*Este artículo fue creado utilizando entrada de voz con IA (Aqua Voice). Tenga en cuenta que puede haber algunas inconsistencias.
Hola. Soy Hiro del Laboratorio de Full Body Tracking.
Last time, I wrote about how organizing your thoughts is not a talent but a habit.
This time is the practical how-to edition. I’ll write about the “thought reorganization using Claude Code” and the “system for reminding yourself of your own ideas when you’ve forgotten them” that I actually practice.
To cut to the conclusion: when you have AI read your thoughts back to you, you gain a surprising degree of objectivity.
Why “Read Aloud”?
When you read text you’ve written yourself with your eyes, you don’t notice many of the issues.
The author knows the context, so your brain automatically fills in gaps even if something is a bit off. But when you hear it as audio, things change.
When it enters through your ears, you clearly catch things like “Wait, isn’t this part a logical leap?” or “I said the same thing twice here.”
The version of yourself that reads with your eyes and the version that listens with your ears feel like completely different people.
Using this, I play back content I’ve organized in Claude Code using my PC’s text-to-speech.
The Setup: Claude Code x VOICEVOX
What I’m doing is simple.
The Flow
(1) I throw my thoughts and notes into Claude Code
(2) Claude Code organizes them and writes them out as MD files
(3) I have that content read aloud using a speech synthesis tool like VOICEVOX
(4) I listen with my ears and jot down anything that feels off or any new insights
VOICEVOX is a free speech synthesis tool that reads aloud text you input using character voices. You can choose from various voices like Zundamon or Shikoku Metan.
When your thoughts are read aloud in a character’s voice rather than your own, it adds even more objectivity. You get the feeling that “someone else is explaining my ideas,” which makes it much easier to push back and find holes.
When You Converse Through Voice, the Quality of Thinking Changes
Another important thing is conducting your dialogue with AI on a voice basis.
Speaking brings out the flow of your thoughts more naturally than typing on a keyboard.
When typing, you tend to try to polish things into proper sentences. But with voice, things come out in the order you think of them, so ideas you weren’t even aware of sometimes surface.
When you throw this to Claude Code, AI organizes it logically.
Your own intuitive thinking x AI’s structuring ability. This synergy is quite powerful.
In my experience, combining voice input with my workflow produces output about 2-3 times faster than using a keyboard alone. And the feeling that it’s written in my own words is preserved.
Why MD Files in Obsidian?
Let me add some context here: Claude Code alone makes it difficult to create a system for “reminding you of things when you’ve forgotten.”
Claude Code sessions are basically one-time. It doesn’t spontaneously recall past exchanges. There is a memory feature, but it can’t yet perform the trick of surfacing “something you were thinking about three months ago” at just the right moment.
That’s why I save content organized by Claude Code as Obsidian MD files.
Obsidian is a note-taking app that manages Markdown files locally, with the distinctive feature of being able to link files to each other.
As Claude Code’s organized thought logs accumulate in Obsidian, you can search through them later or connect related notes to one another.
The true identity of “reminding yourself when you’ve forgotten” is loading the thought logs accumulated in Obsidian back into Claude Code when needed.
Claude Code forgets every time, but Obsidian doesn’t forget. That’s why I use Obsidian as Claude Code’s long-term memory.
When Using NotebookLM
When using NotebookLM, the approach changes a bit.
NotebookLM is a Google service, so it naturally has great compatibility with Google Docs. Rather than loading Obsidian’s MD files directly, I copy them into Google Docs and then load those into NotebookLM.
NotebookLM’s strength is the ability to ask questions across multiple documents.
You can ask something like “Among everything I’ve written over the past three months, what themes keep coming up repeatedly?” This is another method for “reminding yourself when you’ve forgotten.”
You’d forgotten about it, but you were actually thinking about the same thing over and over. When you notice that, you realize “This must be a really important theme for me.”
About Voice Input
If you’ve read this far and thought “Isn’t voice input kind of a hassle?” — let me address that.
I use a voice input service called AQUA VOICE. It converts spoken words into text in real time, and the accuracy is quite high.
That said, the options for voice input have expanded considerably recently.
The standard voice input on iPhone has improved in accuracy, various Whisper-based apps are out there, and Google voice input is available free on the web.
I plan to write a detailed comparison of these tools in a separate article, so for now I’ll just mention that I use AQUA VOICE.
Regardless of which tool you use, the important thing is having the flow of “speak -> convert to text -> throw it to AI.” The tool doesn’t matter much. What makes all the difference is whether or not you have that flow.
Putting the Whole Flow Together
Here’s how my thought-organizing cycle works.
Input (Daily)
Take notes with voice input (AQUA VOICE, etc.). No keyboard. Speak the moment something comes to mind.
Organization (Claude Code)
Throw notes to Claude Code. Claude Code structures them and writes them out as MD files. They get saved in Obsidian.
Objectification (VOICEVOX Read-Aloud)
Have the organized content read aloud by VOICEVOX. Listen with your ears, and note any inconsistencies or new insights.
Accumulation (Obsidian)
Organized MD files accumulate in Obsidian. They connect through links and are searchable. This becomes your long-term memory.
Rediscovery (NotebookLM)
Load accumulated documents into NotebookLM via Google Docs. Discover your past thinking patterns.
Remembering When You’ve Forgotten
Months later, when you load past Obsidian notes into Claude Code for work, your forgotten ideas naturally resurface. You have the moment of “Oh, that’s what I was thinking about.”
The Essence of This Method
I’ve talked a lot about tools, but the essence isn’t about tools.
The essence is “externalizing your thoughts and returning them to yourself with a time delay.”
Humans forget. That’s not a weakness — it’s how the brain is designed.
So you build a system that’s fine with forgetting. When you’ve forgotten, AI and tools remind you of what your past self was thinking.
In that moment, it feels like your past self and present self are having a conversation.
“The thing I was thinking about back then — I can use it this way now!”
I believe this experience generates value that goes beyond mere information organization.
Resumen
Organize your thoughts with Claude Code, and listen to them read aloud by VOICEVOX for objectivity. Converse with AI through voice input, and generate synergy between your intuition and AI’s structuring ability.
Save organized content as MD files in Obsidian, and rediscover past thinking patterns through NotebookLM.
This is the full picture of “how to remind yourself when you’ve forgotten.”
No special equipment or expensive software needed. All you need is the habit of externalizing your thoughts and a flow of tools to catch them.
The era of thinking only inside your own head is over. Think together with AI, remember together. That’s my method of organizing my mind.
¡Gracias por leer! ¡Síguenos en X y suscríbete a nuestro canal de YouTube!