So even their thinly-vailed claim of using the data for support and "issue diagnostics" falls flat, because they don't even provided support for the device at all. It's just data collection.
Even claiming it's for academic research datasets would be a better excuse.
Seems far more likely that they use a microservices architecture, and while they decommed the frontend that allows users to adjust their thermostats, they haven't bothered decomming the telemetry data ingestion services yet.
I still have the screenshot somewhere. About four years ago, me and my wife were at her sister's home. As I talk to her husband about the paved driveway he fixed up the day before, in front of their Nest camera doorbell thingymajig, he says the words "yeah I saw I could buy like airport grade tar, and I thought to myself, why would I need that for a car?" the next day, I pop open Instagram. Lo and behold, something I never googled, an ad for airport grade tar shows up on Instagram. I'm told that "no they're not listening" and "they use your debit card info to make educated guesses" how in God's green earth would this have happened, I have never bought any level of tar for one, nor googled it, nor ever intended on it.
Idk man, I'm a skeptic that they're not listening in some weird way. Not to mention both of my wife restrict which apps get camera and microphone access. It's uncanny that things we talk about but never google / look up wind up as ads within a day if not hours.
Another fun one was the time a friend was telling me about a niche ramen, by brand name on Discord. I pop open Facebook, what do I see? The EXACT ramen brand is the very next ad. If they aren't watching us for ad revenue I'm going to go crazy with all these insane coincidences.
I would not at all be surprised if they are listening. But isn't a simpler explanation that your in-law was googling driveway options, clicked through a link to see what the heck airport grad tar even is, and then google saw you were in the same vicinity and guessed you'd have similar interests? I wonder how many other ads you had in common that week, or if he saw ads for the underwater basket weaving course you purchased, etc.
That still raises the question of HOW because I never used their Wi-Fi neither has my wife. So Nest uniquely identified my voice, captured the conversation and told Facebook I must want airport grade tar?
>Idk man, I'm a skeptic that they're not listening in some weird way. Not to mention both of my wife restrict which apps get camera and microphone access. It's uncanny that things we talk about but never google / look up wind up as ads within a day if not hours.
Surely this is easy to test? Come up with a list of 100 topics. Of those, randomly choose 50. Work them into your conversations, and collect all the ads you've seen. Note down how ads you get for the 50 topics you've chosen compared to the topics you haven't chosen. Better yet, give your phone to your friend and have him say the ads, so you don't get confirmation bias.
This is all very easy to do, and the conspiracy that facebook/google/whatever is secretly listening to you isn't exactly fringe either. Yet, I'm not of any rigorous testing that proves it's real. While absence of evidence isn't evidence of absence, it's a good reason to be skeptical.
In 2019, I have done an experiment: talk about a conspicuous topic in presence of an unlocked Google Pixel. Minutes later there were suggestions of relevant to that topic articles in Google app.
It is not difficult to believe that if it is available to Google itself, it will be available to whomever that pays for it. That is their business model, after all.
The relevant settings are part of your Google account[0], and as of 7 months ago (which is the last time I checked) KB said:
> Web & App Activity saves your searches and activity
from other Google services in your Google Account.
You may get more personalized experiences…
> When Web & App Activity is on, you can include audio recordings from your interactions with Google Search, Assistant, and Maps as part of your activity.
Speaking in presence of a phone with relevant settings turned on certainly strikes me as something that can qualify as “interaction with Google Assistant”.
The KB article does not mention ads, but I would not be surprised if more uses of your voice activity were hidden somewhere in ToS.
>In 2019, I have done an experiment: talk about a conspicuous topic in presence of an unlocked Google Pixel. Minutes later there were suggestions of relevant to that topic articles in Google app.
That's hardly a conclusive experiment when there wasn't a control group, and sample size was 1. That's the issue with the "facebook/google/whatever is secretly listening to you" conspiracy theories: plenty of anecdotes, zero rigorous testing.
Let's also consider this theoretically from business point of view. If you were an ad network, and you could listen to what your customers talk about, would you use it or not?
If the implication is that ad networks would want to, and therefore it's happening, that's terrible logic. I'm sure the US government would love to convince everyone they landed on the moon and won the space race, without actually having to send a rocket to the moon, but that doesn't mean that means the moon landing was faked.
Part of the point of the space race was to flex on how technologically ahead we were with the implication that our weapons systems were also more advanced, so I suspect that given the choice to actually have more advanced technology vs. merely convincing everyone they did, they would prefer to have actually had that capability. Surveillance companies would presumably also prefer to actually have such capabilities while not broadcasting that fact lest people avoid them or demand they be regulated.
That's where the "randomly choose 50" comes in. Even if your initial list of 100 topics was biased, randomly picking from that list allows you to do a comparison between topics that got randomly picked vs ones that didn't. In other words, a randomized controlled trial, the gold standard in experiments. If the group that got randomly picked got a 40% hit rate (ie. that corresponded to ads), but group that didn't randomly get picked got 41%, then you can probably conclude they're not listening to you, even if a 40% hit rate seems spooky. On the other hand if there were significant differences (eg. 40% vs 20%), then there might be other stuff going on worth investigating.
As soon as Cody Kociemba open sources his firmware I'm going to apply it to my Nest 1st gen. It is ridiculous Google is turning off the support, it is probably because they don't want to support the old API or something...
I just threw my unsupported Nest in the trash, just like Google wanted, and got something else.
After all, that’s what they want me to do. Well, their first preference would be for me to buy another one.
Even if I were to put some kind of open source thing on it, I don’t want visitors to my house to see Nest hardware and for Google to get that slight amount of free advertising.
Real time pressure information is valuable when being reported from lots of geographically dispersed devices.
Its a way of tracking weather fronts.
The "discontinued" device is still a little google money printer.
So even their thinly-vailed claim of using the data for support and "issue diagnostics" falls flat, because they don't even provided support for the device at all. It's just data collection.
Even claiming it's for academic research datasets would be a better excuse.
Seems far more likely that they use a microservices architecture, and while they decommed the frontend that allows users to adjust their thermostats, they haven't bothered decomming the telemetry data ingestion services yet.
I still have the screenshot somewhere. About four years ago, me and my wife were at her sister's home. As I talk to her husband about the paved driveway he fixed up the day before, in front of their Nest camera doorbell thingymajig, he says the words "yeah I saw I could buy like airport grade tar, and I thought to myself, why would I need that for a car?" the next day, I pop open Instagram. Lo and behold, something I never googled, an ad for airport grade tar shows up on Instagram. I'm told that "no they're not listening" and "they use your debit card info to make educated guesses" how in God's green earth would this have happened, I have never bought any level of tar for one, nor googled it, nor ever intended on it.
Idk man, I'm a skeptic that they're not listening in some weird way. Not to mention both of my wife restrict which apps get camera and microphone access. It's uncanny that things we talk about but never google / look up wind up as ads within a day if not hours.
Another fun one was the time a friend was telling me about a niche ramen, by brand name on Discord. I pop open Facebook, what do I see? The EXACT ramen brand is the very next ad. If they aren't watching us for ad revenue I'm going to go crazy with all these insane coincidences.
I would not at all be surprised if they are listening. But isn't a simpler explanation that your in-law was googling driveway options, clicked through a link to see what the heck airport grad tar even is, and then google saw you were in the same vicinity and guessed you'd have similar interests? I wonder how many other ads you had in common that week, or if he saw ads for the underwater basket weaving course you purchased, etc.
That still raises the question of HOW because I never used their Wi-Fi neither has my wife. So Nest uniquely identified my voice, captured the conversation and told Facebook I must want airport grade tar?
>Idk man, I'm a skeptic that they're not listening in some weird way. Not to mention both of my wife restrict which apps get camera and microphone access. It's uncanny that things we talk about but never google / look up wind up as ads within a day if not hours.
Surely this is easy to test? Come up with a list of 100 topics. Of those, randomly choose 50. Work them into your conversations, and collect all the ads you've seen. Note down how ads you get for the 50 topics you've chosen compared to the topics you haven't chosen. Better yet, give your phone to your friend and have him say the ads, so you don't get confirmation bias.
This is all very easy to do, and the conspiracy that facebook/google/whatever is secretly listening to you isn't exactly fringe either. Yet, I'm not of any rigorous testing that proves it's real. While absence of evidence isn't evidence of absence, it's a good reason to be skeptical.
In 2019, I have done an experiment: talk about a conspicuous topic in presence of an unlocked Google Pixel. Minutes later there were suggestions of relevant to that topic articles in Google app.
It is not difficult to believe that if it is available to Google itself, it will be available to whomever that pays for it. That is their business model, after all.
The relevant settings are part of your Google account[0], and as of 7 months ago (which is the last time I checked) KB said:
> Web & App Activity saves your searches and activity from other Google services in your Google Account. You may get more personalized experiences…
> When Web & App Activity is on, you can include audio recordings from your interactions with Google Search, Assistant, and Maps as part of your activity.
Speaking in presence of a phone with relevant settings turned on certainly strikes me as something that can qualify as “interaction with Google Assistant”.
The KB article does not mention ads, but I would not be surprised if more uses of your voice activity were hidden somewhere in ToS.
[0] https://support.google.com/websearch/answer/54068?hl=en&co=G...
>In 2019, I have done an experiment: talk about a conspicuous topic in presence of an unlocked Google Pixel. Minutes later there were suggestions of relevant to that topic articles in Google app.
That's hardly a conclusive experiment when there wasn't a control group, and sample size was 1. That's the issue with the "facebook/google/whatever is secretly listening to you" conspiracy theories: plenty of anecdotes, zero rigorous testing.
Let's also consider this theoretically from business point of view. If you were an ad network, and you could listen to what your customers talk about, would you use it or not?
If the implication is that ad networks would want to, and therefore it's happening, that's terrible logic. I'm sure the US government would love to convince everyone they landed on the moon and won the space race, without actually having to send a rocket to the moon, but that doesn't mean that means the moon landing was faked.
Part of the point of the space race was to flex on how technologically ahead we were with the implication that our weapons systems were also more advanced, so I suspect that given the choice to actually have more advanced technology vs. merely convincing everyone they did, they would prefer to have actually had that capability. Surveillance companies would presumably also prefer to actually have such capabilities while not broadcasting that fact lest people avoid them or demand they be regulated.
I'm tempted to try this experiment, but I get stuck at "what if my list of 100 things is subconsciously skewed by ads I'm already seeing?"
It's the ads I'm seeing, but aren't registering on a conscious level that concern me. I think we see far more ads than we are aware of.
That's where the "randomly choose 50" comes in. Even if your initial list of 100 topics was biased, randomly picking from that list allows you to do a comparison between topics that got randomly picked vs ones that didn't. In other words, a randomized controlled trial, the gold standard in experiments. If the group that got randomly picked got a 40% hit rate (ie. that corresponded to ads), but group that didn't randomly get picked got 41%, then you can probably conclude they're not listening to you, even if a 40% hit rate seems spooky. On the other hand if there were significant differences (eg. 40% vs 20%), then there might be other stuff going on worth investigating.
I both love and am horrified by this answer. What if the airport grade tar was there all along?
The other wild thing is I was severely out of town, the ad was very local but way too uncomfortably relevant.
As soon as Cody Kociemba open sources his firmware I'm going to apply it to my Nest 1st gen. It is ridiculous Google is turning off the support, it is probably because they don't want to support the old API or something...
I just threw my unsupported Nest in the trash, just like Google wanted, and got something else.
After all, that’s what they want me to do. Well, their first preference would be for me to buy another one.
Even if I were to put some kind of open source thing on it, I don’t want visitors to my house to see Nest hardware and for Google to get that slight amount of free advertising.
Real time pressure information is valuable when being reported from lots of geographically dispersed devices. Its a way of tracking weather fronts. The "discontinued" device is still a little google money printer.