(Or it could be that everyone working at Google has been carefully chosen to not have such concerns; I do get that feeling sometimes too.)
The engineering team refused to take the less expensive route, and insisted that the mute button physically disconnect the circuit, so that no future engineering team could decide to stealth "unmute" the microphone through software.
To this day, you can disassemble an Amazon Echo device and you will find a physical disconnect of the mic circuitry when you push the mute button. Don't want an "always listening" smart speaker? Just keep it muted, and a red LED circle informs you that the mic is physically disconnected.
I'm proud of the approach that Amazon takes to privacy. Privacy of customer data is considered the most important thing to Amazon, and this customer obsession (the #1 leadership principle) permeates the organization.
Disclaimer: I'm a principal engineer at Amazon.
for some reason i'm thinking this point of view isn't held company wide.
That's a rather poor choice of words. I prefer "helping police departments catch criminals", myself. 'cuz, you know, police departments exist for a reason.
Police have a legitimate, important societal purpose, and have historically abused and over-surveilled minority populations in a way that's highly problematic.
There's a compelling use case for facial recognition in law enforcement. There's also a compelling case that it needs to be closely scrutinized and regulated.
They aren't always good guys, and it's ok to have legitimate concerns about cooperation with law enforcement.
Update to clarify reasons for this characterization: Parent used the words "refused" and "insisted," which strongly suggest conflict between the pro-privacy engineers and others at Amazon involved in the project. And "so that no future engineering team could decide to stealth 'unmute'" suggests a lack of trust in long-term company management. Nothing in this story supports the later statement that "Privacy of customer data is considered the most important thing to Amazon."
This type of product design decision happens all the time. Whenever you're considering component costs, you have to evaluate all of the options. You're mischaracterizing it as a fight between engineers and management.
And second, Amazon did the right thing and listened to them, when they didn't have to. They could have given the project to a different team, reassigned people, or even fired them.
Instead, they had the sense to listen to their engineers, which was the right thing to do.
Who are these engineers? Have they ever spoken publicly about this stance?
https://security.stackexchange.com/questions/154343/can-a-sp...
So while I agree with you, it _also_ wouldn't surprise me if someone raised the concern, but that person was on a different / more silo'ed team and therefore the concern never reached the execution stage. Herego too many management layers and/or nodes of entropy for communication.
But that's just a hypothesis.
And when you go looking, sure enough, there it is in the backlog down in Priority "We'll get to it when we get to it, after all these other more important things that needed to be completed by last week."
https://en.wikipedia.org/wiki/Robbins_v._Lower_Merion_School...
https://en.wikipedia.org/wiki/Robbins_v._Lower_Merion_School...
"The school district intentionally did not publicize the existence of the surveillance technology. It also actively sought to conceal it.[23][41]" https://en.wikipedia.org/wiki/Robbins_v._Lower_Merion_School...
I've worked on several products that had capabilities that we were told by attorneys could not be "advertised" (i.e., no references to them) until the complete feature was ready to be announced.
The '1st order' response of marketing doing their job is 'who cares?'. Few people care about tech tidbits that are not user oriented.
There's already tons of things to worry about and address - and every single bit of copy takes up valuable space.
The issue also does not fit into the standard communications framework: Hey, should we should tell people that there is a microphone, even if it is not working and does nothing? How do we even do that? "Hey, your alarm has a microphone!" Wouldn't that seem odd, why does it have a microphone?! You'd even have to kind of explain it: "Your product has a microphone so that one day in the future, you might enable some other features that don't exist yet"
"this has privacy implications" - no - it only has the perception of privacy implications. Because Google is not actually intervening on people's privacy, it's unlikely they really thought about the need to give people an unneeded affirmation.
Maybe they had a discussion about it, maybe it just didn't rise to the level of 'very important'.
Only with a very specific concern for a subgroup of customers who are wary of these things, would someone have enough leverage to get that "Hey, there's a microphone that does nothing!" notice on the box.
Google is not doing specific evil. They are not trying to infiltrate your homes and take nude pictures of you so they can look at them or sell them.
They are systematically evil, in the sense that it makes sense for them to sell you voice/video features that you want in order that they might provide you even better services. And their AI will use nudies of you in a backwards way to learn more about you.
They're getting evil due to their scope of influence, and negative externalities, much like FB has problems with 'Russian interference in elections' - i.e. not a problem they are trying to create, not a problem they want, just a sensitive byproduct of their product and massive success.
I've got to disagree here: it has privacy implications because at Google's end they can issue a software update and now be monitoring all audio, and the T&C no doubt say they can be unilaterally varied without notification. That all means that Google think I accept Google snooping on me; of course I don't but in court Google lawyers would say I did, and them being in a position to do that is important.
Also, if Google can enable the mic at some point then it's likely that's a possibility for a third party (crackers), OR that Google could do it in response to a legal demand from a government.
There seems no good reason to me that the specification summary can't say "microphone - not in use yet, reserved for future applications"; with a sentence somewhere expanding on that explaining they want the ability to improve the device later and do shipped a mic because it could be useful to expand the products capabilities.
Some of us actually do read instruction books; it would no doubt get some column-inches in a positive way ("what changes might Google make").
Google know everything they make is going to get a tear-down and that mics are going to be discovered in short-shrift: it's ignorant to not anticipate that. In fact one really has to assume they knew this "issue" would come up.
If/when Google enables this tech, it think it will be clear enough. If not, then definitely would be time to raise a real fuss.
"There seems no good reason to me that the specification summary can't say" - I agree with you there - however, I suggest that they mighn't have even have thought of it. That would be the prudent thing to do, but again, there's no established procedure for it.
There probably was not even a legal review hint etc. because as I say, they were not using it.
So it's questionable, weird, but not nefarious.
What they will do, with your ostensible permission, with a signed T&C - is to me, far more nefarious.
I can see this slipping through the cracks between different job roles' responsibilities, although after this incident they'd probably go through a post mortem and find a way to incorporate new checks into the product launch process.
Not a good look for Nest...