One of the most important changes in how I work over the past month: talking to my computer by voice. Let’s break down why it matters right now, why you might have missed it, and the real use cases.
Why now?
Voice input doesn’t look like anything new. The feature showed up on phones 10 years ago. Speech recognition on computers has been around a long time too. Seems like there’s no innovation here. But something changed.
Remember what voice input on your phone looked like 10 years ago? It was unusable. Easier to just send a voice message. Your voice got transcribed literally: every “uhh,” “well,” “you know” made it in. And the text came out all wrong.
It’s all down to the difference between spoken and written language. Spoken language is more redundant — you can repeat something a few times, explain it from different angles. We use more words when we talk. Which is why dictated text reads strangely on a screen.
This is exactly where LLMs came to the rescue. Voice input through tools like Superwhisper or VoicePal runs your speech through an LLM, turning your spoken language into written language.
Remote work changed the rules
Another reason voice input didn’t take off earlier: COVID and remote work. When you work in an office — especially an open space — talking to your computer out loud can feel awkward; easier to type. Now that a huge number of people work from home, it’s a different story. And by the way, even at home, talking to your computer feels odd at first — but after a few days you get used to it, and it becomes completely natural.
So, I hope I’ve convinced you to give it a try. But what for?
The real use cases that work for me
1. Capturing material by voice
I dictate all my long messages and work notes into Notion. For one thing, it’s fun. Yes, your hands reach for the keyboard out of habit — that’s normal. It’s a pattern that’s been with us for decades. But the moment you start doing it by voice, you’ll realize you think and lay out your thoughts differently.
For me, personally, it’s now much easier to dictate a meeting summary and tidy it up a little before sending than to type it out on a keyboard. You can write articles and blog posts this way too. Personally, I still find it hard to start talking from a blank page — I need some kind of outline. But once it exists, dictating the text becomes an easy and genuinely interesting task.
2. Talking to LLMs
Turns out communicating with ChatGPT, Claude, and other LLMs is also fun in voice format. Not in the sense of real-time voice chat — I mean typing your prompt by voice.
Why is it great? Because spoken language is more redundant, so you can pass along more of the extra details you might be too lazy to type out. When we type, we try to squeeze into some number of characters. When we speak, there’s no such limit. So a big prompt is much easier to dictate.
3. Journaling
For over 20 years I’ve been trying to keep a journal regularly. I keep coming back to it, then disappearing again. Sitting down at the end of the day to write a coherent text — I don’t always have the energy for that. But it turns out I always have the energy to speak that text into a microphone.
For this I use Cursor (I’ll definitely make a separate video about it). I usually record “today’s note” in several passes over the course of the day. I just say, “Add to today’s entry that I finished the article about voice input and started reading Neal Stephenson’s new book.” Cursor figures out on its own which file to put it in and how to make the right link to the book and the article. This makes journaling radically easier.
So if you’ve long wanted to keep a journal but the keyboard held you back, voice input is a good solution.
4. Learning
There’s a rule: if you want to learn something, you have to read it or watch it, and then write it down in your own words. Or tell a partner. I love this rule, but I write very little down. I don’t like reading and typing at the same time, when I could just hit Ctrl+C and Ctrl+V.
With voice input, my learning process looks like this: I read an article, watch a video, or listen to a podcast. The moment I find an important idea, I hit pause, go into my knowledge system, and create a note, dictating in my own words what I’ve just heard.
Then I read the resulting text and fix the mistakes and the structure. It turns into a system: I read something, ran it through myself, said it out loud, and looked at the result with my eyes. The information passed through several “different parts of the brain.” And most importantly — that information stayed in my knowledge system, and I can come back to it later.
If you have questions, definitely drop them in the comments. And if you’d like to share your own discoveries about how you use voice input in your daily life, please write — I’d love to hear them.