The way that all this “AI” processing has been trained, it almost always fails for anyone who doesn’t fit the white middle-class aesthetic. Voice-to-text generative AI processing will screw up for people with accents, including non-native speakers; also someone who slurs their words, or talks in African-American Vernacular English. Also, it requires someone to know how to speak and listen in a language. Clicking on icons and inputting commands is the same regardless of what language you speak. This just reeks of out-of touch nepo-baby executives.