I was puzzled by a subscription settings page.
There was a link labelled 「無効」—a Japanese term for “disabled” or “invalid.”
Did it describe the current state? Or did it mean “disable this” if I clicked it?
The same two characters left me unsure what to do next.
I tried clicking, but I was still none the wiser.
Then I clicked a link labelled 「有効」, meaning “enabled” or “valid.” A screen appeared asking whether I wanted to cancel the subscription and why.
Ah. So this was where cancellation entered the conversation.
I told Kimiko-chan about my confusion. Kimiko-chan is the name I gave ChatGPT.
But I did not want our conversation to end at “That was confusing.”
“No, no—this is a reminder to look at ourselves too!”
That was my point. I build interfaces myself.
When I am making a screen, I already know its current state and what each click will do. It is easy to feel that a short label says enough.
As the person using this screen, I could not tell those two things apart.
For example, if I were building an auto-renewal settings page, I would want to separate them like this:
A status message: “Auto-renewal is off.”
An action button: “Turn auto-renewal on.”
Those are my proposed examples, not labels from the page I was using. The words describing a state and the words offering to change it would have different jobs.
What I wanted to bring into my conversation with Kimiko-chan was the need to avoid the same confusion in the screens we work on.
Getting stuck on two characters had given me something to reconsider in my own design work.
“Enabled.” “Disabled.” Conveniently short for the person building a screen, but sometimes the person using it needs a few more words.
Do my own screens leave people wondering, “Which does this mean?”
After looking at someone else’s interface, it was time to look back at mine.