How to write effective prompts to get the most out of ChatGPT


ChatGPT is a natural language processing model developed by OpenAI that can provide natural responses to text input. This technology is used in various fields such as online customer support, chatbots, and text-based communication. However, creating appropriate prompts is necessary to effectively utilize ChatGPT. This article explains how to write prompts effectively to make the most of ChatGPT.

What is ChatGPT?

ChatGPT is a neural network model trained by Deep Learning that can generate natural responses like humans based on a large amount of training data. It has a wide range of applications, including online customer support, chatbots, and natural language processing research.

  • Natural responses using human-like language, not mechanical responses
  • Achieving smooth dialogue with users

How to Write Prompts Effectively

Prompts are used to indicate the direction of the response that ChatGPT generates. Effective prompt writing is also essential to get the most out of ChatGPT. This section describes the purpose and types of prompts and introduces some specific points on how to write them.

Purpose and Types of Prompts

A prompt is an instruction or question that you give to ChatGPT. There are various types of prompts, and you should choose the appropriate prompt for the purpose you are giving to ChatGPT. The main types of prompts are as follows

  • Question-type prompt: A prompt to ask ChatGPT for some information.
    For example, “What is your favorite color?” etc.
  • Explanatory prompts: Prompts to explain something to ChatGPT.
    For example, “What is artificial intelligence? etc.
  • Task-type prompts: Prompts that ask ChatGPT to do something.
    For example, “Please tell me the weather.

There are different types of prompts depending on the purpose.

Specific Writing Points for Prompts

The following are some pointers on how to write specific prompts.

  1. Creative and specific questions:
    It is important to include an interesting question in the prompt. However, questions that are too broad or too simplistic may result in inaccurate ChatGPT-generated responses. Therefore, it is important to include specific and creative questions. For example, “What is your favorite novel?” rather than, “What is the most memorable novel you have read recently? Please tell us why.” By including more specific and in-depth questions, such as, “What is your favorite novel?
  2. Appropriate tone and language for the purpose:
    Prompts should have the appropriate tone and language to match the purpose. For example, if you are seeking feedback, a positive, upbeat tone is appropriate. On the other hand, when providing information, clear and precise language should be used. Also, when posing questions, it is important to use a friendly tone. By keeping in mind the appropriate tone and language for the purpose, responses are more likely to be clear and relevant.
  3. Prompt and Response Consistency:
    Consistency between prompt and response is critical. If the prompt is inadequate or inappropriate, the response generated by ChatGPT will also be inappropriate. Therefore, it is important to consider prompt and response consistency. For example, if the prompt is “What is your favorite color?” and ChatGPT generates a response with content such as “What is my favorite movie? Therefore, it is necessary to consider the consistency between the prompt and the response and to check whether the response to the prompt is appropriate.

Common prompt failures and how to remedy them

Inadequate prompts affect the responses generated by ChatGPT. Below are some common prompt failures and how to remedy them.

Unclear topic

“Let’s talk about the movie.”

This prompt is very broad and does not indicate what topic to talk about; ChatGPT cannot determine a specific topic for such a prompt, and the response may be ambiguous.

“Let’s talk about a movie you recently saw.”

This prompt indicates to ChatGPT to provide a specific topic, talk about a movie you recently saw; ChatGPT is able to generate a response relevant to this topic, which is more likely to be more specific and relevant.

Abstract Question

“What do you think about our company?”

This prompt does not provide a specific question or feedback instructions and does not clearly indicate to ChatGPT what the topic should be. This means that ChatGPT cannot provide a specific answer to the question and may not receive a satisfactory response.

“Tell us what you think about our services.”

This prompt prompt prompts ChatGPT to suggest specific improvements to the service by giving specific instructions regarding the response; ChatGPT can then provide specific feedback on how to improve the service based on this prompt.

Examples of ChatGPT Applications and Successes

ChatGPT is an innovative technology in the field of natural language processing.
Below are some examples of successful applications of ChatGPT., an AI assistant that utilizes OpenAI’s GPT-3, is an automatic sentence generation AI assistant that helps users create content such as blog articles, presentations, product descriptions, and advertising taglines using GPT-3. can save time and improve productivity by reducing the amount of labor required to create complicated text.


DALL-E, which utilizes GPT-3, is an image generation AI developed by OpenAI that can automatically generate a variety of images simply by giving instructions in natural language. DALL-E is an image generation technology with a wide range of applications, and is expected to be used in fields as diverse as automotive design, clothing design, and architectural design.

Brava Oven

Brava Oven is an AI microwave oven that utilizes GPT-3 to automatically make optimal cooking settings simply by inputting information such as type of food, cooking time, temperature, and type of ingredients in natural language. Brava Oven also constantly monitors the cooking progress and stops cooking at the optimum time. because of its natural language processing enabled by GPT-3, Brava Oven is easy for users to operate, allowing even those without advanced cooking skills to prepare delicious food without hassle.


Luminar, an AI writer developed by Coto Works of Japan, uses natural language processing technology to automatically generate human-like text. Specifically, it can create documents such as blog posts, product descriptions, news articles, and press releases in a short period of time. In addition, by specifying the field of the text, tone, target audience, etc., more appropriate text can be generated. Furthermore, Luminar’s artificial intelligence automatically learns to improve quality the more times it is used.


Easychat, the AI concierge, is an AI chatbot developed by Japan’s LINE, which utilizes natural language processing technology to support corporate customer service and marketing. Easychat can respond to customer inquiries quickly and automatically 24 hours a day, 365 days a year. In addition, Easychat can learn from past customer inquiry histories to provide more accurate responses. Furthermore, Easychat can handle not only chatbots but also a variety of information such as voice and images, allowing for more flexible responses.

Conclusion and Future Prospects

Summary of the Importance of Prompts and Effective Writing

This paper has described ChatGPT, a large-scale language model, and discussed the importance of prompts. Prompts are a very important part of ChatGPT, and writing them effectively can improve the quality of ChatGPT output.

When writing prompts, keep the following in mind

  • Include specific information and requirements
  • Be precise and clear
  • Be written in a way that is appropriate to ChatGPT’s characteristics

Prospects for ChatGPT’s future development and potential

ChatGPT is one of the most accurate language models developed by OpenAI. It is expected that ChatGPT will be further developed in the future with the release of GPT-4, which is expected to be able to use larger datasets and perform more advanced natural language processing.

  • ChatGPT will be able to perform more advanced tasks.
  • Enables even more human-like natural language processing
  • Expand the range of applications for ChatGPT, making it applicable to a wider variety of fields