While doing my daily set 火 listening came up but the female voice sounded like “き” not “ひ.” I even compared it to the audio for the “you selected” answer and they sounded identical. I clicked redo question and when it came up it was the AI generated male voice but it sounded like “ひ.”
Do you think there might be a future feature where card problems could be reported within the test so the exact card could be flagged for you to check out? I know it is a lot of work.
Our male voices are very "stereotypical" in different ways, and they have done a TON Of work over the years (through a crowdsourcing platform), so their voices were prime pickings for the "ai takes everything"-pocalypse.
