I have been waiting for some kind of audio fingerprinting for security systems for years. Though personally I'd prefer a training model which would also let you

I have been waiting for some kind of audio fingerprinting for security systems for years. Though personally I'd prefer a training model which would also let you teach the device other sounds (like "this beeping means the stove is on"), I'd rather be able to set up notifications for those, but even having my phone tell me "I heard breaking glass at home" is a start. Otherwise I don't see much here that is new, presumably there's still no voice recognition on the non-wake-word speech (it's just noticing there are people around, talking to each other, or that you talk to yourself, your pet, etc).