Google is listening, always

[Koolau]

You are not paranoid.



FB is listening too.

Did you forget a letter?

 

[marko]

So here's a spooky one, albeit not really about Google 'listening' on the pod but more about 'knowing more about us than we suspect':

Yes, we all seem to have stories about ads popping up for a product we have searched.

Last week, I was searching on VRBOs for places in Europe. Predictably, ads starting showing up for VRBO. No surprise.

Here's the surprise: The ads for VRBO had photos of a very familiar property! The property is owned my best friend in Carmel. Yes, his property is on VRBO but I did not search anywhere near Carmel. I was searching in Europe and my friend is 3000 miles away from me here in Boston.

My guess is that somehow Google knows this property is owned by my friend, knows he is in my contact list and put the two together. (Obviously, not smart enough to know that if I want to stay in Carmel, I can do it for free by staying with my buddy)

 

[Winemaker]

I have to throw out a disclaimer to DW once a week at least. I constantly get advertisements for women's lingerie, but my wife buys her own in person in a retail store, doesn't use my laptop, and we have different email addresses. And we certainly don't talk about underwear, and we leave our cells in a different room/floor at night.

 

[ERD50]

I have to throw out a disclaimer to DW once a week at least. I constantly get advertisements for women's lingerie, but my wife buys her own in person in a retail store, doesn't use my laptop, and we have different email addresses. And we certainly don't talk about underwear, and we leave our cells in a different room/floor at night.

There is a simple explanation.

-ERD50

 

[Winemaker]

There is a simple explanation.

-ERD50

Which is?

I have Adblock, which also blocks pop ups. I only get advertisements on one website, which is an ecomonic/politicals site, which I allow because I don't pay for a subscription. But I get bra and womens sleep wear advertisements.

 

[CaptTom]

A buddy (older gentleman) asked me for help the other day deleting an unwanted icon from his phone. Apparently he had "somehow" managed to get a link to a p*rn site on it and didn't want his wife to see it, or to ask his grandchildren how to delete it.

Back to the original subject, as I understand it the phone is always listening for the "wake" phrase, and only sends the ensuing question to the server. Think about it. If it were continually streaming all the sounds around you, your data usage would be sky-high. I'm not really worried about that, although it's funny when the phone hears something which sounds like the wake phrase, and tries to answer whatever you said next.

 

[Winemaker]

No porn visits here, but today's ads on the site were for mechanical actuators and pizza ovens.

 

[FLSUnFIRE]

I really don't think there is listening for ads as much as really good data mining of other available info. I get ads for weird stuff after people visit if they use my wifi and thus are affiliated to my IP. Walmart.com shows me purchases made in store going back quite a ways for any credit cards that I've used online and thus associated with my identity/account. While I protect myself some I've given up on anonymity and in general find it convenient... it's nice to have most ads I suffer be somewhat relevant. (as I type this, a "Depends" ad is in the sidebar.... I hope they are not that good at predicting my future needs!)

 

[steveark]

A few weeks ago I was talking with my son and he mentioned an electronic coffee mug he had (Ember).

I knew he had it, but never knew the name. My PC was on when we talked. I have never searched for it (nor has DW).

The last 2 weeks I have been getting adds for this on all sights that I do not ADBlock.

Since we did a lot of Zoom during Covid, I realized that I left the microphone on (it is off now).

I might be paranoid, but just because I am paranoid doesn't mean they aren't out to get me

I'm afraid you are paranoid, Google never has nor never will use anything picked up on a microphone to send you ads.

 

[Koolau]

I really don't think there is listening for ads as much as really good data mining of other available info. I get ads for weird stuff after people visit if they use my wifi and thus are affiliated to my IP. Walmart.com shows me purchases made in store going back quite a ways for any credit cards that I've used online and thus associated with my identity/account. While I protect myself some I've given up on anonymity and in general find it convenient... it's nice to have most ads I suffer be somewhat relevant. (as I type this, a "Depends" ad is in the sidebar.... I hope they are not that good at predicting my future needs!)

Heh, heh, if there is longevity in your family... If we live long enough, we'll all need 'em. They may just be playing the odds by sending you preemptive ads.

 

[CardsFan]

I'm afraid you are paranoid, Google never has nor never will use anything picked up on a microphone to send you ads.

OP here. You might be right, but you sound awful certain. How can you possibly know that for sure?

 

[Koolau]

I'm afraid you are paranoid, Google never has nor never will use anything picked up on a microphone to send you ads.

Oh, my!

My current (and previous) tag lines come flooding back.

In case you forget, I've occasionally used "Just because you're paranoid doesn't mean they aren't out to get you." My current tag line is self explanatory and there are no exceptions to it that I'm aware of (at least if we all live long enough.) But if you have proof/data/company statements/preferably legal opinions to offer, I'd love to take a look as I was wrong once and of course, YMMV.

 

[ARW44]

Several years ago, after our geezer Wednesday bike ride, our discussions turned to the Citroen DS19. The trip home in the cr is about 30 miles, and a DS19 passed me on the way. The last time I saw one of those was probably 1973.

 

[catii]

No ads, sporadic links, or other surprises for us.

AdBlock Ultimate
Ghostery
Privacy Badger
uBlock Origin

My Alexa sits right in the middle of my diningroom table, which happens to be a few feet from my parrot's cage. She's an African Gray, & the damn bird talks constantly, even in her sleep! Think of it as the permanent equivalent of white noise, but using our voices.

So, although we sometimes have to raise our voices to get Alexa to respond, it's well worth not being "bugged!"

 

[NW-Bound]

What impresses me about the Google Home devices is that they can still hear me while they are playing music at a fairly high level.

You would think their own speaker would deafen their microphone. Imagine yourself singing while trying to listen to someone else talking.

 

[Winemaker]

I'm afraid you are paranoid, Google never has nor never will use anything picked up on a microphone to send you ads.

I would think Google would be listening in order to collect data for AI systems. I can remember reading about a robot taking pizza orders on the phone and sounding like either a typical teenager working at Pizza Hut or Luigi, the owner of Luigi's Wood Fire Pies. It was amazing to listen to how AI had nailed the interaction.

 

[ERD50]

What impresses me about the Google Home devices is that they can still hear me while they are playing music at a fairly high level.

You would think their own speaker would deafen their microphone. Imagine yourself singing while trying to listen to someone else talking.

Since the device is also generating the sound, it should be easy to cancel it out to a high degree with an inverse phase.

Your voice would then stand out as a difference signal. edit/add: Saturating the mic could be an issue, with the music being generated so close to the mic, but that can be accommodated in the design. They probably also use passive noise cancellation - apply the music signal to both the front and back of the mic element. That provides a pretty high degree of cancellation.

-ERD50

 

[FIREarly]

Ok, google ain't going away. But they need to step up their game. I find that my ads start coming in AFTER I buy something. Come on g-man, you can do better than that! Once I've bought whatever I'm on to the next thing. Catch up!

hahaha

 

[steveark]

OP here. You might be right, but you sound awful certain. How can you possibly know that for sure?

A skeptic that debunks all sorts of stuff did a whole show on it including information provided by Google and phone manufacturers. He even did some field tests and it sounded like very reliable information. Google would have to be flat out lying and they are smarter than that. Skeptoid.com is his web site. He's very fact based.

 

[Koolau]

A skeptic that debunks all sorts of stuff did a whole show on it including information provided by Google and phone manufacturers. He even did some field tests and it sounded like very reliable information. Google would have to be flat out lying and they are smarter than that. Skeptoid.com is his web site. He's very fact based.

Heh, heh, I feel so much better - until I look at my tag line yet again. I'm sure Google would be totally honest with him. And, of course, he would be honest on his web site.

Let me just ask: Is it possible for Google to misuse your voice inputs? If the answer is yes, I don't want one in my house. Call me paranoid all you want but, just because you're paranoid.........

 

[NW-Bound]

Since the device is also generating the sound, it should be easy to cancel it out to a high degree with an inverse phase.

Your voice would then stand out as a difference signal. edit/add: Saturating the mic could be an issue, with the music being generated so close to the mic, but that can be accommodated in the design. They probably also use passive noise cancellation - apply the music signal to both the front and back of the mic element. That provides a pretty high degree of cancellation.

-ERD50

The problem is that the signal received by the microphone is not identical to the signal sent to the speaker.

You have distortion and frequency response of the speaker and microphone to contend with. You have acoustic coupling as well as mechanical coupling, reverberation and a whole lot of hairy stuff. It's not that easily cancelled out.


PS. All this makes me curious. It should not be hard to set up an experiment using nothing more than a PC hardware, a small speaker and a microphone. You can use the PC hardware to send out a sound segment, and use the PC to also sample the microphone to see how close to the original signal you get back.

 

[Sunset]

A skeptic that debunks all sorts of stuff did a whole show on it including information provided by Google and phone manufacturers. He even did some field tests and it sounded like very reliable information. Google would have to be flat out lying and they are smarter than that. Skeptoid.com is his web site. He's very fact based.

Whew.... I feel better now, as I know companies never lie

Wonder which company the FBI use to search through certain users communications ?

"Carnivore, later renamed DCS1000, was a system implemented by the Federal Bureau of Investigation (FBI) that was designed to monitor email and electronic communications. It used a customizable packet sniffer that could monitor all of a target user's Internet traffic. Carnivore was implemented in October 1997. By 2005 it had been replaced with improved commercial software."
https://en.wikipedia.org/wiki/Carnivore_(software)

 

[old medic]

The other day a retired friend called... not uncommon for us to chat a while about different stuff... He starts reading me parts of an article about the 10 top cars that may become collectors in the future... I'm on my phone, no computer or Alexa type thing...
Now every time I open phone browser... I get links to " 10 top auto" this or that.
Same guy...we were camping, phone in my pocket and he drags out and goes to talking about their "name brand" portable smokeless fire pit.
Got ads for this "name brand" pit for the next several weeks...

 

[folivier]

Your friend is obviously an internet influencer.

 

[ERD50]

The problem is that the signal received by the microphone is not identical to the signal sent to the speaker.

You have distortion and frequency response of the speaker and microphone to contend with. You have acoustic coupling as well as mechanical coupling, reverberation and a whole lot of hairy stuff. It's not that easily cancelled out.


PS. All this makes me curious. It should not be hard to set up an experiment using nothing more than a PC hardware, a small speaker and a microphone. You can use the PC hardware to send out a sound segment, and use the PC to also sample the microphone to see how close to the original signal you get back.

Yes, but the response of the mic/speaker is easily modeled to a fairly accurate degree. That coupled with the likely passive noise canceling they use is probably good enough.

And if not, the active method you describe can be done. My surround sound theater receiver has that function. They provide a microphone on a long cord, plug that in, set it at your listening spot with all the speakers positioned, and it runs a series of test tones, measuring delays and response, and provides an offset for it all. A simpler version of this could be done pretty easily,

-ERD50