I wish there was a way to use Whisper for continuous recognition! Unfortunately, Whisper doesn't support streaming — and AFAIK it's unlikely that it ever will.