Prompt Factorization Studies (6) mini — Summary: Visualizing <Thinking> with P×R×C×F [Column] The Science of Thought (No. 85)

🎯 The Aim of This Column (3 Points)

• Reconfirm how dialogue with AI can be "designed" through the "Prompt Factorization" model (P×R×C×F).

• Clearly grasp the overall picture of prompt design by reorganizing the four factors—P (Purpose), R (Role), C (Context), and F (Format)—with practical examples.

• Explore the possibilities of a new thinking technique in the AI era—"Prompt Factorization Studies"—from the perspective that "questions can be created."

◆ Introduction: What kind of questions have we asked in this series?

In this "Prompt Factorization Studies" series, we have been breaking down the structure of prompts to have better dialogues with generative AI like ChatGPT.

Why do such a thing?

The answer is that the "questions we can pose" to AI will determine the quality and quantity of our intellectual production from here on out.

In the past, a question was a "means to obtain an answer."

However, in dialogue with AI, unless the question itself is"designed", you cannot reach the expected answer.

A question is not something you just throw out; it is something you design and "create."

🔍 Why was "Prompt Factorization" necessary?

For example, a vague question like "What do you think about this?"

A human might be able to grasp the meaning from the preceding conversation or situation, but ChatGPT is an AI without "context."
Therefore, you must explicitly state in "words" what purpose you have, who you are as, what you are talking about, and in what format.

In other words, a prompt is not a spell to extract information, but an"interface for thinking."

What is an interface for thinking?

To put it simply, this "interface for thinking" is—

🔍 To put it a little more carefully...

In other words, an "interface for thinking" is,

.

🧠 Why is a prompt an "interface for thinking"?

For example, when you ask ChatGPT a question,
the AI cannot understand it if you are just thinking about it.
However, by putting it into the form of a "prompt" and sending it, the AI becomes able to respond to that question.

That means you have extracted your thoughts as a "blueprint"
and put them into a form that the AI can understand.

✅ In summary:

An interface for thinking is,

The fact that a "prompt" can be that—is the core of Prompt Factorization Studies.

🔁 Reflection question: How have you been "asking" ChatGPT?

In this series, we have thought about prompts by breaking them down into the following four elements.

• $${P}$$ (Purpose) = Purpose: What do you want it to do?

• $${R}$$ (Role) = Role: Who do you want it to answer as?

• $${C}$$ (Context) = Context: What are you talking about?

• $${F}$$ (Format) = Format: How do you want it to answer?

Did the questions you have sent to ChatGPT so far include these elements?
Or, have you been unconsciously drawing out vague answers with vague questions?

In this final installment, we will re-examine "what a prompt is" based on the "blueprint of questioning" we have learned so far.

And finally, I ask you, as a "designer of questions,"what kind of questions do you want to create?

◆ Re-presentation: Prompt Factorization Formula and Prime Factor Model

Let us summarize our considerations so far into a single formula.

💡 Prompt Factorization Formula

$${Prompt = P × R × C × F}$$

This formula shows that the “questions” we pose to generative AI can be structured from the following four perspectives.
Each element can be further broken down into “prime factors,” which dramatically increases the precision of prompt design.

• $${P}$$ (Purpose)
  　┗ The type of desire for thought
  asking “What do you want me to do?”　┗ Prime factors: Understanding, Judgment, Action

• $${R}$$ (Role)
  　┗ The design of persona
  asking “Who should I speak as?”　┗ Prime factors: Knowledge level, Stance, Social perspective

• $${C}$$ (Context)
  　┗ The structuring of topics
  asking “What are we talking about?”　┗ Prime factors: Domain, Subject, Situation

• $${F}$$ (Format)
  　┗ The output format
  asking “How do you want me to answer?”　┗ Prime factors: Style, Structure, Constraints

Thus, a “prompt” is not just a single sentence, but a “blueprint for inquiry.” By consciously combining each element, ChatGPT’s output will return more accurately, more deeply, and more in line with your intentions.

Now, let us review this in the form of a formula.

🌈 Factorization of Prompts

$${Prompt=P×R×C×F}$$

• $${P}$$ = Purpose: What do you want me to do?

• $${R}$$ = Role: Who do you want me to answer as?

• $${C}$$ = Context: What is the topic about?

• $${F}$$ = Format: In what form do you want the answer?

💡 Prime Factorization of Each Factor

🪶$${P(Purpose)=P_1(Understand)×P_2(Evaluate)×P_3(Generate)}$$

$${P_1}$$ = Understand
　┗ The desire to "know" or "grasp." Example: Summarize / Explain / Visualize

$${P_2}$$ = Evaluate　┗ The desire to "compare," "choose," or "assess." Example: Compare / List pros and cons / Grade

$${P_3}$$
= Generate　┗ The desire to "create," "propose," or "act." Example: Write / Brainstorm / Create a post

🪶$${R(Role)=R_1(Knowledge Level)×R_2(Stance)×R_3(Perspective)}$$

$${R_1}$$ = Knowledge Depth
　┗ Expert / Layperson / Beginner / Child, etc. At what
"depth" do you want me to speak?
$${R_2}$$ = Tone / Attitude　┗ Strict / Gentle / Calm / Enthusiastic / Humorous, etc. With what

"emotion" do you want me to speak?$${R_3}$$
= Perspective　┗ As a parent / As a boss / As a user / As a person from the future, etc. From what
"standpoint"
do you want me to speak?

🪶$${C(Context)=C_1(Domain)×C_2(Target)×C_3(Situation)}$$

$${C_1}$$ = Domain
　┗ Specifies the
”field of knowledge”, such as medicine, business, education, philosophy, or social media.
$${C_2}$$ = Target　┗ Indicates the

”specific topic or object”, such as an article, product, idea, statement, or personal experience.$${C_3}$$
= Situation　┗ A factor of
”when, where, and to whom you are speaking”. For example, "just before a presentation," "before posting on social media," or "for students."

🪶$${F(Form)=F_1(Style)×F_2(Structure)×F_3(Length/Constraints)}$$

$${F_1}$$ = Expression Style
　┗ Specifications regarding tone, vocabulary, and writing style, such as gentle, academic, casual, or polite.

$${F_2}$$ = Structure　┗ The "shape" of the output, such as bullet points, paragraphs, comparison tables, or a story format.

$${F_3}$$ = Length & Constraints
　┗ Specific volume requirements, such as within 200 characters, in 3 lines, in one sentence, or readable in 5 minutes.

In the next section, we will examine how to design better prompts by applying this factorization model to actual examples.

◆ Practical Example: Redesigning the Question

Let's experience how much the "quality" of ChatGPT's response changes by altering the design of the question.

This time, we will compare using the following source text as an example:

📝 Source Text (Target for Revision):

🎡 Before: Vague Prompt

Example Response:

📌 Explanation:

• The expression has been slightly polished, but the meaning and structure remain largely unchanged.

• Due to the absence of design regarding who it is for, what the purpose is, and how it should be polished, it remains a mechanical correction.

🎯 After: Designed Prompt

Example Response:

📌 Explanation:

• P (Purpose) = Empathy for anxiety and providing reassurance

• R (Role) = For beginners, gentle, mother's perspective

• C (Context) = Career education, blog post, for SNS

• F (Format) = Conversational tone, opening introduction, 300-character limit

→ The 'Before' is text proofreading, while the 'After' is 'a message that reaches the heart.' This truly shows the difference resulting from designing 'what you want from AI'.

🧠 Summary

The difference in prompts is, in other words, 'the difference in the design of the question'.

By clarifying what you are asking the AI and what kind of answer you are seeking, the AI's response evolves from mere correction to empathy, co-creation, and dialogue.

◆ Finally: You are the owner of the question

When you ask ChatGPT a question, that question is imbued with the shape and intent of your own thinking.

What is the purpose of that question?
As whom do you want it to speak? About
what topic do you wish for it to respond, and in what format?

ChatGPT demonstrates its power precisely when faced with clear questions. 'Prompt Factorization Studies' is a method to visualize the structure of those questions and realize deeper, richer dialogues.

A prompt is an interface for your thinking.
Becoming conscious of its design is the first step toward improving the quality of your own thinking and dialogue.

Furthermore, this factorization-based thinking will be useful not only for prompts to AI, but in all situations, including business, education, and daily life.

❓ A final question for you

What kind of question would you like to factorize?
What new dialogues and discoveries might be born from that question?

🧠 This column was written in co-creation with Generative AI/ChatGPT. It is an attempt to explore new forms of intellectual collaboration between humans and AI by focusing on the 'structure of questions'.

Profile | Kazuomi Matsunaga

🧭Thinking Navigator

Biography:Graduated from Kyushu University, Faculty of Engineering (Materials Engineering) ⇒ Engineer at a metal materials manufacturer in the Chukyo region (Production Technology) ⇒ Advertising sales for a local free paper in Nagasaki Prefecture ⇒ University staff member (← Currently here). Living in Nagasaki Prefecture, in my 50s. Both of my children have graduated and left for college, and I am now living with my wife. I am a writer who lets my thoughts play daily at the intersection of "questions" and "curiosity." Recently, I have been writing about my unique perspectives on themes such as education, stories, psychology, AI, and design, centered on my serialized column, "The Science of Thinking." I am exploring "free knowledge design" that goes beyond existing frameworks. My motto is "Imagination opens up the world."
In my professional life, while working in university and education-related fields, I value carefully observing the moments when "people start to like something" and the scenes where "questions sprout."
🧭 I hope to deliver articles that leave you with a "new perspective" after reading them. I would be happy if I could do that.
📮 Feedback and messages of support are always welcome!