We use different APIs to transcribe audible and readable notes, all computer based. Depending on for instance background noise the result may have errors or be empty. It also may lead to errors in transcription that could be funny or upsetting.
We apologize for that, but at least you have your Braintoss reminder :)