Article·Nov 18, 2025

I Type 107 WPM and I Still Use Voice AI When I Work. Here's Why.

Typing speed isn't the real bottleneck to efficiency. What actually is may surprise you.

4 min read

By Sharon Yeh

Last Updated

I can fly across a keyboard. My typing speed consistently clocks in at 107 words per minute (see below for proof). So when people see me talking to my computer when I have a keyboard, they're puzzled. "Why bother with voice when you can type that fast?"

Here's the thing: typing speed isn't the real bottleneck.

The Hidden Friction in Fast Typing

Here's something to try: next time you're writing an email or document, pay attention to what's actually happening. You'll notice you pause – sometimes just for a split second – before typing each sentence. "Should I say 'Thanks for the update' or 'Appreciate the heads up'?" "Is this too casual?" "Does this sound right?" You might not realize it, but you're constantly making these micro-decisions, even when your fingers are flying across the keys.

This is what I call the "translation tax" – the cognitive overhead of converting messy thoughts into structured text. No matter how fast your fingers move, you're still spending mental energy on phrasing instead of ideas.

With voice, you skip that translation entirely. You just say the thought as it comes – messy, unstructured, imperfect – and let the system handle converting it into coherent text. It's closer to how your brain actually works. Instead of spending energy on sentence construction, you spend it on the ideas themselves.

Beyond Words Per Minute

Voice AI isn't just for capturing thoughts faster. Sometimes the friction isn't in the writing at all – it's in the execution. Consider something as simple as sending a Slack message. With traditional typing, that means: clicking into Slack, picking the right workspace, finding the correct channel, drafting the message, and hitting send. With voice, that entire sequence can collapse into a single command: "Draft a playful update about the product launch and send it to the team channel."

That's why I still use voice AI regularly, despite my typing speed. For me, it's less about words per minute and more about:

  • Lowering friction between thought and expression
  • Staying in flow instead of pausing to structure text
  • Shifting cognitive load to the assistant for tasks like drafting, summarizing, or rephrasing
  • Reducing clicks and context-switching for routine actions

Does that mean I never use my keyboard? No! It’s all about choosing the right tool for the right context – voice for idea generation and delegation, keyboard for editing and precision.

When I Use the Keyboard

  • Quick Slack replies – "Thanks, shipped!" or dropping a link
  • Editing blog drafts – tightening sentences, fixing typos, polishing flow
  • Coffee shop work – when speaking aloud would annoy everyone
  • Filling out forms – name, email, address fields

When I Use Voice

  • "Draft and send a fun team update about our product launch" – let AI handle the tone and structure
  • "What's the weather in Berlin next Tuesday?" – quick info while staying focused
  • Walking brainstorms – thinking through strategy while pacing around
  • "Summarize this 47-message Slack thread" – listen to the recap instead of reading
  • Morning brain dumps – getting messy thoughts out before they're ready to be structured
  • "Set a reminder to follow up with Sarah tomorrow at 2pm" – quick task management

Keyboard = scalpel for precisionVoice = accelerant for getting started

The Hybrid Reality

In practice, it's rarely an either-or choice. My typical workflow looks like this:

  • Voice to generate or delegate – getting the raw material out of my head
  • Keyboard to refine and finalize – polishing it into something ready to ship

Voice is the accelerant that gets me started and handles the routine work. The keyboard is the precision tool that makes it shine. Together, they create a system with much less friction between thought and action.

The goal isn't to replace fast typing with voice AI – it's to use both tools strategically. Speed isn't just about how fast you can press keys. It's about how quickly you can move from idea to execution, with as little friction as possible along the way.

And that's why, even at 107 WPM, I still talk to my computer.

I use Deepgram Saga as my go-to voice AI tool – try it out here!

Unlock language AI at scale with an API call.

Get conversational intelligence with transcription and understanding on the world's best speech AI platform.