI as a Mentor, Not a Crutch
Tools like **Cursor**, **GitHub Copilot**, and **Gemini Code Assist** are your best friends—if used correctly.
* **Avoid:** "Write a login page for me" Copy/Paste Done. (You learn nothing).
* **Adopt:** "Explain the logic of this authentication function line-by-line."
* **Verification is the New Coding:** Your job is no longer just writing code; it's *verifying* that the AI's code is secure, efficient, and fits your specific needs.
---
## 4. Helpful Tools & Resources
* **Gemini Guided Learning:** You can use my **Guided Learning** mode to break down complex topics step-by-step. Just ask me to "Help me learn [topic]" and click the **Learn** chip.
* **Interactive Environments:** Use **Replit** or **Google Cloud Shell Editor** to write code in your browser without any complex setup.
* **AI-First Editors:** **Cursor** is a popular fork of VS Code that has AI deeply integrated into the workflow, making it easier to see how changes affect your whole project.
> **Pro Tip:** If you can't explain what a block of code does in plain English, you don't "own" that code yet. Delete it and ask the AI to explain the concept again until it clicks.
**Would you like me to create a 30-day personalized learning roadmap for a specific language like Python or JavaScript?**
Proper Prompt:
Prompt:
I am a complete beginner with zero coding experience. I want to learn Python so I can eventually automate simple tasks at my office job.
Please give me a 4-week learning roadmap. For each week, tell me:
What concept to focus on
One specific practice exercise
Which free resource to use (website or YouTube channel)
Assume I can study 1 hour per day.
Output:
Learning to code is a superpower for the modern office, and Python is the perfect "Swiss Army Knife" for it. Since you have 1 hour a day, we’ll focus on the "80/20 rule": learning the 20% of Python that will do 80% of your office work.
Here is your 4-week roadmap to becoming an office automation hero.
---
## **Week 1: The Building Blocks**
* **Focus:** **Syntax, Variables, and Data Types.** You need to learn how to store information (like names, prices, or dates) and how to talk to the computer.
* **Practice Exercise:** Create a "Salary Calculator." Ask the user for their hourly rate and hours worked per week, then have Python print out their total monthly pay after a 10% tax deduction.
* **Resource:** **[Programming with Mosh (YouTube)](https://www.youtube.com/user/programmingwithmosh)** – Watch his "Python Tutorial for Beginners" (first 60 minutes).
---
## **Week 2: The "Brain" (Logic & Loops)**
* **Focus:** **If-statements and Loops.** This is where automation happens. "If" statements let Python make decisions (e.g., *if* the price is over $100, flag it), and "Loops" let Python do repetitive work (e.g., check *every* row in a list).
* **Practice Exercise:** Build a "Password Validator." Write a script that asks for a password. If the password is "Secret123", print "Access Granted." If not, use a loop to let them try again up to 3 times.
* **Resource:** **[FreeCodeCamp (YouTube)](https://www.youtube.com/c/Freecodecamp)** – Search for their "Scientific Computing with Python" playlist or their 4-hour Python beginner course.
---
## **Week 3: Handling Files & Text**
* **Focus:** **Reading and Writing Files.** In an office, your data lives in files. You’ll learn how to open a `.txt` or `.csv` file, change the text inside, and save a new version.
* **Practice Exercise:** The "Automatic File Organizer." Create a script that looks at a list of filenames (like `report_2025.txt`, `image1.jpg`, `data.csv`) and "sorts" them by printing out which folder they should go into based on their extension.
* **Resource:** **[Automate the Boring Stuff with Python](https://automatetheboringstuff.com/)** – Read **Chapter 9: Reading and Writing Files** (the website is free to read).
---
## **Week 4: The Office Superpower (Excel & CSV)**
* **Focus:** **The Pandas Library.** Most office work is just moving data around in Excel. Python has a tool called "Pandas" that can read 10,000 rows of Excel data in a fraction of a second.
* **Practice Exercise:** Download a sample CSV of sales data. Use Python to find the total sum of the "Sales" column and save only the rows where the "Status" is "Pending" into a new file.
* **Resource:** **[Corey Schafer (YouTube)](https://www.youtube.com/user/schafer5)** – Look for his "Pandas Tutorial" series. He is widely considered the best teacher for clear, technical breakdowns.
---
### **How to stay on track:**
1. **Don't memorize:** You don't need to remember every command. Professional coders Google things 50 times a day.
2. **Use an IDE:** Download **VS Code** (free) to write your code. It will highlight your errors and make learning much easier.
3. **Consistency > Intensity:** 1 hour every day is much better than a 7-hour marathon on Sunday.
**Would you like me to generate a simple "starter script" for Week 1 that you can try running right now?**
2. Omitting Crucial Context
AI cannot read your mind. Provide:
- Social-historical context
- Cultural context
- Temporal details
- Physical constraints
- Authorial intent
3. Overloading With Too Many Tasks
Too many instructions can confuse the model and cause it to overlook important parts of your request.
Example:
Prompt:
I'm planning a 10-day trip to Japan in the spring. Please create a detailed day-by-day itinerary for Tokyo, Kyoto, and Osaka. For each day, suggest restaurants for breakfast, lunch, and dinner that are budget-friendly. Also, give me a full packing list for spring weather, teach me 50 basic Japanese phrases, explain the entire history of Japanese tea ceremonies, compare the Shinkansen bullet train to the JR Rail Pass and tell me which is better, list the top 10 cultural etiquette rules I must know, and recommend the best pocket Wi-Fi rental companies. Oh, and also tell me how to use a Japanese toilet.
Output:
Japan in the spring is a dream—cherry blossoms, crisp air, and a perfect blend of high-tech and history. Since you’re hitting the "Golden Route" (Tokyo, Kyoto, Osaka), here is your comprehensive guide to doing it right in 2026.
---
## 📅 10-Day "Golden Route" Itinerary
### **Tokyo (Days 1–4)**
* **Day 1: Arrival & Shinjuku.** Drop bags, head to Shinjuku Gyoen for blossoms, and see the Tokyo Metropolitan Govt Building (Free views).
* *Lunch:* **Ichiran Shinjuku** (Famous tonkotsu ramen).
* *Dinner:* **Omoide Yokocho** (Yakitori in "Piss Alley").
* **Day 2: Harajuku & Shibuya.** Meiji Jingu Shrine, Takeshita Street, and the Shibuya Crossing.
* *Lunch:* **Kyushu Jangara Ramen** (Harajuku).
* *Dinner:* **Genki Sushi** (Conveyor belt fun in Shibuya).
* **Day 3: Asakusa & Akihabara.** Senso-ji Temple and the electric town.
* *Lunch:* **Asakusa Kagari** (Udon/Soba).
* *Dinner:* **Coco Ichibanya** (Iconic Japanese curry).
* **Day 4: TeamLab Borderless & Odaiba.** Futuristic art and bay views.
* *Lunch:* **DiverCity Plaza Food Court** (Huge variety).
* *Dinner:* **Yano-tei** (Budget-friendly Teppanyaki).
### **Kyoto (Days 5–7)**
* **Day 5: Arashiyama.** Bamboo Grove and Monkey Park.
* *Lunch:* **Otsuka** (Famous roast beef bowls).
* *Dinner:* **Musashi Sushi** (Sanjo-dori).
* **Day 6: Higashiyama.** Kiyomizu-dera, Sannenzaka/Ninenzaka streets, and Gion.
* *Lunch:* **Nishiki Market** (Street food samples).
* *Dinner:* **Chao Chao Gyoza** (Pontocho area).
* **Day 7: The Classics.** Fushimi Inari (gates) and Kinkaku-ji (Golden Pavilion).
* *Lunch:* **Kura Sushi** (Near Fushimi).
* *Dinner:* **Kyoto Ramen Koji** (Inside Kyoto Station).
### **Osaka (Days 8–10)**
* **Day 8: Osaka Castle & Dotonbori.** History by day, neon lights by night.
* *Lunch:* **Tokumasa** (Curry udon near the castle).
* *Dinner:* **Mizuno** (Okonomiyaki in Dotonbori).
* **Day 9: Universal Studios Japan (USJ).** Super Nintendo World is a must.
* *Lunch/Dinner:* Eat within the park or at **CityWalk** (Lawson convenience store for budget).
* **Day 10: Nara Day Trip.** Bowing deer and the Great Buddha.
* *Lunch:* **Nakatanidou** (Fresh mochi and snacks).
* *Dinner:* **Shinsekai Kushikatsu** (Deep-fried skewers) before heading home.
---
## 🚄 Shinkansen vs. JR Pass (2026 Update)
**Verdict: Buy individual tickets.**
Following the massive price hikes in 2023/24, a 7-day JR Pass now costs **¥50,000**. A round-trip ticket from Tokyo to Osaka costs roughly **¥29,000**. Unless you are going all the way to Hiroshima or Hokkaido and back within 7 days, the JR Pass is no longer worth it.
> **Tip:** Buy your Shinkansen tickets via the **SmartEX** app or at the station machines.
---
## 🎒 Spring Packing List
Spring (March–May) is beautiful but fickle. It can be 18°C during the day and 7°C at night.
* **Clothing:**
* Light puffer jacket (Uniqlo Heattech is your friend).
* Layers: T-shirts + light sweaters/cardigans.
* **Walking shoes** (You will walk 15k+ steps daily).
* One "nice" outfit for temples or upscale dining.
* **Essentials:**
* **Portable Power Bank** (Phone batteries die fast with Google Maps).
* Compact umbrella (or buy a "konbini" one for ¥500 if it rains).
* Coin purse (Japan is still very cash-heavy).
---
## 🗣️ 10 Essential Phrases (50 is a lot! Start here):
1. **Arigatou gozaimasu** (Ah-ree-gah-toe go-zai-mass) – Thank you.
2. **Sumimasen** (Su-me-mah-sen) – Excuse me/Sorry.
3. **Kore o kudasai** (Ko-re oh ku-dah-sai) – I’ll take this, please.
4. **Okaikei onegaishimasu** (Oh-kai-kay oh-neh-guy-she-mass) – Check, please.
5. **...wa doko desu ka?** (...wa doh-ko dess-ka?) – Where is ...?
6. **Eigo?** (Ay-go?) – English?
7. **Oishii!** (Oy-shee!) – Delicious!
8. **Wakarimasen** (Wah-kah-ree-mah-sen) – I don't understand.
9. **Hai / Iie** (Hi / Ee-eh) – Yes / No.
10. **Irasshaimase!** (Ee-ra-shai-ma-seh) – Welcome! (You’ll hear this in every shop).
---
## 🍵 History of the Tea Ceremony (*Chanoyu*)
* **9th Century:** Tea arrives from China via monks.
* **12th Century:** Zen monk Eisai introduces powdered green tea (matcha).
* **15th Century:** It evolves from a luxury for elites into a spiritual practice.
* **16th Century:** **Sen no Rikyu** perfects the "Wabi-cha" style, emphasizing simplicity, humility, and the aesthetic of imperfection (*Wabi-sabi*).
---
## ⛩️ Top 10 Etiquette Rules
1. **No Tipping:** It can actually be seen as confusing or rude.
2. **Escalators:** Stand on the **Left** in Tokyo; **Right** in Osaka.
3. **Silence on Trains:** No phone calls; keep talking to a whisper.
4. **Trash:** There are no public bins. Carry a small bag for your trash.
5. **Shoes Off:** Remove shoes at temple entrances and indoor restaurants with tatami.
6. **Chopsticks:** Never stick them vertically in rice (funeral rite).
7. **Bowing:** A small nod is enough for most interactions.
8. **Tattoos:** Check "tattoo-friendly" lists if you plan to visit an Onsen (hot spring).
9. **Walking & Eating:** Generally frowned upon; step to the side of the street to eat.
10. **Cash Trays:** Place money in the small tray at the counter, not directly in hands.
---
## 📶 Best Pocket Wi-Fi & 🚽 The Toilet Guide
* **Wi-Fi:** **Japan Wireless** or **Ninja WiFi** are the industry leaders. Pick up at the airport.
* **The Toilet:** 1. **The Panel:** Usually on the wall or the side of the seat.
2. **The Icons:** Blue spray = Rear; Pink spray = Front/Feminine.
3. **Stop:** The big orange/red square button.
4. **Flush:** Often a separate lever on the tank or a sensor on the wall labeled **大** (Big) or **小** (Small).
**Would you like me to find the specific cherry blossom forecast dates for your travel window?**
Proper Prompt:
Prompt:
I am a complete beginner with zero coding experience. I want to learn Python so I can eventually automate simple tasks at my office job.
Please give me a 4-week learning roadmap. For each week, tell me:
What concept to focus on
One specific practice exercise
Which free resource to use (website or YouTube channel)
Assume I can study 1 hour per day.
Output:
Learning to code is a superpower for the modern office, and Python is the perfect "Swiss Army Knife" for it. Since you have 1 hour a day, we’ll focus on the "80/20 rule": learning the 20% of Python that will do 80% of your office work.
Here is your 4-week roadmap to becoming an office automation hero.
---
## **Week 1: The Building Blocks**
* **Focus:** **Syntax, Variables, and Data Types.** You need to learn how to store information (like names, prices, or dates) and how to talk to the computer.
* **Practice Exercise:** Create a "Salary Calculator." Ask the user for their hourly rate and hours worked per week, then have Python print out their total monthly pay after a 10% tax deduction.
* **Resource:** **[Programming with Mosh (YouTube)](https://www.youtube.com/user/programmingwithmosh)** – Watch his "Python Tutorial for Beginners" (first 60 minutes).
---
## **Week 2: The "Brain" (Logic & Loops)**
* **Focus:** **If-statements and Loops.** This is where automation happens. "If" statements let Python make decisions (e.g., *if* the price is over $100, flag it), and "Loops" let Python do repetitive work (e.g., check *every* row in a list).
* **Practice Exercise:** Build a "Password Validator." Write a script that asks for a password. If the password is "Secret123", print "Access Granted." If not, use a loop to let them try again up to 3 times.
* **Resource:** **[FreeCodeCamp (YouTube)](https://www.youtube.com/c/Freecodecamp)** – Search for their "Scientific Computing with Python" playlist or their 4-hour Python beginner course.
---
## **Week 3: Handling Files & Text**
* **Focus:** **Reading and Writing Files.** In an office, your data lives in files. You’ll learn how to open a `.txt` or `.csv` file, change the text inside, and save a new version.
* **Practice Exercise:** The "Automatic File Organizer." Create a script that looks at a list of filenames (like `report_2025.txt`, `image1.jpg`, `data.csv`) and "sorts" them by printing out which folder they should go into based on their extension.
* **Resource:** **[Automate the Boring Stuff with Python](https://automatetheboringstuff.com/)** – Read **Chapter 9: Reading and Writing Files** (the website is free to read).
---
## **Week 4: The Office Superpower (Excel & CSV)**
* **Focus:** **The Pandas Library.** Most office work is just moving data around in Excel. Python has a tool called "Pandas" that can read 10,000 rows of Excel data in a fraction of a second.
* **Practice Exercise:** Download a sample CSV of sales data. Use Python to find the total sum of the "Sales" column and save only the rows where the "Status" is "Pending" into a new file.
* **Resource:** **[Corey Schafer (YouTube)](https://www.youtube.com/user/schafer5)** – Look for his "Pandas Tutorial" series. He is widely considered the best teacher for clear, technical breakdowns.
---
### **How to stay on track:**
1. **Don't memorize:** You don't need to remember every command. Professional coders Google things 50 times a day.
2. **Use an IDE:** Download **VS Code** (free) to write your code. It will highlight your errors and make learning much easier.
3. **Consistency > Intensity:** 1 hour every day is much better than a 7-hour marathon on Sunday.
**Would you like me to generate a simple "starter script" for Week 1 that you can try running right now?**
4. Not Iterating
Prompt engineering is iterative.
Refine → Test → Improve.
5. Not Including Constraints
Tell AI both:
- What to do
- What not to do
6. Forgetting to Specify Output Format
LLMs default to unstructured paragraphs.
Instead, request:
- Bullet points
- Tables
- JSON
- Step-by-step
- Code blocks
Structured outputs improve clarity and usability.
7. Not Using Prompting Techniques
Without techniques, you’re underutilizing AI.
Step-by-Step Guide to Effective Prompt Engineering
(Photo Source: CertiProf)
Not all prompting techniques are equal. Each produces different behavioral responses.
Using the wrong technique may cause:
- Hallucinations
- Token waste
- Logical failure
Before prompting, choose an LLM such as Google Gemini, ChatGPT, or another preferred model. Prompting techniques generally work across platforms.
Core Prompting Techniques
1. Role Prompting
Assign a role to shape tone, expertise, and reasoning.
Template
Role: [Assign a role]
Task: [Define the task]
Context: [Provide necessary data]
Constraint: [Define limitations]
Example (Software Development)
Role: Act as a Senior Frontend Developer who mentors junior developers.
Task: Review my CSS code below. I am trying to center a "Profile Card"
perfectly in the middle of the web page.
Context:
.background-container {
display: flex;
justify-content: center;
}
Constraint: You are strictly forbidden from giving me the missing CSS
properties. Instead, explain the difference between the "main axis" and the
"cross axis" in Flexbox. Then ask me one guiding question.
Role prompting shapes how the AI behaves — mentor, critic, tutor, consultant, etc.
2. Zero-Shot Prompting
No examples are provided. The AI relies purely on its training.
Template
Instruction: Define the rule
Context: Provide necessary data
Task: Generate response
Best for:
- Translation
- Summarization
- Sentiment classification
- Simple transformations
3. Few-Shot Prompting
Provide examples to establish a pattern.
Template
Context 1: Example input
Output 1: Example output
Context 2: Example input
Output 2: Example output
Context n: New input
Output n:
This improves consistency and formatting accuracy.
4. Chain-of-Thought Prompting
Encourages step-by-step reasoning.
Ideal for:
- Mathematics
- Logic
- Complex problem-solving
Example
Find the integral:
∫ x cos(x) dx
Reasoning:
- Identify as product of algebraic and trigonometric functions.
- Choose Integration by Parts.
- Let u = x and dv = cos(x)dx.
- Apply formula uv − ∫vdu.
Final Answer:
x sin(x) + cos(x) + C
Chain-of-thought prompting improves transparency and logical clarity.
Real-World Applications of Prompt Engineering
Software Development
(Photo Source: n8n Blog)
AI assists with:
- Code generation
- Debugging
- Explaining logic
- Refactoring
However, the quality of code depends heavily on prompt structure.
Role prompting works well for:
- Code reviews
- Mentorship simulation
- Enforcing best practices
Education and Learning
(Photo Source: Discovery Education)
AI can:
- Act as a tutor
- Generate quizzes
- Provide step-by-step math solutions
- Explain topics at varying levels of difficulty
Chain-of-thought prompting is particularly powerful in education.
But remember:
Output quality depends on input quality.
Ethical Considerations and Responsible Use
(Photo Source: TechHQ)
AI is powerful, but imperfect.
Risks include:
- Bias
- Hallucinations
- Data privacy concerns
- Academic dishonesty
- Overdependence
Avoid sharing:
- Personal identifiers
- Confidential data
- Sensitive credentials
Responsible AI use requires awareness of institutional and company policies, as well as international ethical frameworks such as UNESCO’s principles on AI ethics.
These include:
- Do no harm
- Privacy protection
- Transparency
- Accountability
- Human oversight
- Fairness and non-discrimination
After reading this article, remember not only the techniques, but also the ethics of using AI. In our community, we stand by responsible and ethical prompt engineering.
The Future of Human–AI Interaction with Prompt Engineering
AI is evolving from a niche tool into a foundational skill.
Prompt engineering is expanding into context engineering — designing not just prompts, but entire AI interaction environments.
Modern LLMs now:
- Ask clarifying questions
- Suggest refinements
- Automate improvements
- Offer structured response options
AI is evolving alongside us.
Even if AI advances rapidly, responsible prompt engineering ensures:
- Ethical outputs
- Human-centered systems
- Rights-based AI development
To Conclude
Artificial Intelligence is not truly “artificial.”
It is:
- Inspired by humans
- Built by humans
- Used by humans
- Affecting humans
Prompt engineering is no longer a niche skill.
It is becoming a core literacy of the AI age.
When two intelligences collide, prompt engineering becomes the bridge.
References
Amazon Web Services. (2026). What is Prompt Engineering?
Bergmann, D. (2025). What is Machine Learning? IBM.
Google Cloud. (2026). Prompt Engineering for AI Guide.
Harvard University. (2023). Getting Started with Prompts.
UNESCO. (2026). Ethics of Artificial Intelligence.
OpenAI. (2025). Usage Policies.
Interaction Design Foundation. (2025). What is Human-AI Interaction?
Top comments (0)