Hacker News new | ask | show | jobs
by raw_anon_1111 201 days ago
We need to separate three separate issues. I work with call centers and I always need to discuss all three

1. Voice to text transcriptions

2. Text to understanding

3. Adding capabilities where it can do something with #2.

The voice to text that Siri uses seems to be worse than when you are dictating using voice to text from the keyboard.

The latter gets close to 100% with my southern native English accent and does okay when I’m trying to speak Spanish. Siri messes up with English a lot more and it’s a lost cause when I try to speak Spanish.

2 comments

I have a British accent and the speech-to-text from the keyboard is also terrible.

Honestly with these assistants I'd rather just type my query. Voice input is embarrassing and error-prone. The only place that voice input is useful is in the kitchen.

Counter anecdote: I also have a British accent, and while I find Siri as shit as everyone else in this thread the dictation built into the iOS keyboard very rarely has a problem with my accent. I'm fairly close to Received Pronunciation, which I'd guess is one of the easier British accents for Siri to understand.

(I do often get frustrated with dictation quirks that don't have anything to do with my accent, like it choosing the obviously wrong option when their are multiple words that sound the same, especially its insistence on assuming I'm saying the name of a contact rather than the common noun that sounds the same.)

When they switched to "ML" (this was way before ChatGPT ate the world) Siri's speech to text went to absolute shit.

I suspect they had a very carefully hand-crafted model before, and replaced it with an ML model "that will fix itself over time" - and it never did.