When I asked Claude to rewrite the same rules for my child and grandparents, the wording was completely different
It was the day before summer vacation when I showed an AI a notice for a service aimed at children. It quietly pointed out, "The perspective of the parents is weak." Even though the content was supposed to be the same, I was surprised at how much the wording changed depending on who was reading it.
In this article, I will introduce how to rewrite messages to family members based on "who you are writing to," along with the results of actually trying it out with Claude. It includes real-life examples of how the way you convey things to children, grandparents, and partners changes significantly even with the same rules. The estimated reading time is about 5 minutes.
I thought, "If it's family, they'll understand"
Actually, I hadn't been that careful about messages to my family. I had been getting by with roughly the same phrasing for my children, grandparents, and partner. The point about the "weak parental perspective" in the opening episode was about a work announcement, but I realized that this point might also apply to my own family communications. Even though what they want to know, what they are worried about, and the appropriate level of formality should all be different depending on the person, I had been trying to get by with just one way of phrasing things all along.
Today's demonstration: I had Claude write 3 patterns for the same rule change
During summer vacation, we decided to change our child's game time from "up to 2 hours a day" to "1 hour after homework is finished." First, without specifying the target audience, I just asked Claude.
Prompt: Write a message about changing summer vacation game rules. Claude's response: I'm changing summer vacation game time to "1 hour after homework" starting today! It's to keep our daily rhythm in check, so thanks everyone.
It wasn't bad, but I felt a bit hesitant to send this as-is to the grandparents. Since we are in a position where they look after our grandchild, I felt it would be rude to send it with this level of casualness. So, I specified the recipients and asked again.
Prompt: We are changing our child's game time during summer vacation from "2 hours a day" to "1 hour after homework." Please write a short message conveying this change to the following three people: (1) The elementary school 3rd-grade child (gently, including the reason), (2) The grandparents (a request for when they look after the grandchild, politely), (3) The partner (also asking for their input, confirming a joint decision).
(1) For the child: To [Name], the game rules for summer vacation will change a little starting tomorrow. It used to be 2 hours a day, but from now on, it will be "1 hour after homework is finished." The reason is that if you play games from the morning, your daily rhythm gets messed up. Once you finish your homework, you can enjoy yourself to the fullest for the next hour!
(2) For the grandparents: To Dad and Mom, thank you as always for looking after [Name]. I'm writing to let you know that we've changed the game rules for the summer vacation. It used to be up to 2 hours a day, but we've changed it to 1 hour after homework is finished. It would be helpful if you could keep this rule while [Name] is at your house. If it's difficult, please don't push yourself and feel free to do as you always have.
(3) For the partner: I wanted to consult with you about something. I'm thinking of changing the summer vacation game time from "2 hours a day" to "1 hour after homework." They've been playing games lazily since the morning, and I'm worried about their daily rhythm. Do you think we can go in this direction? Let me know if you have any concerns.
Even though it was the same fact of a "rule change," it wrote with different focuses: reason and a sense of reward for the child, gratitude and a request for the grandparents, and consultation and confirmation for the partner. Honestly, I didn't expect there to be such a difference.
Comparing what changed in the three different approaches
When I lined up the three messages that actually came out, there were mainly three points that had changed.
📋 Three-way check sheet
Recipient: The child / Tone: Gentle, with reasons / Focus: Not disrupting the daily rhythm / Added remark: Positive, saying they can enjoy themselves afterward
Recipient: Grandparents / Tone: Polite, based on gratitude / Focus: Request for cooperation / Added remark: Easing the burden by saying it's okay not to push themselves
Recipient: Partner / Tone: Flat, consultative / Focus: Stance of deciding together / Added remark: Confirmation by asking "Do you think we can go in this direction?"
The content to be conveyed (the details of the rules) is the same for all three. The only thing that changed was specifying "who it is for." Just that alone completely changed the temperature of the text.
Actually, just adding a word about the target audience completely changes the AI's response
This was my biggest discovery this time. While AI may seem omnipotent, it doesn't actually infer 'who the text is intended for' on its own. The reason my initial messages felt somewhat aimless was that I hadn't specified the target audience. Conversely, by simply adding a phrase like 'for children' or 'for grandparents,' the AI's response becomes something entirely different. The amazing thing isn't the AI itself, but the fact that the results change depending on how we ask. This applies not only to family communication but likely to professional writing as well.
Comparing the texts side-by-side
Reading the three texts side-by-side, I honestly don't think I would have come up with the idea of adding a phrase like 'it's okay if you can't manage it' for the grandparents' version on my own. Usually, I tend to lean toward a requesting tone like 'please make sure to follow this,' without considering the burden on the person being asked. For the children's version, I often just convey the rules before explaining the reasons, so having a conclusion like 'after that, you're free to have fun' makes the message feel different to receive. Whether or not I use the AI-generated text as-is is one thing, but realizing my own writing habits was the biggest takeaway.
Things to keep in mind when using it
While convenient, there are a few things to be careful about. I recommend reading through the AI-generated text in your own words before sending it. Especially for the grandparent generation, the phrasing chosen by the AI can sometimes be stiffer or, conversely, too casual than expected. Also, for safety, do not include specific personal information about your family (real names, addresses, school names, etc.) in the prompt; use placeholders like 'XX' instead. AI suggestions are just a starting point, and you should make the final decision on whether or not to send it yourself.
AI terms explained in 1 minute
The term 'prompt,' which appeared many times here, refers to the instructions or questions you give to an AI. Even when asking about the same content, simply adding conditions to the prompt like 'who it is for' or 'what tone to use' will significantly change the returned answer. As demonstrated here, just adding the target audience is enough to create a sufficient difference, so please try starting with that.
Summary
There are three main points today. First, the points emphasized in family communication differ depending on the recipient. Second, simply adding the target audience to your AI prompt will significantly change the response. Third, it is safe to always read through the text with your own eyes before sending it. As a next step, the next time you have an opportunity to communicate something to your family, try telling the AI 'who the recipient is' before making your request. You will surely get a different sentence back than before.