I Stopped Typing for a Week. Here Is What Actually Happened.

I set the rule on a Monday morning: no typing for work output. Emails, documents, notes, Slack messages. All of it had to be spoken. Passwords and code were exempt. Everything else was fair game for the experiment.

By Monday afternoon I had already cheated twice.

Day One: The Friction Is Real

The first thing I noticed was how often I reached for the keyboard by reflex. Not for long documents. For tiny things. A two-line reply to a colleague. A note to myself. A search query. These micro-typing moments are so automatic that breaking them felt like learning to write with the other hand.

I set up VoiceInk on my Mac and assigned a thumb key on my keyboard as the push-to-talk trigger. Press, speak, release. The text appeared in whatever field I was in, almost instantly. The setup took ten minutes. The habit change took considerably longer.

My email replies were slower on day one. Not because transcription was slow, it was not. But because I was not used to speaking in complete sentences without rehearsing them first. Typing lets you draft privately, word by word. Speaking commits you to a direction immediately. That felt exposed, even with no one in the room.

Day Three: Something Started to Shift

By Wednesday I had written more in two days than I typically write in four. Not because I was trying harder. The volume just came out faster when I was not rationing words through my fingers.

I also noticed that my writing was less hedged. My typed prose tends toward qualifications: "this might suggest," "in some cases," "one could argue." My spoken prose was more direct. I said what I meant because I was not stopping to second-guess each clause before committing it to the page.

A colleague asked about one of my emails: "Did you have more time this week? That was the clearest thing you have sent me in months." I had not had more time. I had just talked instead of typed.

The Hard Parts Nobody Mentions

Open offices, or even shared apartments, make voice input complicated. I work from home, but my partner was on calls from the next room two days that week. I dropped back to typing those afternoons rather than narrate documents over her meetings. If you share a space, this is a real limitation.

Proper nouns were also a recurring frustration. My own last name came out wrong three times. A client's company name required spelling out every session. Modern transcription is extremely good, but it is not telepathic. Unusual words take a few corrections before the system learns them, or you learn to say them more clearly.

Day Seven: The Honest Accounting

By Friday of the second week I had no desire to go back to all-typing. But I also had no desire to go fully voice. What I found was that the two modes are genuinely complementary, not competitive.

For long-form writing, emails, documents, and notes: speaking is faster and the output is often better on first pass. For editing, code, and anything requiring precise character-by-character input: typing is the right tool. Treating them as rivals is the wrong frame.

The most useful thing the experiment gave me was a recalibrated default. Before, I typed everything without thinking. Now I make a small, deliberate choice at the start of each task. That choice has made me faster overall by a margin I can actually feel, not just measure.

Whether You Should Try It

You do not need a full week of commitment to learn something from dictation. A single afternoon of using your voice for everything you would normally type is enough to surface the moments where it helps and the moments where it does not.

The barrier to starting is low. If you have a Mac and any external microphone, you have what you need. The rest is just talking.