Why Talking Out Loud Produces Better First Drafts
The blank page is not a typing problem. It is an editing problem. When you type, you edit every sentence as it appears, and that constant self-correction is what stalls a first draft. Talking out loud sidesteps it, because your mouth moves faster than your inner critic.
Key takeaways
- Most people speak around three to four times faster than they type, so ideas land before you second-guess them.
- Speaking bypasses the inner editor, which is the real cause of blank-page paralysis.
- A spoken draft is meant to be messy: AI cleanup fixes filler and punctuation afterward.
- On Mac, BlaBlaType turns speech into text in any app and keeps every word 100% on-device.
The inner editor is what freezes a first draft
Writing a first draft and editing a draft are two different jobs, and they fight each other. Drafting wants speed and momentum. Editing wants precision and doubt. When you type, both happen at once: you write four words, delete two, rephrase, and lose the thread. That back-and-forth feels productive, but it is the single biggest reason a first draft never gets finished.
Speaking separates the two jobs by force. You cannot easily backspace a spoken sentence, so you keep going. That momentum is the whole point of a first draft. You are not trying to write well yet. You are trying to get the raw material out of your head and onto the page where you can actually work with it.
Speaking is faster, and speed changes how you think
The plain mechanical fact helps too: most people speak around three to four times faster than they type. That gap is not just about finishing sooner. When your output keeps pace with your thinking, you follow an idea to its natural end instead of losing it while your fingers catch up. If you want the background on how this is measured, the concept of words per minute covers both speaking and typing rates.
This is why dictation feels different from typing rather than just quicker. You talk in full thoughts. You explain something the way you would explain it to a colleague, and that conversational register is usually clearer and warmer than the stiff prose people produce when they type carefully. You can always formalize later. Getting a human voice on the page first is the hard part.
But spoken drafts are messy, right? Yes, and that is fine
The usual objection to dictation is that raw speech looks rough on the page. It does. There are filler words, half-restarts, and almost no punctuation. That used to mean a spoken draft created as much cleanup work as it saved. Modern on-device AI cleanup changes the math. It removes filler, fixes punctuation and grammar, and adapts the tone, so you get a readable first draft instead of a transcript to untangle.
Here is what that transformation actually looks like on a real spoken sentence.
so um basically what i wanted to say is like the launch went well i think but uh we need to you know follow up with the people who signed up and didnt actually try it yet
The launch went well. We need to follow up with the people who signed up but have not tried the product yet.
Same idea, captured in one breath, now readable. You spoke for six seconds and skipped the part where you would normally stare at the cursor rewording that sentence four times. This is also why voice works so well for routine writing like email. If your inbox is the bottleneck, our guide on how to dictate emails on Mac walks through the exact setup.
How to actually draft by talking on a Mac
Talking out loud only helps if the words land where you are working, without breaking your flow to copy and paste. That is the difference between real system-wide dictation and a separate transcription window.
| Approach | Types into any app | Cleans up speech | Stays on-device |
|---|---|---|---|
| Typing by hand | Yes | No | Yes |
| Built-in Mac Dictation | Yes | No | Mixed |
| File transcription apps | Files only | Rarely | Yes |
| BlaBlaType | Yes | Yes | Yes |
BlaBlaType runs system-wide, so a single shortcut lets you dictate into your editor, notes app, email, or an AI chat, and the on-device AI cleanup happens before the text appears. Speech recognition and cleanup both run locally on Apple Silicon, so your audio and transcript never leave the Mac. That last point matters when the draft is unpublished, confidential, or client work. If privacy is your first concern, we break it down in the privacy stack. Apple also offers built-in dictation if you want to try the idea before adding AI cleanup.
Draft with your voice, not the blank page
Talk your first draft into any app on your Mac. On-device AI cleans it up, and nothing leaves your machine. No card needed for the trial.
Download for macOSWhen talking out loud is not the right tool
Dictation is for the drafting stage, not every stage. Editing is still a keyboard job: cutting, reordering, and fine word choice are visual, spatial tasks that voice handles poorly. Highly structured writing with tables, code, or dense formatting is also better typed. And if you share a room or an open office, speaking your draft out loud is not always practical. The honest rule is simple: talk to generate, type to refine. Used that way, voice removes the hardest friction in writing, which is starting. For a wider look at the tools, see the best dictation software for Mac, and if you are choosing a plan, the pricing page lays out the trial.
Frequently asked questions
Is speaking really faster than typing for a first draft?
For most people, yes. Most people speak around three to four times faster than they type, so a spoken first draft usually comes out quicker. Speed is not the only benefit though. Talking also keeps you moving before the inner editor can stall you.
Does a spoken draft not come out messy?
It does, and that is fine for a first draft. Raw speech has filler words and loose punctuation. On-device AI cleanup can remove filler, fix punctuation and tidy grammar, so you get a readable draft to edit rather than a transcript to untangle.
Do I need to talk to my Mac like a robot?
No. You talk normally, in full sentences, at your own pace. Modern on-device models handle natural speech, and a custom dictionary teaches the app names and jargon it might otherwise mishear.
Is my voice sent to a server when I dictate a draft?
With BlaBlaType, no. Speech recognition and AI cleanup run 100% on-device on your Mac, so your audio and transcript never leave your machine. That matters when you are drafting sensitive or unpublished work.
Can I edit the draft afterwards?
Yes, and you should. Talking out loud is for the first draft, when the goal is to get ideas down. Editing with your keyboard afterwards is where you shape structure, cut, and polish. The two steps work best kept separate.