“I think that Alexa was listening to something and sent me a message that, um, you probably didn’t want me to hear.”
Amazon: “Alexa is only looking for the wake word and then, ‘Alexa, send a message to,’ and their name.”
I do believe that’s honestly the intent; I’ve also seen continuing examples of absurd results from speech recognition. Even not considering machines, people mis-hear what others say, but you can at least hope they have some judgment about what might be a mistake contextually. It’s hard to expect that kind of judgment from a machine.
Imported from Google+ — content and formatting may not be reliable