New Amazon Echo devices will have local voice processing

Marketing an always-on face tracking camera with voice recognition enabled microphone as offering "privacy" is really something.

Interesting NLP tech:

> Amazon claims it’s the first company to offer this kind of privacy-first technology on smart speakers. Last year, Amazon unveiled the AZ1 Neural Edge processor, which powers current Amazon Echo devices, but its new Echo Show 15 will be powered by the AZ2 processor. Amazon said that this processor can do 22 times more TOPS (trillions of operations per second) than the previous generation.

> The company built this module alongside MediaTek, and it will allow for on-device neural speech recognition for new products.

Aaand, as I expected, it is for English only:

> Amazon says these latency benefits will be available first for English in the US, but it plans to expand to more languages “in time.”