I recorded 1,180 TTS clips to find out what voices misread in LLM output

Sign in to join the discussion — your thebay.events account works here.