The Alexa devices have always bothered me. Not just because of the baseline privacy concerns, but because it seems so damn easy to activate the trigger word on accident. The absolute worst part being a false positive activation of my Sonos system while watching a movie/show. When this happens (and it happens a lot), the audio volume of the show is lowered significantly to allow Alexa to hear the rest of the command.

Pause the show, wait til "she" stops listening, rewind the show very slightly because I missed whatever was just said due to the volume shift, resume the video, instantly pause again at the moment she was originally triggered until she stops listening, resume the show, this time without having to rewind since I paused soon enough not to miss any dialog.

Maddening.

And the fact that all of these false positives get sent straight to Amazon doesn't make me feel great.

I didn't need a HomePod for its audio quality since I already have great quality with the Sonos system, and couldn't justify the original HomePod's price as a command/control-focused smart speaker with a home already filled with Alexa devices.

All of that to say, HomePod mini is perfect for me.

I can throw out my Echos, disable Alexa on the Sonos system, and scatter a couple of Minis to replace the Echos.

This is something I didn't realize I wanted.

South Park, being; well, South Park - actually did an episode taking advantage of this and trolling thousands of viewers:

https://mashable.com/2017/09/14/south-park-alexa-google-home...

(not really SFW content, strong language)

I suspect that the device is constantly recording the audio in your room, and sending that to Amazon’s servers in the cloud. Because I doubt that the little puck has enough processing power inside it to handle speech recognition.

This allows Amazon the ability to continually upgrade their servers, without the need to upgrade the puck.

And only when the server detects it’s prompt, “Alexa”, will it send a response back to the puck. This is how the little device appears to be so responsive.

And not surprisingly, this bit of functionality is not marketed by Amazon. After all, who the hell in their right mind, would want to be constantly monitored, and have all their private conversations at home be secretly recorded by Amazon.

In fact, I suspect that all the voice assistant devices from other manufacturers, like Google, are doing exactly this.

I have no issues personally with using cloud services like this but they very much used weasley language in the presentation and website to obscure the fact that Apple's cloud absolutely will get access to your unencrypted raw utterances following their wakeword detection, and cause people to conclude that that wasn't happening.

As a side-note I find it a bit surprising that with all the history Apple has with Siri and the meaningfully powerful on-device ML acceleration that they still can't download and run a recognition model on-device that wouldn't require streaming your speech to Apple's servers...

On device you can only ask it to play songs on the device, calling people in the phone book etc.

One thing I've learned is the identity is absolutely everything in (mass) surveillance.

The more places we can disconnect or obscure this as end users (or ideally consumers) the better.