I wonder how a fine-tuned STT model running locally in the aircraft could fair vs an experienced ATC operator listening over the radio. I bet we're at the point

I wonder how a fine-tuned STT model running locally in the aircraft could fair vs an experienced ATC operator listening over the radio. I bet we're at the point where the speech model is more accurate.