You're Not Bad at Writing Prompts, You Just Don't Know How to Talk!
Discover how simple prompts can unlock AI writing success. Learn to communicate clearly with AI using easy models like 3W1H and Purpose-Task-Result.
A few days ago, I gave several in-person talks about AI writing.
During the talks, I demonstrated some of the prompts I frequently use.
The common reaction from the audience was, "Why are the prompts so simple?"
I often get asked by participants, "Teacher, do you have some super powerful prompts that you're secretly hiding?"
Most people think AI writing, or using AI in general, is something very sophisticated, and believe that without some tricks, it's impossible to use.
However, based on my own perspective and over a year of practice, I can confidently say that’s not the case.
The purpose of AI is to allow ordinary people like us to interact with it through natural language.
So, natural language—what we normally speak or type—is the best way to communicate with AI.
The key is, that you just need to express yourself clearly.
Yes, when using AI, especially for writing, there is no profound or lofty threshold. The biggest challenge is whether you can clearly articulate your needs.
I often use particularly simple prompts in my writing, sometimes even just a single sentence.
But as long as that sentence is clear and conveys the task you want AI to execute, then that prompt is effective.
So here's the problem: many people can't express themselves clearly. Some people don’t even realize this until they start using AI and notice their usual way of communicating isn’t working.
I've encountered this with a student in my writing class.
She told me that she thought her communication skills were decent, but when interacting with AI, she found herself often unsure of how to express herself.
So, how do we express ourselves clearly?
We can actually rely on some classic expression models.
For instance, I always advocate the 3W1H model.
The 3Ws are: Who, What, Why.
"Who" refers to the role you assign to the AI;
"What" is telling it what task you want it to perform;
"Why" is explaining the purpose and the background behind the task.
Finally, "How" is instructing AI on how to carry out the task.
For example, if I want to write an article introducing a newly launched coffee machine and hope to generate interest in the product among readers, I would use the 3W model to describe the request:
Who: “An experienced barista.”
What: Write an article introducing the features and benefits of this newly launched smart coffee machine.
Why: State the purpose—help readers understand the machine’s features and inspire them to buy it.
How: “Describe it in a lighthearted and fun tone, making readers feel that it’s easy to use and very convenient.”
This way, the AI would approach the task from the perspective of a barista, conveying the machine’s advantages in a professional yet approachable manner, making complex functions easy to understand, and ultimately achieving my goal of engaging readers.
This is one of the prompts I use most often in my everyday writing with AI, and it works really well.
Besides 3W1H, I also frequently use a very simple "Purpose-Task-Result" model.
This model is not only clear but also very direct, making it particularly suitable for conveying needs accurately in a short amount of time.
For example, when I want AI to write an inspirational blog post, I would describe it like this:
Purpose: Motivate readers through the article;
Task: Write a story about bouncing back from setbacks;
Result: Make readers feel hopeful and empowered.
This prompt is very concise and straightforward, so AI quickly understands what I want without needing to guess my intent.
So, back to the original question, does AI writing really require any advanced skills?
In my opinion, the most important skill is simply becoming better at expressing yourself and communicating your ideas clearly.
AI is just a magnifying glass—it amplifies our expression and makes it easier to turn our thoughts into words.
Perhaps, this is the greatest charm of AI writing: in learning how to communicate with machines, we also learn to have better conversations with ourselves.