I've been keeping a running gag with my acquaintances over YouTube's recommendations for a while now, because they're so improbably and time-sensitively on the nose.
Yesterday was the worst one yet. We were driving and my daughter was filling out MadLibs by hand in a paper booklet in the backseat. One of the fill-ins we came up with was "pantyliner".
As soon as we got home from the event, I sat down and YouTube gave me an above-the-fold recommendation for a Japanese pantyliner commercial.
I've seen enough. This word never comes up in day-to-day conversation, it's not in my interests, and it's one of the least exciting topics imaginable. None of me, my wife, nor my daughter had hands free to search it up at that time, and besides, we were totally preoccupied with driving and our event until I got home a couple hours later.
For the record, we're an iPhone family and were in a Tesla.
I had been convinced by, I think, a Simon Willison article that this isn't happening. But he's wrong. This is happening.
I wonder if the sequence of online events that lead you to “organically” come up with pantyliner was also enough for the ad algorithms to do the same.
Maybe subconsciously in videos or other browsing activity pantyliner adjacent things were discussed.
Not saying companies aren’t using audio for ads, but it just feels like there would be more leaks from the big COs if it were ubiquitous in everyday life
I also think there’s something weird going on with YouTube recommendations. The thing that I’ve noticed is that people I follow on TikTok are showing up in my YouTube recommendations, and I’ve never interacted with their YouTube accounts. Most times I didn’t even know they had a YouTube account.
I use the same email address on both accounts, but it seems weird that TT would export its follow graph to YT for any reason.
How is that surprising? Maybe you just get recommended the same stuff on multiple platforms because it reflects your interests? If I am interested in Retro computing and get the 8-Bit Guy in my recommendations on multiple platforms, would you conclude that the platforms colluded? Occams Razor tells me what’s the more likely conclusion
Are Google's cookies / ads / analytics used by TikTok? Does TikTok send notifications to your email address (is it Gmail?)? Do the TikTok accounts embed YouTube / related links in their profile anywhere? Quite easy for Google to get tracking info. So many ways that metadata gets leaked.
I suspect tech companies spy us in ways they don't disclose, but
anecdotal evidence like this is not useful. The most likely culprit is cognitive bias: if you hadn't heard that word, you wouldn't have noticed the recommendation. [1]
I do believe in this. I had this experience last year where I suddenly started thinking of a pillow product I had first seen 5 years before (and had never looked into since then).
In the following days, I started noticing the ads for that product everywhere -- in the physical world! Not online, not in my browsing, but in shops and buses and street signs.
Got me wondering which came first, the ads, or me thinking of the product? Probably the ads....
His family is surrounded with microphones, cameras and sensors. The company has a long track record of unwarranted surveillance.
In the most recent case against Google for wiretapping, all claims survived the dismissal and summary judgment stages. Google settled instead of showing a jury that what the company does is not wiretapping. Why would anyone trust this company, except out of necessity.
Anecdotes are not proof, but are anecdotes needed to form a reasonable suspicion of Google conducting unwarranted surveillance.
For some reason, HN commenters will oppose the notion of Google/Meta using microphones for data collection but how much does that matter when we already know Google uses any available means it can get away with.
I've seen similar things ... except that the internet-connected device in the situations where I become aware of this does not have a microphone that could be used by software. So I suspect that the cognitive bias explanation mentioned by others is more likely.
It is happening, and has been for a long time. When my first daughter was a baby, so around 2015, my wife and I were discussing options for helping her prop herself up, and came up with the idea (on our own, as far as I know) that a horseshoe shaped pillow would be perfect. I picked up my phone, opened the Amazon app, and typed the letter "h". The app autocompleted "horseshoe shaped pillow" based on just the one letter.
These are common baby items. If you searched for baby stuff on Amazon or anyone in your household sharing your IP did, and especially if you purchased baby stuff on Amazon, they're going to recommend baby stuff more often. It's called an a priori algorithm. It looks at what other people purchased that have purchased similar things as your search/purchase history and recommends based on that.
A long time ago, my friend bought a Hyundai Getz. We used to play the Getz game - nothing complex, but whenever you saw a Getz you'd yell "Getz" and get a point.
After playing for a few days, we were seeing easily 20-30 per day. I'm sure I'd seen them many times before (they were one of the most popular cars in Australia for a few years before that), but I'd never really noticed them. I've played the game with a few other car models with friends since, and it always results in a comment like "Wow, there are so many X on the road".
I understand why your story feels chilling to you and how it would convince you, however it's more likely that:
a) in the 100+ ads you see per day[0] you've seen panty liners before but it was never relevant to you (like a Getz) or
b) you've seen ads for panty liners recently but not conciously taken it in, resulting in the inverse of what you thought happened - your MadLibs suggestion was due to an ad.
I'm very confident many companies would be happy to spy on you via microphone if practical, but outside of malware I'm yet to see anyone proving this actually happening. We have billions of smartphones in the world and millions of tech people, many who would have to be keeping the secret all these years and many others who somehow couldn't find evidence of this. Occum's razor just doesn't add up here.
Haha... I'm surprised to learn my wife and I aren't the only people playing the "Getz" game! We started around 15-20 years ago, when every second hire car in Queensland was a Getz.. they're a little harder to come by now, but I still get a little thrill from calling one before she's seen it :)
I'm worried about falling prey to conspiratorial thought, but I'm in the same boat. I can't think of any other reasonable explanation that would explain the advertising.
It's happening and we are being gaslit by people saying oh it would never happen. Apple and Google would never allow it to happen. They're both capitalist corporations driven by profit alone. There's immense profit in this. It's happening and it makes me want to throw my phone in the river.
There are a few acquaintance over the years that have been involved in various sectors of social media and various 3 letter government agencies. The one thing they have always said is, just assume they doing to most morally questionable things to get data - they just won't public admit until it is absolutely vital to do so.
Slightly off topic but I did like one of them being so frank about Linux and open source. They worked for a while with the NSA, all they said was "Linux has 100 million lines of code. You have to fooling your self to think we didn't slip in hundreds of backdoors onto that thing". The same can be said of almost any other system though.
I have long suspected that the gaslighting thing of "it is just subconscious coincidence" is just a very neat cover for these things. There might be a slight influence but it is nowhere near as powerful an effect as they would like you to think. With advertising swaying people, they can move a few percent of people a few percent in one direction but it doesn't really target individuals effectively in guaranteeing results. Only as an aggregate.
> They're both capitalist corporations driven by profit alone. There's immense profit in this.
This is exactly why I hate this dumbass conspiracy theory. You're totally correct, they are both capitalist corporations driven by profit alone. And the fact is that if they were secretly recording us in direct opposition to what they have said on the record it would be an absolute disaster for their bottom line.
I've tested this before, and I'd recommend you all do it too. Its very easy: Pick a product you never need/buy/see advertised/talk about. Personally, I live in an apartment in a big city, so I used the word "mulch". Now stand in your empty apartment and talk to yourself about that product. This can be mostly meaningless babble like "I love mulch. I need to buy mulch. Mulch is best when brown." I did this for ~30 minutes.
Now the hard part: DO NOT tell anyone your word, do not type it into anything, its best you don't even mention the experiment to people. If you tell your word to people then they might type it into a computer which would taint the experiment.
Have you tried a control where you instead think of a product and see if you get ads for it, because what you're describing can also just be confirmation bias.
Yep, did the same thing with "Smoked Salmon". I went a step further and decided on this without speaking it about 2 months in advance so that I can get a clear baseline on ads I was getting.
On talking lovingly about smoked salmon, it took 2 days for it to turn up.
I have also seen it happen with folks talking about Jeep for instance.
I don't use any sort of voice-based smart assistant like Google Home / Alexa.
Since this experiment, I've been a lot more conscious of microphones. My desktop's microphone and my laptop's microphone both have hardware kill switches. My desktop's webcam no longer has a built-in microphone. My television and cellphone still have microphones without kill switches.
I analysed[0] many apps on the App Store when Apple started forcing app authors to disclose their use of data, at the time nine apps collected audio in a context that could be construed to mean use for ad targeting.
It’s been four years so I should really redo this research.
Very curious how zero investigative journalists have managed to go undercover and buy one of these ads, despite the fact that they are apparently so widely available.
When an ad is purchased, the buyer doesn't know how it's targeted, they only know if it was effective. The ad seller would necessarily keep the "proprietary details" of how the ad was targeted from the buyer.
Nobody's even been able to offer evidence that these apps are somehow bypassing iOS and Android permission systems to listen in on the microphone in the background without the OS itself knowing.
It's all just really dumb fearmongering that ultimately hurts the credibility of privacy advocates as a whole.
The sad thing is, these companies are collecting mass amounts of sensitive data and using it to drive ads, and we should be doing more about it. There's absolutely no reason to fabricate nonsense like this, the reality is already terrifying enough to creep people out.
> Nobody's even been able to offer evidence that these apps are somehow bypassing iOS and Android permission systems
Maybe a lot of the folks this happens to haven't denied microphone permission? Or maybe someone around them hasn't? Facebook has a calling feature, chances are people (or others around them) have given it access without realizing it. I have certainly caught Facebook Messenger snooping on my location when it had no business to, merely because I had granted the permission days ago when sharing my location with my friend, and forgotten to deny it again. It spooked me out and made me realize how they trick you into sharing info you didn't intend. I don't see why they wouldn't opportunistically exploit mic permissions just the same. Hell, that might be why they created the calling feature in the first place!
Maybe not those devices in particular, but there are plenty of terms of service and spokespeople on the record saying utterances are captured and passed on to third parties. One TV company CEO famously said you should leave the room to have a private conversation.
To prove that it's possible to buy ads targeted on microphone audio. If this kind of targeting was really as widely available as the article implies, a journalist would be able to do it fairly easily and write an article about it.
The meaning of this is that the proposed technology doesn't exist and is bullshit. If "CMG Local Solutions" were actually able to listen in on your conversations, investigative journalists would be all over this trying to prove/disprove and understand the privacy implications. The fact that they haven't speaks volumes.
These kinds of topics come up frequently ("Your phone is secretly tracking you!!!") and this may sound a bit harsh but it always is interesting to me how people aren't able to use basic critical thinking skills to debunk them. It's like I sometimes say "The conspiracy theory doesn't even make sense even if you take everything it presupposes at face value":
1. As you point out, this isn't even some "super secret government" tracking you. They are basically advertising that anyone selling any random shit could take advantage of this feature. So obviously it would be trivially easy for anyone with the motivation to go and buy some of these ads to get clarity if some voices were actually being tracked. There is literally no way for this technology to be hidden given how it is described in TFA.
2. Companies don't need these secret tracking technologies, because the overt tracking ones already give them tons of information: your location, what you searched for, browsing history, your friends and family, etc. etc.
3. Given that ads are already super targeted, whenever this story comes up you always get these anecdotes of "I was talking about some totally random thing like pantyliners and then THE VERY NEXT AD I SAW WAS FOR pantyliners! Indisputable proof!" I've certainly had a similar experience. But it's certainly not that hard to see how these kinds of coincidences would be quite common given how good ad targeting already is.
Like the saying goes, "Anyone who is a conspiracy theorist has never been a project manager..."
These companies are themselves claiming they can do this.
They could be lying, but it’s not at all unreasonable to take their word for it when so many people report dramatic specific hits on their speech.
If the CIA put up a page on their web site saying that they really did sell crack in the 80s, would you still try to debunk that claim? (to pick a random example)
> I appreciate that Gizmodo actually calls out ridiculous nature of the claims. The original 404 Media article took every claim at face value and stirred some additional hyperbole to boot.
Just because these capabilities exist doesn’t mean they exist broadly. They may have a few apps in their network that had a reason for the user to provide microphone access and then they exploit it for this. But it’s likely relatively obscure apps that aren’t well scrutinized. They might be able to target hundreds of users with it, and play up the “modern” approach and then do some lookalike thing to pad the audience into something that appears more effective to the customer.
I have always denied Instagram access to my camera, microphone, photos, and location, on top of disabling background refresh. Sometimes, upon switching to the app, I get this screen: https://i.imgur.com/Sj8Dikg.png
This has always struck me as odd, because it means the app would have started both the camera and microphone if it had permissions, without me clicking the right buttons to get to it.
I'm too lazy right now to see if I can trigger the screen, and whether iOS would show the microphone/camera activity icons, but maybe there's a bug that's actively being exploited by Meta. I don't know how active the microphone stays if it has permissions, but I wouldn't put it past them to send all the data captured by the sensors even if I'm not posting a story. They did it with "status updates" on Facebook back then, when they gathered the data even if I deleted whatever I wrote without posting.
Readit News
Yesterday was the worst one yet. We were driving and my daughter was filling out MadLibs by hand in a paper booklet in the backseat. One of the fill-ins we came up with was "pantyliner".
As soon as we got home from the event, I sat down and YouTube gave me an above-the-fold recommendation for a Japanese pantyliner commercial.
I've seen enough. This word never comes up in day-to-day conversation, it's not in my interests, and it's one of the least exciting topics imaginable. None of me, my wife, nor my daughter had hands free to search it up at that time, and besides, we were totally preoccupied with driving and our event until I got home a couple hours later.
For the record, we're an iPhone family and were in a Tesla.
I had been convinced by, I think, a Simon Willison article that this isn't happening. But he's wrong. This is happening.
Maybe subconsciously in videos or other browsing activity pantyliner adjacent things were discussed.
Not saying companies aren’t using audio for ads, but it just feels like there would be more leaks from the big COs if it were ubiquitous in everyday life
I use the same email address on both accounts, but it seems weird that TT would export its follow graph to YT for any reason.
[1] https://en.wikipedia.org/wiki/Frequency_illusion
In the following days, I started noticing the ads for that product everywhere -- in the physical world! Not online, not in my browsing, but in shops and buses and street signs.
Got me wondering which came first, the ads, or me thinking of the product? Probably the ads....
His family is surrounded with microphones, cameras and sensors. The company has a long track record of unwarranted surveillance.
In the most recent case against Google for wiretapping, all claims survived the dismissal and summary judgment stages. Google settled instead of showing a jury that what the company does is not wiretapping. Why would anyone trust this company, except out of necessity.
Anecdotes are not proof, but are anecdotes needed to form a reasonable suspicion of Google conducting unwarranted surveillance.
For some reason, HN commenters will oppose the notion of Google/Meta using microphones for data collection but how much does that matter when we already know Google uses any available means it can get away with.
edit: found some info on this here https://news.ycombinator.com/item?id=38688101
The data broker industry is huge, shady, unregulated, and in many cases offshore.
After playing for a few days, we were seeing easily 20-30 per day. I'm sure I'd seen them many times before (they were one of the most popular cars in Australia for a few years before that), but I'd never really noticed them. I've played the game with a few other car models with friends since, and it always results in a comment like "Wow, there are so many X on the road".
I understand why your story feels chilling to you and how it would convince you, however it's more likely that:
a) in the 100+ ads you see per day[0] you've seen panty liners before but it was never relevant to you (like a Getz) or
b) you've seen ads for panty liners recently but not conciously taken it in, resulting in the inverse of what you thought happened - your MadLibs suggestion was due to an ad.
I'm very confident many companies would be happy to spy on you via microphone if practical, but outside of malware I'm yet to see anyone proving this actually happening. We have billions of smartphones in the world and millions of tech people, many who would have to be keeping the secret all these years and many others who somehow couldn't find evidence of this. Occum's razor just doesn't add up here.
[0] https://www.thedrum.com/news/2023/05/03/how-many-ads-do-we-r...
https://www.businessinsider.com/google-exec-nest-owners-shou...
They've been warning us themselves.
Slightly off topic but I did like one of them being so frank about Linux and open source. They worked for a while with the NSA, all they said was "Linux has 100 million lines of code. You have to fooling your self to think we didn't slip in hundreds of backdoors onto that thing". The same can be said of almost any other system though.
I have long suspected that the gaslighting thing of "it is just subconscious coincidence" is just a very neat cover for these things. There might be a slight influence but it is nowhere near as powerful an effect as they would like you to think. With advertising swaying people, they can move a few percent of people a few percent in one direction but it doesn't really target individuals effectively in guaranteeing results. Only as an aggregate.
This is exactly why I hate this dumbass conspiracy theory. You're totally correct, they are both capitalist corporations driven by profit alone. And the fact is that if they were secretly recording us in direct opposition to what they have said on the record it would be an absolute disaster for their bottom line.
Deleted Comment
Now the hard part: DO NOT tell anyone your word, do not type it into anything, its best you don't even mention the experiment to people. If you tell your word to people then they might type it into a computer which would taint the experiment.
Within a week I was seeing ads for mulch.
On talking lovingly about smoked salmon, it took 2 days for it to turn up.
I have also seen it happen with folks talking about Jeep for instance.
I guess I could choose like 10 words and draw two to talk about and two controls to think about to notice any selection bias.
I don't use any sort of voice-based smart assistant like Google Home / Alexa.
Since this experiment, I've been a lot more conscious of microphones. My desktop's microphone and my laptop's microphone both have hardware kill switches. My desktop's webcam no longer has a built-in microphone. My television and cellphone still have microphones without kill switches.
It’s been four years so I should really redo this research.
0: https://hugotunius.se/2021/01/10/the-apps-that-listen-to-you...
The page has been removed from their website, but it's archived:
https://web.archive.org/web/20230927000839/https://www.cmglo...
It's all just really dumb fearmongering that ultimately hurts the credibility of privacy advocates as a whole.
The sad thing is, these companies are collecting mass amounts of sensitive data and using it to drive ads, and we should be doing more about it. There's absolutely no reason to fabricate nonsense like this, the reality is already terrifying enough to creep people out.
Maybe a lot of the folks this happens to haven't denied microphone permission? Or maybe someone around them hasn't? Facebook has a calling feature, chances are people (or others around them) have given it access without realizing it. I have certainly caught Facebook Messenger snooping on my location when it had no business to, merely because I had granted the permission days ago when sharing my location with my friend, and forgotten to deny it again. It spooked me out and made me realize how they trick you into sharing info you didn't intend. I don't see why they wouldn't opportunistically exploit mic permissions just the same. Hell, that might be why they created the calling feature in the first place!
1. As you point out, this isn't even some "super secret government" tracking you. They are basically advertising that anyone selling any random shit could take advantage of this feature. So obviously it would be trivially easy for anyone with the motivation to go and buy some of these ads to get clarity if some voices were actually being tracked. There is literally no way for this technology to be hidden given how it is described in TFA.
2. Companies don't need these secret tracking technologies, because the overt tracking ones already give them tons of information: your location, what you searched for, browsing history, your friends and family, etc. etc.
3. Given that ads are already super targeted, whenever this story comes up you always get these anecdotes of "I was talking about some totally random thing like pantyliners and then THE VERY NEXT AD I SAW WAS FOR pantyliners! Indisputable proof!" I've certainly had a similar experience. But it's certainly not that hard to see how these kinds of coincidences would be quite common given how good ad targeting already is.
Like the saying goes, "Anyone who is a conspiracy theorist has never been a project manager..."
They could be lying, but it’s not at all unreasonable to take their word for it when so many people report dramatic specific hits on their speech.
If the CIA put up a page on their web site saying that they really did sell crack in the 80s, would you still try to debunk that claim? (to pick a random example)
> I appreciate that Gizmodo actually calls out ridiculous nature of the claims. The original 404 Media article took every claim at face value and stirred some additional hyperbole to boot.
This has always struck me as odd, because it means the app would have started both the camera and microphone if it had permissions, without me clicking the right buttons to get to it.
I'm too lazy right now to see if I can trigger the screen, and whether iOS would show the microphone/camera activity icons, but maybe there's a bug that's actively being exploited by Meta. I don't know how active the microphone stays if it has permissions, but I wouldn't put it past them to send all the data captured by the sensors even if I'm not posting a story. They did it with "status updates" on Facebook back then, when they gathered the data even if I deleted whatever I wrote without posting.
</faraday cage hat>