Voice to Text (with AI)

In addition to using AI in Obsidian, I’m also allergic to voice-to-text.

E.g., if I do it in a car or so, I mostly need to spend time fixing or updating words that it wrongly interprets, or I also ramble much more when I talk than when I actually type. Again, writing is thinking, so if you skip that, you also skip the thinking.

But as you said, for notes or a brain dump, that is feasible.

As written in Keep AI Out of Your Obsidian Vault, voice-to-text can be another approach to use with Obsidian, which makes sense when you are in a car or so.

But because I talk much more than I write, and because AI or translator services don’t always understand all my words, it’s much more work for me than just typing it out. Plus, I usually skip the thinking, though sometimes talking it through can add to the process.

I never started voice to text it because, for me, it was always about skipping the thinking when I just talk. And even more, I had to check whether the transcriber was correct or had misunderstood me, so in the end, more work and less thinking. The only time I use it is in the car.

But Slow Living, Slow Technology, or as Daniel Kahneman said: Thinking, fast and slow - there are two systems, depending on what you need to do, especially creative things, the second, slow system is much better.

So for me, Slowing Down and focus on the thining with typing is more rewarding and useful than quickly talk something into a microphone.

Origin: Li, and Using Obsidian with AI
References: Audible