AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
How to Reduce Translation Quality in Hyper Localized Translations
A lot of human linguist translation was applied on top of machine translation and so the translation quality was generally very high. I think the signal I was looking for was whether it needed to be that high in order to solve the user need because if you're doing your machine translation it's pretty much instant. If you add linguists to try and augment the quality of the translation it then takes a matter of days You have project managers coordinating what language needs to go to what linguist before it's checked back into the repository. On the question is what's that worth. And so for me I think the important signal to look for when you're trying to reduce translation quality is to understand one what