I flipped on this when a nickname I was calling my newborn child privately in my room also happened to be the name of a common consumer food product (which I don't buy and have never been 'in the market for' beyond it just being a generic food item) - and I was getting slammed with those ads for that product. They're listening.
Why is this comment getting so much heat for being confirmation bias? If anything, we should all rally around trying to guess the name in question. Did you name your kid Butter? Bananas? Campbell's CHUNKY Savory Pot Roast?
hahah you actually guessed it so I'll tell you - it was "Chunky man", and yes, I was slammed on social media with a relentless stream of inexplicable adds for CHUNKY soup, until I was like "holy shit"
edit: and before anyone asks, this nickname was only said verbally, in my house, among my family members, and never typed into a computer a single time
What if you had it backwards? You named your child "Chunky" after slowly being operantly conditioned to think about the name from the endless ads you were seeing before?
Ads for this common product shown to you must be because some unnamed entity was listening to you talk to your daughter.
Did I get that right?
I'm not sure I understood everything correctly, if you're sure, we'd certainly appreciate the names of any consumer products that you feel do this, as well as the consumer product.
Any tech product that keeps an audio livestream, then shares the entities with a powerful company later, just so they can show you ads, shouldn't go unnamed.
There are enough anecdotes to be more than confirmation bias
Try it for yourself. Choose a product. Maybe one of the examples here. Talk about it loud with your phone in proximity. Maybe turn off adblocker and see what happens.
I'd assume that this listening functionality only comes with certain apps, that's certainly the case for ultrasonic tracking apps [0]. Probably you have just managed to avoid installing such apps.
Sounds like you might have the problem here if you think you can determine someone is acting schizophrenic via internet comments. The person I am replying to is suffering from confirmation bias about their belief in confirmation bias.
Could you imagine the world of shit Facebook/Google/Apple/etc would be in, if someone from their engineering team proved this was possible? People have claimed this NSA-style "listening in" has been happening for years, just to target ads, but there's never been a single shred of code to explain how it would be feasible, especially via an iPhone.
Google's voice assistant can't even understand what I am saying when I am holding the phone up to my face and clearly pronouncing my question that I am asking directly. And then it spends a minute fussing while it compresses and uploads my voice to send across my crappy cell network.
I'm supposed to believe that it's listening to my every word from my pocket on a noisy road with a half-dead battery?
I think people just don't want to accept that:
1. We fall into very basic trends and interests that are very predictable based on age and gender and expressed interests
2. We do so much mindless scrolling and are not aware of all of the random content we see until we get something that triggers a memory.
If such a capability existed, it would have been made public by now. It would require the complicity of Apple, Samsung, Google et al.
There has not been a single shred of evidence produced to support yearslong claims of devices constantly listening in on you. Intelligence services spend huge money to purchase compromises from private companies, or produce them in house. If they could simply compel a tech company to utilise this supposed technology, they would obtain a warrant to do so.
All of the ‘targeted advertising’ anecdotes on here can easily be chalked up to confirmation bias.
I have a pretty solid understanding of the western IC and I can fairly confidently say that if this technology existed it would be used constantly as it supersedes a typical phone surveillance warrant.
I am a privacy advocate, but this is not one I’ve been especially concerned about.
The amount of baseless skepticism about these marketing technologies is astonishing.
It requires NO cooperation on the part of Apple, Google etc.
For smartphones it just requires allowing microphone access to some app; there are troves of apps with legitimate needs to access the microphone (voice chat in games, voice notes...); and many people accept all the permission requests they're offered, in any case.
Active microphone indicators have been introduced in Android 12 and iOS 14; many people still have older versions; those who don't might not care about them; and everything is still possible when an app with legitimate microphone usage reasons is in the foreground.
On other devices (such as smart tvs) there are often even less problems.
Most of all, there ARE well known services that use microphones to report on ads reach, and the one object of this thread only goes further by detecting arbitrary (pre-chosen) words:
And again, this cmg offering exists, and is openly advertised by a reputable company; why so many people just want to ignore its existence, is beyond me.
This is a warning about people accessing smart devices, predominantly voice enabled cameras by using compromised usernames and passwords for the purposes of Swatting.
It does not specifically make mention of the devices surreptitiously listening in on you. Anyone accessing these devices with compromised credentials would be using them in the same way the actual user would.
A year ago I would have said it's impossible for ads companies to listen to everything you say and then base ads on it, because the sheer amount of data and compute necessary wouldn't be worth it, since 99% of what you say is useless garbage to them.
But today there are some pretty good local speech recognition models. I don't know if it's happening today, but I wouldn't be surprised if ad companies start figuring out ways to do local speech processing and only send the most relevant bits for full processing in the near future.
We'll have to depend on Apple and Google to protect us.
Let's try to imagine a world where that exists, keyword detection exists as well ("ok google" etc), and keyword detection for targeted ads doesn't exist. Can you? I can't.
I can because I've worked in the space. You have to build a model for every key word that you are looking for. Those models take up space and lots of compute to train. That's why you can't set an arbitrary wake word for your Alexa/Google/Siri and you have to choose from a short list. Because those are the only models they have trained.
It would not make sense to train a model for every advertiser and then upload that to the phone. It would only make sense to capture the audio and send it to the cloud for generic speech processing. But that would also not make sense because it takes a bunch of compute to do speech processing, not to mention you'd see all the data being uploaded from your device and the cost of receiving, processing, and storing all that data.
I'm 99.9% sure that this is not happening today, but we are on the verge of the tech being good enough to do local speech processing, and then there is no bandwidth limitations, no storage issue, and the consumer pays for compute.
Only for very reliable recognition you'd need that, rough local speech recognition has existed for decades.
Here's a very basic offline app for Android: https://f-droid.org/packages/org.stypox.dicio .
It works pretty bad for me, with its tiny not specialized models, but still enough for some purposes.
You can use an online model for the confirmation of the recognitions, by the way
> It would not make sense to train a model for every advertiser and then upload that to the phone.
I was assuming that's what they were doing. Maybe you could combine both ways, with really imprecise models, so the phone captures and upload only words that above average chance of matching (so not that much data), and have heavy servers do the rest?
Yes I don't work in this field, but you shouldn't assume that just because you don't know how something can be achieved, then no one else has figured it out either (especially where they're motivated to keep it secret).
Yep, if you can think of a feasible way to do it, you know it will be done. Local filtering makes it totally feasible. The real question is whether the OS is preventing this type of passive access to the microphone.
I'm convinced this is happening. I notice it regularly. Here's one:
I dramatically extolled the virtues of "Corinthian leather" during a dnd campaign and started getting advertisements within 24 hours about Corinthian college. I don't need another degree, I graduated 10 years ago. I'm thousands of miles away from their campus, I live in a cabin in the woods.
Someone in your group searched for "Corinthian leather" on their phone because they didn't get your joke. But because they searched on your wi-fi, your general IP address is now getting targeted.
This happens all of the time in our household. I'm seeing ads for things I never talked about, but are related to my wife's searches because that's as close as they can target.
Same. I know everyone has denied it and security researchers have found no evidence as far as I know, but I'm not letting go. Here's mine:
I was at work one day and my co-worker was dribbling a basketball in the office just talking to me, and I happened to mention that I could ALMOST dunk but never actually got it in no matter how hard I tried. I open my phone and see an ad for "learn to dunk in 30 days guaranteed."
I don't care about basketball. Haven't watched or played it regularly since the 90s, never search anything related to it.
From the context, I would guess that your cabin in the woods has at least one smartphone in it. Does it have an Alexa or other device that is known to listen?
“Smart” TV manufacturers have all the wrong incentives to get 1984-style telescreens fully operational. They plug into the wall so there’s no battery drain to worry about, and the prices are so artificially low (cf non-smart displays ) that they must be doing something to make it all worthwhile.
Most people have a TV, spend a lot of their time near that TV, and probably have bought a 4K screen in the last decade to bring this hardware into their home.
I think the laser focus on smartphones (you immediately thought smartphone when you read the headline, right?) where it’s likely impossible to do wrt power consumption, ignores the other microphone almost everywhere by now
Pick a random old model car and start talking about. Just say the year and model a lot for several days and see if your Facebook feed starts suggesting random pages related to that or similar vehicles.
Of course our devices are listening to us! What use is a voice activated device that can't process speech? That this ad company is leaning into the old "facebook messenger is listening to you 24/7 because it asks for mic permissions" scare to sorta, but not really, defraud its customers (they imply, but never outright say they are listening to ambient speech at all times) is pretty funny to me.
When the page itself was shared weeks ago on social media, after looking at it I suspect it's an April Fool's page that was orphaned and never came down.
It's not linked to anywhere else from the site.
The claims are insane if serious but hilarious if satire.
Gizmodo should be ashamed of themselves. Running an article that says “there’s little to no evidence suggesting that ‘your devices are listening to you’” is obviously and provably false.
Any device that responds to voice commands must be listening. Always. This is obvious.
We also have proof that companies like Amazon do collect the data, store the data, and listen to the data [1].
They try to make this sound like a conspiratorial idea in contrast with debunked ideas. I question their motives when the ideas they call false are so clearly true.
Amazon is listening to voice recordings sent to it explicitly to process. This is much different than the device passively listening to you when it is shut off. And just by the amount of human labor Amazon puts into transcribing the clips should be proof enough at how useless this data would be for automated systems.
The idea that "wake words" equate to your voice being sent off to the internet is pretty easy to disprove. These devices will respond to the wake words even when disconnected from the internet.
Generally such assistants work by continuously listening for a 'wake word', the processing for which happens on device and without recording the audio to disk or sending it to a server. After they hear the wake word, they then start sending the recorded audio to a server to recognise the request (usually after playing some audio cue to let you know they are expecting). The reason for this is twofold: firstly, the obvious privacy concerns. Secondly the cost of streaming that much audio continuously and processing it server-side is very high.
If someone were to show evidence that any such products were in fact sending audio back without being triggered by such a wake word (either continuously, triggered by something else covertly, or without giving a cue that it was listening), it would likely be quite a scandal.
It is technically true that devices with wake words are always listening, but that's not the whole truth. In that idle mode they're only listening for the wake word. Only once you've said the wake word does it start recording. This is evidenced by the fact that you can't set a custom wake word on them (you may only select from a small list on which they were trained).
Nobody has provided evidence that they're always recording and sending (only conspiracy theories). It's possible those theories are true, but there's currently no evidence.
I dunno. My wife and I have made a game of discussing random products/services we have no interest in (like golf, pilot certification, etc.) to test this. Neither of us gets ads for golf or flying lessons.
We even tested a mind-reading hypothesis by writing the topic on a piece of paper and showing it to each other occasionally to put it in the other's consciousness. No ads for those topics manifested either.
Stolen mic data revelations in 10 years wouldn't surprise me but as of now, we cannot reproduce it.
