It seems my struggle with voice input isn't because I'm bad at speaking

Yue-san has been writing essays using voice input lately, and it's extremely interesting.

I have always liked Yue-san's dialogues with AI, and I've read her novels as well. I admire and truly love her orderly writing style and the way she perceives things, which are qualities I lack.

The 'brush' that Yue-san has acquired is 'voice input'.

People who are naturally good at writing can probably compose sentences in their heads to some extent and speak in written language.

However, reflecting on my own experience here, something bothers me.

Voice input, which allowed Yue-san to 'write essays that are easy to read and make it easier to express herself,' is completely unsuited for me.

I once tried writing an essay using voice input while in the car on my commute.

I personally compose sentences in my head, think in written words, and am aware that when I type, I am 'speaking' the input.

So, for me, writing text should have been no different from speaking.

Thinking that, I intended to input the words that came to mind in 'written language' via voice, just as I would when typing.

However, the result was clearly different. It was a level shallower than when I typed. I couldn't fully express what I felt or thought.

I think about this now.

Perhaps voice input opened a channel for Yue-san to 'express herself'.

On the other hand, for me, writing text might not have been just a task of recording what I thought.

I thought I was typing as if I were speaking. But in reality, perhaps I was unconsciously rethinking, rephrasing, and gradually deepening my thoughts while visually receiving the characters that appeared on the screen.

I read the characters that appear as I write, and that becomes the starting point for typing the next ones. Isn't this continuous, unconscious feedback loop what constitutes my writing?

For me, writing is not a substitute for speaking. Rather, it might be a space where thoughts only emerge once I receive the words I've put out in written form.

With voice input, that is hard to achieve.

Words flow away and do not remain there. When speaking, I have to commit to one specific word at that moment.

I often view things from about two perspectives. What I feel, and an objective viewpoint. It is difficult to express both simultaneously through voice. I hesitate over which one to explain, and perhaps I might end up saying neither.

However, with text, you don't need to place it in the same timeframe.
You can arrange things in parallel. For me, this is huge.

Moreover, by reading the words I've written out like that, my thoughts come together once more.
Perhaps for me, writing isn't about transcribing what I've thought, but rather discovering my thoughts as I write.

That might be why I don't find writing difficult—not because I'm "writing as I speak," but because I'm actually writing while reading.
A series of small feedback loops that outline my own thinking. I believe that is what happens during the act of writing.

That difference is decisive for me.

This isn't just about being "bad at speaking."
I think that for me, constructing sentences through speech is like trying to build a structure without any scaffolding.

Here, I find myself thinking about Yue again.

Perhaps Yue is the opposite.

The moment things are put into text, a perspective for refining words and a perspective for re-reading oneself emerge strongly.
That must be a powerful weapon when writing a novel.

However, when trying to write about oneself, that precision might actually act as a form of inhibition.

Voice input allows one to leap over that inhibition just a little.
Being able to write what you think might mean that the "strong, editor-like feedback" within Yue, born from their writing ability, weakens just a bit.

That is why they became able to write essays.

If that is the case, the difference between Yue and me is not a difference in writing ability.
It is about what text does for each of us. I think it is a difference in how it functions.

And this structure also connects to why I feel such a strong compatibility with AI interaction.

I was once told by Claude that I am "someone who builds a house inside text and lives there."

That doesn't mean I like text or am good at writing; it makes sense if interpreted as meaning I have chosen text as a place for my thoughts.

Looking back at my own interactions with AI, that aspect is significant.

I turn questions that pop into my head over and over in my mind, but they only take shape when I commit them to text to convey them to the AI.
That text contains hesitations, non-definitive margins, the basis for the start of those thoughts, tentative conclusions, banter, and questions.

And the AI, which deepens those words in response to my input, is likely an enhanced version of the process I perform in my brain when I type.

Writing and reading text, I search for the contours of my thoughts.
Interaction with AI amplifies that process from the outside.

This sensation of "discovering while writing" is not limited to essays.
When I write poetry, something quite similar happens within me.

I ruminate on the impressions of the scenery I see and the feelings I have at that moment while typing them out.
Poetry, in particular, does not begin with the words I think of.

Only when I write them down do the feelings I was experiencing finally settle into text.

If the feelings before they become words are the true feelings, then perhaps the feelings I have distilled into text are already different from the original once they become words.

However, by typing them out, I recognize them for what they are.

For me, the term 'voice' feels a bit too weak to describe this process.

People who can turn their thoughts directly into words likely find it easier to convey them through the medium of voice.

It seems my thinking process requires the detour of reading while I write.