Google admits Nest security product has a secret mic, insists it wasn't supposed to be a secret

Originally published at:




the microphone has never been on…

Google pays for components and installs them, but never uses them?


Google has a lot on its mind. Some things are bound to slip. But no need to worry. It’s an “opt in” microphone. There’s no way that would ever be abused. Pinky swear.


When I was in school, we were reading, and analyzing, things like 1984, Fahrenheit 451 and Brave New World. The implicit message was, Don’t let this happen. I don’t think we all got the message.


Considering people will sue Apple because their battery doesn’t last forever, I expect a huge lawsuit over this. Or maybe people will like it… who knows? I wonder when Google will announce that there’s also a hidden camera in there?


In the novel “The Light of Other Days” scientists figured out how to use wormholes to view any event in the past from any POV. They could also view anything anywhere in the present. Soon everyone had the ability to employ this technology and the book’s final chapter was a quick exploration of all the ways the world would change as the result of all privacy bring completely stripped away. I seem to recall the authors (Arthur C. Clarke and Stephen Baxter) had a fairly optimistic opinion of the way things would be (e.g. no more corruption, no more lying, no more clothing [since why would anyone feel the need to be modest when they know they can be watched at anytime by anyone]) implying that once everyone got used to the idea their lives would much better.

I’ve never been in agreement with those conclusions but of course it was a science fiction novel and the concept was so far fetched that there was no reason to worry we’d ever have to find out if they were right.


It’ll end with it scuttling around like the head crab from The Thing.


Putting on my tin-foil top hat for a moment: It’s like there’s a deliberate, coordinated campaign to destroy all consumer trust in the privacy protections included in TOS and now hardware. Once trust reaches target levels (very low), companies can then monetize privacy features in the form of monthly subscriptions. It’s like “protection money” but the key difference is…


That’s because, as it turns out, the purpose of adding that sort of material to curricula around the country was to monitor, record and report to the authorities the reactions the kiddos had to it. No problem tho! It’s totally cool, b/c they simply forgot to mention it all those years!

Every OnStar car has a mic. Every Amazon Echo has a mic. Every smartphone has a mic.
But this is part of Google’s plot to secretly monitor us all?

1 Like

I’ve worked on hardware with unused components. The component set was locked down a year or two in advance because we thought they would enable cool features. The cool features were eventually cut, but it was too late to remove the components and still ship on time.

1 Like

I’m wearing this t-shirt right now:


FYI, the Apple Watch series 2 had a barometer, never used or advertised. It was a “new feature” on the series 3. The S2’s was on the wrong side of the water barrier, so they decided not to announce it as a “until the first splash” feature.

Other products at various companies end up with hardware because it was part of an integrated part, or it was hoped to be used but the feature didn’t make the feature freeze date, or otherwise had late software bugs.

It isn’t all that uncommon (unless you are looking at the second gen price squeeze version of hardware!)


The new mantra is
ship it in the update.

doesn’t Elmo have a mic?

Wasn’t it the case in some mid-range products that produces actively crippled the software, so that their upmarket alternative had an additional feature?
I vaguely remember Garmin doing something like that in the GPSMap series. Maybe barometric pressure was on board (SOC’ed?), but turned off, and the housing was different or something?

1 Like

IBM has done that for decades with mainframes. Buy a cheaper mainframe, you get lower performance. Decide you need to upgrade? A tech will come in, mumble a few incantations, and presto, you’ve got the next model up, which is basically unlocking the locked-out hardware.

I don’t know specifically about mid-range products.

I know Tesla has done it with some battery models (BB covered this more then ones, most recently when Tesla enabled “full range” on those cars during a hurricane). Sun Microsystems did it with their (for the era) massively multiprocessor systems, you could edit a text file in /etc that told it how many CPUs you wanted it to use, but if it didn’t match your license you got a call from your sales guy. You were explicitly allowed to do so for short periods of high load though (like a few days) without having to buy the license. DEC VAX systems, and maybe older systems did it. I don’t know if any DEC Alpha systems did. IBM did as well. Technically DEC did it with some restores on the board (“whatever you do don’t cut this trace because doing so will make your computer run 20% faster” made the rounds a few decades back…). I assume there are a ton of others I don’t know about (or knew & forgot about).

I think in general this sort of thing can be good if the NRE is the largest part of the cost, and earning back the NRE is hard at the lower price point. Adding in the features and then having multiple products that enable them or not can recover the NRE on the features, but if they couldn’t have charged for the features they never would have been developed because they couldn’t justify the NRE at the lower model price. (alternately the lower cost model wouldn’t have been made because the NRE on designing the slower one is too costly)

The hard part is you can’t really tell “from the outside” if that model is justified. You don’t really know the test/design costs. It just feels like a ripoff. (or if the license is unenforceable and first sale doctrine lets you enable the feature yourself, it feels like a secret discount!!)