Based on a recent demonstration given by Google, the technology behind the live transcription feature may require a constant internet connection when it officially becomes available. It also says that the new feature will be different from current AI-driven and on-device translation features. As real-time “multilingual transcription” is much more complicated than simply translating written text from one language into another. Further, in its current state, Google says that the feature will not work with an audio file, for now. For the live transcription to be effective, users will have to use live audio capture through a smartphone’s microphone. Oddly enough, Google did suggest a loophole to the issue, whereby a user can simply play an audio file through a speaker and proceed to capture said recording live and get it translated.
The real-time transcripion feature is still in its testing phase, and as exciting as it sounds, Google also hasn’t given a specific availability date for it either. (Source: The Verge)
