If only their machine learning speech to text system was good, that would be alright. But it's extremely inaccurate.