Getting Started with AI: The Fundamentals of Prompting

Modern AI chat tools are powered by Large Language Models (LLMs), which are sophisticated systems trained on vast amounts of text data to understand and generate human-like responses.

LLMs like ChatGPT, Claude, and Gemini don't truly "understand" in the human sense, but they excel at recognizing patterns in language and predicting what text should come next based on your input. This is why the quality of your prompt matters so much. The AI model can only work with what you give it, making clear communication the foundation of getting useful results.

When you interact with an AI model, you're engaging in a deliberate exchange where how you ask matters as much as what you ask. This is where learning how to prompt AI effectively becomes essential.

What Is a Prompt?

A prompt is the input you provide to an AI model to elicit a specific response. Think of it as a set of instructions, a question, or a request that guides the model toward the output you want. While it might look like ordinary text, a well-crafted prompt is actually carefully structured communication designed to maximize the quality and relevance of the AI's response.

Why Prompting Differs from Ordinary Conversation

Talking to an AI isn't quite the same as chatting with a human. AI models don't have memory of previous conversations (unless explicitly provided), they can't read your mind, and they don't share your context. A prompt bridges this gap by providing all the necessary information upfront. While you might tell a friend, "write that thing we discussed," an AI needs explicit details about what thing, in what format, for what purpose, and with what constraints.

Understanding how to prompt AI is a skill.

The clearer and more specific your prompt, the better the AI can align its response with your actual needs.

The Anatomy of a Prompt

Effective prompts typically include several key components:

• Context: Background information that helps the AI understand the situation. For example, "I'm a high school teacher preparing a biology lesson."

• Task: A clear statement of what you want. "Create a quiz on cellular respiration."

• Constraints: Specific requirements or limitations. "Include 10 multiple-choice questions, avoid overly technical jargon, and make it suitable for 15-year-olds."

• Format: How you want the output structured. "Present it as a numbered list with answer choices labeled A through D."

• Examples: When relevant, showing what good output looks like can dramatically improve results.

These elements don't have to always be in a specific order. You also don't have to include all of them to get a good prompt.

Good Prompts vs. Bad Prompts

Bad prompts are vague, ambiguous, or missing critical information.

"Tell me about cells"

...leaves the AI guessing about depth, scope, and purpose.

Good prompts are specific, clear, and complete.

"Explain the difference between prokaryotic and eukaryotic cells in three paragraphs, using analogies a 10th grader would understand"

...gives the AI everything it needs to succeed.

The difference often lies in specificity. Instead of:

"make this better"

try:

"revise this paragraph to be more concise while maintaining a professional tone."

What Is Prompt Engineering?

Prompt engineering sounds technical and intimidating, but it's really just the practice of crafting effective prompts to get better results from AI.

The term shouldn't scare you. It's simply about being intentional with your instructions. Anyone can learn prompt engineering basics and immediately see improvement in their AI interactions. It's less about complex formulas and more about clear communication and knowing a few helpful techniques.

Prompt Engineering Techniques to Improve Your Results

Several established techniques can enhance your prompting when you're learning how to prompt AI:

Role/Persona prompting

Assign the AI a specific role or perspective to shape its responses. Instead of

"How should I invest $10,000?"

try:

 "You are an experienced financial advisor speaking to a risk-averse client in their 30s. How should I invest $10,000?"

This contextualizes the advice appropriately. Though this is not financial advice, and you should probably speak to a professional rather than an AI tool.

Chain-of-thought prompting

Advice from a few years ago may have suggested explicitly asking the AI to "think step-by-step" to improve reasoning, but modern models often do this automatically behind the scenes.

This technique is still valuable when you want to understand the reasoning process, not just get an answer. It serves as an excellent learning aid. For example, instead of asking:

"What's 15% of 240?"

...try:

"Calculate 15% of 240 and show your work step-by-step"

...when you want to understand the methodology.

Few-shot prompting

Provide examples of input-output pairs to demonstrate the pattern you want. For instance:

"Convert these customer messages to formal responses: 
	Customer: 'ur product is broken!!!'
	Response: 'Thank you for contacting us. We apologize for the issue with your product.'

	Customer: 'need refund asap'
	Response: 'We understand your request for a refund and will process this promptly.'

	Now convert: 'this thing doesnt work'"

The AI model picks up on the pattern you demonstrate and replicates it with the answer it provides you.

Prompt chaining

Break complex tasks into smaller steps and prompt for one part at a time. Take the prompts sequentially, using output from one as input for the next. For example,

First prompt:

"List five trending topics in sustainable fashion."

Second prompt:

 "Take topic #3 from your previous response and write a 200-word blog introduction about it."

Iterative Refinement

Start with a basic prompt, then refine based on the output you receive. For example, begin with:

"Write a product description for noise-cancelling headphones,"

...then follow up with:

"Make it more concise and emphasize the battery life feature."

---

Mastering these techniques transforms how to prompt AI from guesswork into a systematic approach. The investment in learning to prompt well really pays off in the quality, relevance, and usefulness of AI outputs. Whether you're drafting emails, analyzing data, or brainstorming ideas, better prompts mean better results.

Interested to learn more about prompt engineering? Take a look at this article.