As we all know(or suppose) that Google is silently listening to us, is there any way to stop him without covering the microphone physically (I know it can be a suggestion)?

Like some trustable app to take the mic permissions so that Google will not be able to use that resource, or something else.

Grab yourself one of these bad boys

Yes

Google isn't silently listening to us

@nathan815
Exactly what a google-employee would say.

@metamourge I'd love to be a google employee! But not smart enough :D

@amitgupta
Ungoogle your phone, e.g. install a custom-rom without google-apps.
Overwriting sensor-data with random-data, using something like xposed-framework + xprivacy-lua might also work, but I'm not quite sure.
Since Gapps are system-apps, they have permissions to overwrite pretty much all other measures.

@metamourge oh, I don't use android. But, sure, if you think they're actually constantly streaming voice data I'd be interested in seeing the evidence for this (network activity perhaps; I'm sure this would be a huge data hog).

Yeah, google isn’t doing anything immoral, questionable or wrong.

Just like how a “bug” accidentally grabbed the WiFi traffic google maps cars passed by.

Maybe there will be another “bug” which will randomly start to record mix and video to make sure the sensors work ...

What do I know, let me put ALL my trust in a company which makes money from advertising and I am the product ...

@rusty-hacker We're talking about the claim that Google constantly streams voice data here. Nobody said everything Google does is moral.

Google claimed that there were 2 billion active Android devices in 2017. Assuming a very low quality sound with 8kB/s this would make about 15 PetaBytes/s traffic, not to mention the computing power needed to actually run a voice recognition software that can handle 15 PetaBytes/s.

Why do you guys think that they're passing the voice data and not the transcribed text? Which your phone or any device can locally do?

@neriald I don't use voice recognition on my phone, so I have no clue how good it actually is. Sending the transcribed text would only make sense if the phone can distinguish different speakers reliably, even when they are speaking simultaneously (imagine you having a conversation while a movie is running in the background), in different languages and dialects. If that can't be done by the phone's software a more sophisticated voice recognition software would have to be run, which implies somewhat that the voice data has to be transmitted.

@TobiSGD I use siri and speech to text all the time and 99% of the time main speaker is separated without any problems, also when you're running such big scale monitoring you don't care few missing words.

And it is not so hard to place a dictionary with keywords which would go into higher surveillance mode once there is a hit...

As you can see in a minute I could come up with various ideas on my own, now apply that to billion dollar company forced by governments to implement that.

@neriald OK, so it may even be possible.
Now, of course, that it may be possible does not imply that it actually happens.

@neriald I doubt constant text-to-speech processing would be feasible with mobile battery life constraints. While charging? Sure.

Now the question is, where is the evidence for this? Even if it is possible, that doesn't imply this actually happens as TobiSGD said.

@nathan815 true it would drain battery or maybe it already does.

I don’t really have proof if I had we wouldn’t need to talk about it. However there are sources like wikileaks which tells about hacking tools/backdoors for your phone, pc, smart tv.