When we discussed this - at long length - during the original Echo development program, we consistently came to the conclusion that it should be hardware, not software, to maintain customer trust.

IMHO, This was the correct choice.

No one I know would trust a software mute.

Yeah... I thought it was pretty clear. While still hoping that we could build something (product & reputation) that users wouldn't want to turn off.

Are you still involved? If so, is there some way I can request a feature to turn off all of the "by the way" and other follow-ups? I find them incredibly annoying. I usually answer "no and don't ask again" but it doesn't seem to understand.

I left some years ago. Yeah, not a huge fan of these "by the ways" but agree that having them would be less bad if there were a "don't ask again."

My question for you is why not advertise this loudly? Or, maybe you did and no one cared? Because presumably it's a bit more costly per unit to do it this way, and very few people are ever going to understand and care.

Not that I think you made the wrong choice, I think it's a good one, but I'm just curious how many consumers you thought would actually understand it, or how that information would be distributed

It's a balancing act. We built it, to build trust. But talking about it too much would - we thought - highlight concerns and increase worries that we hoped (most) users wouldn't have.

As an example, we were very clear and upfront in the original product detail page that users' voices went to the cloud, as we were not trying to hide that. Even so, some people tried to play a bit of "gotcha" and fear-mongering about the fact, as though we were not being transparent.

Appreciate that, a lot.

Particularly with software auto-updates, a software mute is simply untrustworthy.

True. And in the early (internal) alpha days, false positives for wake word detection were quite high. So we felt the pain from eating our own dogwood; even though we spent huge amount of science / engineering resources on reducing those false positives, we knew it would never get to zero while maintaining acceptable true positive rates.

I know you. :)

#doppler4lyfe

Took me a couple of minutes to figure out who you are :)

Still proud of this product!

This is a good point, I had suspected that the mute function was, in fact, implemented in software.

That they went ahead and implemented in hardware is encouraging. Someone over there took the feature seriously enough to take it this far. I am sure it added cost over a software-only solution.

Software based mute means a compromise of the kernel would destroy the control provided by a software mute.

They could have, but I believe in the past they have claimed it to be hardware.