This Simple Audio-to-Article Workflow Saves Me Hours.
A note before you read: The audio above is me, thinking out loud, mostly unedited (save for a pass through Descript to cut filler words and silences). The article below is an AI-generated draft from that transcript, which I then edited for clarity and personality. That's the format here, and this first one is literally about why I'm doing it this way.
(If you want to try this yourself, I broke down my exact workflow at the bottom of this page.)
Welcome to the KJ CAMRON Audio-Sonic Experience.
I have a new-ish thing I'm doing. And this is it.
I talk too much. Always have. Thinking out loud is how I brainstorm, how I work through ideas, how I figure out what I actually believe about something. The problem is most of that thinking just lives in my head, evolves for a while, and eventually dies there. Or I sit down to write it out and eight hours later I have a thousand words and a migraine.
So I'm changing the workflow.
From now on, when an idea is strong enough to say out loud, I'm going to record it. Then I'll drop the transcript into Claude, get a first draft back, edit it for clarity and voice, and post both: the raw audio at the top of the page and the cleaned-up article below it. You can listen to the chaos or read the organized version of it. Your call. I’d recommend both.
There's a reason I'm keeping the audio. Most people — writers, consultants, creators, thinkers of all kinds — present you the finished product of their thinking. Clean. Polished. All the dead ends removed. And that's valuable, but it's not the whole picture. What I find genuinely interesting is watching an idea evolve in real time, before it gets refined. The process is where the real thinking lives. The messy part before the clarity. Miley Cyrus said it herself: it's the climb. (Yes, that Hannah Montana reference was in the transcript. Yes, I kept it. You get to hear the chaos in my head.)
I want to be clear about a few things:
Not everything on this site will be done this way. Some pieces will be fully written, the old-fashioned way. Writing is a practice I'd never drop entirely, as it forces a kind of slowness and precision that speaking doesn't. I need to keep that muscle. But for ideas that arrive fast and want to get out fast, this is the new lane.
Also, I'm recording these on my phone’s built-in recorder app, wherever I can find a quiet-ish place. Not in an audio studio. Not planned. Just spontaneously, when something is worth saying. Expect that unpolished energy.
One more thing: I'll always be transparent at the top when an article was drafted this way. You deserve to know what you're reading and how it was made. The thinking is mine. The voice is mine. The draft was Claude's. The edit was mine. That's the deal.
Anyway. Welcome.
Want to try this yourself? Here's my exact workflow.
The whole thing takes under 30 minutes once you have the tools set up.
1. Record. Use whatever's in your pocket. Voice memos on iPhone, Samsung's built-in recorder, doesn't matter. Aim for 3–5 minutes. No script, just talk toward a point. And if you don’t like the sound of your own voice, I promise, you’ll get over it after a few reps.
2. Clean the audio. Run it through Descript. Their "Remove Filler Words" and silence-removal features will cut a 6-minute ramble down to something tighter. Export the cleaned audio and the transcript.
3. Draft the article. Drop the transcript into Claude (or your AI tool of choice) with a simple prompt: "Turn this transcript into a clean, readable article. Keep the voice and any personality, remove the repetition." You'll get a workable first draft in seconds.
4. Edit for voice. Read it back. Fix anything that doesn't sound like you. Add anything you glossed over. This is the only part that requires real attention — budget 10 minutes.
5. Publish both. Embed the audio at the top of the page. Put the article below it. Label it clearly: the audio is unedited thinking, the article is the AI-assisted write-up that you edited. Readers can choose their format.
That's it. The thinking is yours. The voice is yours. The AI just handles the transcription gap.
