You know those ideas that show up at the worst possible moments to write — while you're driving, walking, doing the dishes? We mutter them, promise ourselves we'll jot them down, and almost always lose them. Voice is the most natural way to capture a thought. The hard part has always been turning it into something useful. Let's see how, step by step.
Why speaking is faster than writing
The numbers are clear: on average we speak at 120-150 words per minute and type at about 40. But speed isn't the real advantage. The real advantage is that, when we speak, we think out loud: we don't stop to fix the punctuation, we don't judge the sentence while it's forming. The thought comes out whole, with its tone and its useful digressions.
Writing, on the other hand, switches on your inner editor straight away: you slow down, you polish, and you often lose the original idea along the way. To capture, it's better to speak; to refine, it's better to write. The trick is to separate the two moments — and let the tool build the bridge.
From voice to text: how it works
Automatic transcription uses speech recognition to convert audio into text. Quality depends a lot on the language: an engine optimized for the language you speak handles accents, proper names, clipped words and that “messy”, real speech we all use every day much better.
Itomori transcribes in your language and — an important detail — keeps the raw transcript as the source of truth. Your original thought stays intact, even when you derive cleaner versions from it. It's a guarantee: the tool doesn't rewrite what you said behind your back; it always starts from your words.
Don't stop at text: from note to content
This is the real leap, the one most dictation apps don't make. Once the thought is text, it usually just sits there: a raw block to rework by hand. Itomori, instead, turns it into the format you need:
- a speech or an outline for public speaking;
- a post for social media, already in the right tone;
- a tidy email, with ideas in the right place;
- a lesson or structured teaching material;
- a summary of the key points.
The same raw thread becomes several things, without starting from scratch every time. It's the Create phase of the Capture · Connect · Create method.
Three concrete use cases
- The idea on the go. 60 seconds of voice in the supermarket queue become the draft of a post: you record it now, and refine it tonight with a fresh mind.
- The lesson or the talk. You think through the outline of a talk out loud while you walk; Itomori structures it into a speech with an opening, points and a close.
- The newsletter. You collect three voice notes scattered across the week and turn them into an email that almost writes itself, instead of staring at a blank page on Sunday evening.
Practical tips for recording well
A transcript is only as good as the audio you give it. A few precautions make a big difference:
- Speak off the cuff, not perfectly. Hesitations can be removed later; a lost idea doesn't come back.
- One note, one thought. Three short notes are better than a ten-minute monologue on five topics: they connect and are found more easily.
- Cut down the noise. In the car or on the street is perfectly fine, but avoid loud music or wind blowing straight into the microphone.
- Say the context at the start. A sentence like “idea for the launch post” helps you (and the AI) find and use the note.
Voice or keyboard? When to use which
Voice doesn't replace writing: it comes before it. They're good at different things, and using each at the right moment changes the result.
- Use your voice when the idea is still warm and shapeless, when your hands are busy, when you want to think out loud without judging yourself, or when you need to capture something quickly before it fades.
- Use the keyboard when you need to be precise (numbers, codes, exact quotes), when you're refining a text that will be around for a long time, or when you're somewhere you can't talk.
The ideal sequence, almost always, is: voice to capture, keyboard to refine. Itomori is exactly the bridge between the two — you record the raw version, then step in only where it's needed.
What to do with the raw transcript
A common question: if I then generate a “clean” post, why keep the original transcript, with all its hesitations and “I mean”s? It matters a lot. The raw version is the source of truth: it's what you actually said, in your words and in the order your mind put them. Clean versions are interpretations; the original isn't.
In practice that means two things. First: you can always go back to the source if a rewrite has “softened” a point you wanted to be sharp. Second: search and connections work on the raw text, so you find a note even with the spontaneous words you used in the moment, not just the cleaned-up ones. That's why Itomori never overwrites the original: it keeps it.
“I'm no good at talking into a microphone”
It's the most common objection, and it's almost always untrue. You're not making a podcast: you're talking to yourself. You don't need a trained voice, or complete sentences. The “I mean”s, the pauses, the second thoughts are all perfectly fine — in fact, it's often in the second thought that the good idea arrives.
The secret is to treat it like a phone call to a friend you're explaining something to: start with “so, I wanted to tell you that…” and let it flow. The first two or three times it feels strange; by the fourth it's the most natural thing in the world, and you realize you're capturing thoughts you used to lose every single time. And the transcript gives you back an orderly text anyway: your job is only to say, not to perform.
Try it on Itomori
The best way to understand it is to record your next thought instead of writing it. See how it works on the Product page, and if you often work in a team, see how to capture and summarize meetings.