Every time I read some technical description about why this isn't happening, the technical description seems convincing.
However...
A friend tested the theory a few years ago. He doesn't own a swimming pool, doesn't want to, and has never expressed any desire to. He put his and his wife's phone on the table and said to the wife (loudly), "Why don't we look into pool fencing?". She agreed with him. Shortly after, on both of their phones, on a particular social network, they were inundated with ads for....pool fencing.
Think about what this implies. If your phone is listening, it’s listening all the time, right? So like 12-18 hours of continuous audio every day. That’s a lot of ad triggers. Way too many to actually be served with ads during your browsing time, which is a strict subset of your total audible proximity to your phone (plus ad inventory is a strict subset of what you view on your phone).
So how does the phone + ad networks decide which words to prioritize to trigger which ads when?
So for this anecdote to be true, not only would the phone have to be listening, but the targeting algorithm would need to decide to actively exclude all the other audible triggers from that time period, and fill your limited ad impression inventory with the one phrase you were intentionally testing.
How would it do that? Especially if this is indeed an outlier one-off topic of conversation that you cover in a single sentence. There would not be contextual clues (like repetition over time) that might indicate you are actually “in market” for a pool fence.
To me this is the problem with these anecdotal tests. You understood that that was an important phrase in the context of ad targeting. But how did the automated ad system know it should serve you ads on that topic, and not one of the many other advertisable topics you talk about over the course of several days? Or that your phone hears over several days?
1) App stores the trailing two minutes of speech in memory.
2) If the app detects a consumption-related trigger word, the related conversation is flagged for transmission to the server.
3) Flagged audio block is converted to text. Consumption related verbs ("buy", "purchase", etc) are identified. The syntax of the sentence clearly indicates which noun is the target of a given consumption-related verb ("new car", "pool fencing")
Your voice is unique and can be fingerprinted to ID you (see Alexa devices). Add in things like positive sentiment analysis, changes in vocal inflection/intonation and context surrounding spoken products like purchase inference/intent and you can probably triangulate a threshold for showing products with high likelihood of purchasing intent.
Really smart people have been working on these things at Google for decades and that’s barely scratching the surface of this nuanced discussion. CPU/GPU has only gotten faster and smaller with more RAM available and better power management across the board for mobile devices.
Anything is possible if there is money to be made and it’s not explicitly illegal or better they can pay the fines after making their 100x ROI.
My phone can listen all day every day. It listens for "hey google" and it can listen and passively tell you songs that are playing. It's not outside the realm of possibility to do their audio fingerprinting on keywords and what not. The advertising potential makes it extremely juicy
The system knows to serve you ads about the new topic because it's new. You're already getting ads for the stuff you're normally talking about. The new topic stands out easily.
It doesn't have to be your phone. Could be your TV or any other device.
Most importantly there's just patterns of behavior. Companies are absolutely desperate for every scrap of data they can get on you. Why would they not capture audio from your mic?
You’re so right. We should just trust the computers in our pockets, hands, and nightstands 24/7/365 running proprietary operating systems, firmware, and sensor suites phoning home as much targeting data as they can possibly collect — but not that! What could they possibly gain from harvesting that?
>So how does the phone + ad networks decide which words to prioritize to trigger which ads when?
The same way they analyze your email and web searches. Basically, statistics.
>To me this is the problem with these anecdotal tests. You understood that that was an important phrase in the context of ad targeting. But how did the automated ad system know it should serve you ads on that topic, and not one of the many other advertisable topics you talk about over the course of several days? Or that your phone hears over several days?
Buddy, so many people have witnessed this happening for at least 10 years and even done experiments at this point that it's common knowledge. I know for a fact that one of my friends now has a phone that is especially receptive to hearing me say things around it, because our conversation topics ALWAYS come up in my searches, ads, and feeds shortly after. Think about that. Someone else's phone sends data to a cloud that I never gave permission to. It then puts that together with data from MY phone about where I was (perhaps even the devices chirping at each other!). The aggregation happens within a week then I see relevant ads. I've seen this happen dozens of times. It's no coincidence.
As far as the article, I'm not even going to read it. It's got to be stupid. We know from leaks, reverse-engineering, and personal experience that this spying is going on. I question the source of this article, but I suppose we should never underestimate the lengths someone will go to in order to feel that they are smarter than the rest of us with our eyes open.
Why did they pick a swimming pool? Did they see people in their area installing pools? I think that's often people's best guess, is that the "random" thing people use to test this actually isn't random and subconsciously they already had this topic seeded to them.
Something similar -- while on a family visit at my parents' house, my brother was talking about his upcoming Hawaii trip, Specifically he was going over a snorkeling adventure he signed up for.
For the next week or so, I got many ads on my phone about underwater packages for Hawaii, along with ads for various snorkeling and swimming gear. Now I had never researched any of that on my phone, however obviously my brother has. And the ad trackers saw that both my phone and his had communicated out over the same IP address (my parents wifi) on other random internet connections, so that is probably why they were then targeting my tracker cookie with ads that would be related to his tracker cookie. (This is all technically "easy" for the trackers to do, and seems logical that they would, because "why not").
On an unrelated note, I was making a peanut butter sandwich, started browsing some sites, and started getting ads for Skippy peanut butter. My phone must have smelled the peanut butter in the air.
I know my iPhone isn’t listening to me. And I know about my friend’s activity influencing the ads I get served, and my demographic, and location, and all of that. And my random idea for a test word being predictable in a shocking way.
But, recently I started thinking about the average user, who will install anything and approve any permissions requested without reading it. And imperfect App Store reviews approving a Trojan horse accidentally.
Am I positive someone hasn’t inadvertently allowed mic access to a malicious party? I wonder if that person’s phone may, in fact, be listening to them.
No, they deliberately chose a topic they had absolutely no interest in, to try and avoid confirmation bias. It’s not impossible that what you describe is actually what happened to an extent though, a lot of the recommendations and ads on FB do seem to have a “what people around you / in your network like” factor.
Confirmation bias makes it hard to extract much from these types of anecdotes. On a daily basis you might be talking about dozens of products. If your lookback period is a few days, that could 100s of products, and you'll get spooky coincidences pop up from time to time from pure chance alone.
Technilogical causes are much more likely than accidental causes for such effects to appear, in today's world.
Occam's Razor and the answer to the question, "What kinds of companies are at work in the environment?" push that probability in a specific way, because the motives and means are definitely there. Do you think they are the kinds of companies that would waste such an opportunity?
Their Chief Councel's recommendation depends on how slimy they are, right?
What would happen if they got caught? Slap on the wrist would be all, if that, no?
Something similar happened to me with backpack zippers. It convinced me the phone is listening and serving me ads despite the technical explanations that it isn't.
I was walking to work and my backpack zipper broke getting off the elevator. When I got to my cube I set my phone on the desk and said to my coworker, "damn, my backpack zipper just broke!" 45 minutes later I was in a meeting and checked my phone and backpack zipper ads appear. I had never googled backpack zippers before, never seen backpack zipper ads.
Literally the only proceeding thing before getting these was was telling my coworker that mine had just broken.
But this is just selection bias. If a hundred people do that and one gets an ad, it’s proof. Nevermind the 99 others who never saw a thing and wouldn’t bother posting.
The only way to test this would be to have your anecdote together with the complete marketing profiles of your friend and his wife. If such a profile could even be compiled in principle, from it we would be able to tell whether your friend or his wife had generated any non-audio pool-related signals, or whether they had seen other pool-related ads recently. Also, it'd be nice to know how often people in their marketing categories receive ads for pool fencing. Could be an astonishing coincidence.
It’s definitely a difficult one to test in a scientific way. But they 100% had no interest in pool fencing, living long-term in a rental townhouse. They chose the phrase specifically to be something they had no interest or search history in.
The most striking example happened to me while watching a documentary about siberian cats.
We were watching it in Italian, our main language, and I wanted to know more about it, as I typed "g", the first result was "gatto siberiano", exactly the cat I was looking for. Way too specific.
Another time as my girlfriend said she was interested how much a specific model of a watch a friend of him costed, the very same happened, as I typed the first few letters the very watch brand and model appeared.
Since then, I just don't care about how much technical description I can read, nothing's gonna convince me of it being a coincidence.
I’m fascinated that this urban legend persists among tech people because it’s so easy to disprove.
Did you know that you can set up a proxy from your phone and capture all traffic from it? It would be so trivial to find the traffic from your phone. There are ways to MITM and inspect the traffic, too.
There are also many people doing static reverse engineering of phone apps looking for security vulnerabilities. To believe this urban legend, you’d also have to believe that none of them have ever encountered this hidden voice analysis code.
If we ignore that, you know there are OS-level security controls on apps, right? iOS and Android don’t make it easy for apps to use the microphone constantly and run in the background to process it.
Finally, if we ignore all of that, how can anyone believe that these companies are recording conversations but none of their employees have ever chosen to blow the whistle? We’ve seen numerous FAANG “whistleblowers” come through with everything down to trivial or baseless complaints, but nobody has blown the whistle on these supposed widespread spying programs?
The whole urban legend is preposterous to anyone who has any experience with apps or phone security, let alone common traffic analysis or reverse engineering tools. I don’t understand why the myth is so persistent among even some technical people.
I'm not sure if the legend is true or not. But this argument doesn't really disprove it. The devices don't need to send full audio recordings. They are powerful enough these days that they can do a cheap on-device audio analysis and tagging, and then upload the (very small) tags. It doesn't need to be Siri quality analysis because it doesn't matter if the analysis is incomplete or sometimes inaccurate. They would just be scanning for certain keywords.
As for whistleblowing... Is there really that much to whistleblow about it? We already know that ad-based companies like Google are collecting our data every chance they get, because they make billions of dollars from it. They're scraping our emails, studying our GPS location, paying attention to who we are in proximity with, etc. The level of surveillance is incredible and people don't really care. It wouldn't be headline news to find out that they are taking advantage of yet another side channel.
>Did you know that you can set up a proxy from your phone and capture all traffic from it?
The phone knows about your proxy. There are phones - actual brands - that were caught on sending secret telemetry to their manufacturer, but only when not listened - definitely only on mobile data, no wifi, and I assume with cert pinning.
I know a person who was researching this and they needed a Faraday cage and a BTS to conduct experiments. So it's not exactly trivial.
The difference is that these were small Chinese brands that were not even that popular in my country - and still someone researched this. Imagine how much research Android and Iphone get, and there's not a single proof of and wrongdoing. Now that is unlikely.
This is just flat earth for technophiles. They don't really want to know the truth, they just want to enjoy their fantasy of living a conspiracy theory.
It is interesting how people always come up with anecdotes like this but none of them try repeating the experiment multiple times.
You might think the pool fencing example might be an extreme coincidence, but far weirder things happen every day. And what made your friend consider pool fencing as an example if they don't like pools? Maybe something they saw recently gave them the idea? Hmm...
Had this happen. “Airport tier tar” was the phrase someone said near me. Saw ads on Instagram for such a niche thing the next morning. Not only did I see ads they were insanely local. I have never needed to buy tar.
Then theres the time a friend told me about a very specific brand of Ramen, I opened up Facebook, and there it was, very first ad.
There is a video of Zuck denying they "recording peoples microphone" -- but how he said it with a smirk I took him to mean "we do on-device transcription and only send back keywords"
You see thousands of ads of every type every day and ignore them. Now you’re doing a test and consciously looking for ads related to pools. Of course you’re going to find something.
its called a noise gate: basic audio gear that triggers a function based on SPL(sound pressure level),which would be a reliable way to trigger a capture event and....the rest, without listening to everything.Change in tempo and pitch could also be good for an "event trigger".
the start of digital audio goes back to the 1980's
and the full suite of capabilities is trivial for any phone, as they are integrated extensivly to cancel background noise anyway.
And with so called digital voice assistants running, I cant be surprised.
My main point would be that ,everyone is convinced that there phones are spying on them,
its one more thing to make them flinch
and grimace, argueing about it will only draw deeper lines.And that, is where we are.
So...... the listening isn't very good? Because recommending a swimming pool simply based on the single word pool is just terrible.
Either they have the most technically impressive spying system that can't do anything right or it's just not happening and people are making connections where there isn't really any.
I’m not sure what point you’re attempting to make here, but they chose the phrase “pool fencing” and were rapidly inundated with ads for pool fencing, which, in isolation, would suggest the listening is extremely accurate.
we haven't done an experiment like that but I've had family know what medium certain topics were expressed and discussed over and those topics that landed in the ads that startled with their topic intersection were topics discussed only verbally vs typed into some search field or connected to some other web interaction etc.
That's kind of the smoking gun when you can create a disjoint set of topics and a disjoint set of mediums of communication delivery and see what shows up in the ad space from those discussion topics strictly expressed verbally.
But why did he pick pools? What if he lives in an upper middle class suburban neighborhood where everybody has pools? And what if he slowed his scrolling just a little too much on a pool ad on Instagram? What if he actually, kinda does think about getting a pool?
Who knows.
I'm just saying, the technical, ethical, and legal implications of creating an ad network that surreptitiously slurps up audio 24 hours a day in violation of the claimed terms of service without anybody leaking anything about it is a conspiracy that seems less likely than people just being more predictable than they would like to believe.
Whatever made him use pool fencing as his random example is probably also why the ad showed up. Maybe it's the season for that stuff, he saw other ads earlier, or other friends talked about it. He may not consciously remember that, but it could make him more likely to think of it again later. In other words he talked about it because of the ads, not the other way around.
...unless there were actually several thousand people who performed this experiment, got a negative result, and therefore don't remember it or post anything about it.
> Every time I read some technical description about why this isn't happening, the technical description seems convincing.
Having knowledge of the technical limitations and challenges myself, I used to be on board for this argument, but now less so.
All of the technical arguments against the listening seem to ignore the "Ok, <DEVICE>" or "Hello, <DEVICE>" initiating phrases for the voluntary surveillance devices people put in their rooms, and offer only a worst case defense ~"how could they process everything everyone is saying?!"
Why is it such a stretch to imagine these devices grab Direct Objects and Subjects and store those singular items for ad keywording?
We have cookies and know how they work, why is it difficult to extrapolate?
simonw is a breathless proselytizer of LLMs and likely is suffering from "a man's salary depending on misunderstanding" and all that.
It bears repeating, "these corpos are raising billions and hiring former alphabet heads to their boards for reasons other than just making you a better programming assistant."
> Why is it such a stretch to imagine these devices grab Direct Objects and Subjects and store those singular items for ad keywording?
How about because Apple say they don't do that, and can and do get sued if they say things like that which are not true?
(Sadly I make basically no money at all from my "breathless proselytizing" of LLMs. I hope to fix that this year, someone should pay me for this stuff!
You know I've written more negative things about LLMs than almost anyone else, right? 121 posts tagged AI ethics right here: https://simonwillison.net/tags/ai+ethics/ )
My experience working at one of the companies that gets accused of this a lot is that many colleagues wish we were as evil as claimed because it would be so much easier do their jobs that way than struggling through the reality of it which is endless red tape over the tiniest issues that have even the slightest proximity to privacy. So I've been a bit skeptical too.
Exactly. The big companies are scared of lawsuits and trying to get approval for something like that would be a nonstarter. As a matter of fact the device folks at the same company would be working hard to kill such an idea in its infancy because it’s already an uphill battle to sell always-listening or always-watching devices to consumers because of the creepiness factor.
And people also are terrible at math. Modern ML (regression & neural nets) are ridiculously good at predicting stuff you might be interested in, particularly when rich data sources like browsing and e-commerce histories are available; the decision to show the ad to you at some point almost certainly was made long before any audio-to-marketing pipeline could act on it.
Yup - having worked at Google Display Ads (arguably the epicenter of such talk), I personally only ever witnessed people walking the walk, privacy-wise. The threats to our privacy are quite public and not at all illegal; IMO data brokers and 3P browser trackers are at the top of the list, but all of Google’s known ills are there too (location tracking, exchange monopolization, allowing predatory advertisers, gestures broadly at chrome, etc etc etc).
They don’t need to be listening to us, and wouldn’t know how to even begin hiding it if they were. Something like that would require tons of compute and thousands of conspirators risking massive backlash, all to prop up a relatively tiny part of their business.
> Convincing people of this is basically impossible
Absolutely correct IME, btw. This is one of those things a smart engineer learns not to argue online, or at the Christmas dinner table for that matter. People tend to stand their ground on this one and move quickly to accusations of bias and naïveté…
I'm not trying to change your mind, but this response (from another user) was flagged, so I'm providing a pull quote.
> A marketing firm called Cox Media Group has recently revealed that it is listening to user conversations via their smartphones through its so-called "Active Listening" Software. With this, the company will push advertisements that users will see on certain platforms based on the heard conversations as unveiled by a report.
Well said. I've worked in adtech and this aligns with my experience. Alphabet probably wouldn't even make that much more money compared to its current ad program. There's no shortage of supply in display advertising.
My proof? Yesterday I was driving home and I saw an old Toyota Previa minivan and thought to myself "Oh, a Previa, you don't see those very often these days." When I got home, I started scrolling through my Google News and it showed me an advert for the new Previa.
I agree with Simon: you basically aren't going to convince someone that their phone doesn't listen to you and serve you adverts based on it, because they've run into instances where it seems like it.
I think people are prone to believing that their phone is listening to them out of an instinct to believe they are a lot more unique than they actually are. A clash between western individualism and hyper-efficient consumerism.
That's not me taking a stance on whether it's true, though. There would be a lot of fame in it for a whistleblower, but on the other hand if Google or Meta figured out a way to do it with a low chance of it being proven, why wouldn't they?
“Be part of the richest portion of the middle class and never have to worry about money again orrrrr mysteriously die 6 months from now leaving no identifiable impact”
> out of an instinct to believe they are a lot more unique than they actually are.
This is trivializing people in my opinion. The non-trivializing interpretation would be that for whatever reason people are skeptical that a black box that they don't understand very well, don't know how to audit, and don't know how to exert low level control over is doing things that they don't want it to do.
When framed that way it is immediately clear that this is an incredibly reasonable stance to take. The default assumption should always be that a third party who has a vested interest is pursuing it.
As an example. It is paranoia to assume that a 1970s era vehicle with almost no electronics in it is reporting on me to the manufacturer. It is willful ignorance bordering on delusion to assume that a vehicle manufactured in 2020 is not.
I think you're right on the money here. Most people are surprisingly predictable, and if your algorithm based on whatever works for at least 30% of people at least once a year, which is very low bar if you think about it, the population at large will notice
Had something similar happen to me. I wanted a mid sized ladder. Didn't search for it or even tell anyone that I was looking for a ladder but suddenly started getting a bunch of ads for ladders. Eventually I figured out that when I went hiking a few days earlier, I forgot to turn off my Fitbit after I got done. On the way home, I stopped at Walmart and looked at the ladders they had. I could see on my Fitbit account my path through the state park and through the Walmart, including the section with the ladders that I dwelled in front of for a while. That was enough to trigger some set of parameters somewhere and get me ladder advertisements.
> suddenly started getting a bunch of ads for ladders
The trouble is, what does "suddenly" and "a bunch of" mean? This doesn't sound very specific, let alone scientifically rigorous. How many ladder ads were you getting previously, and how would you know? Exactly when did the rate of such ads increase? Unless you have these details, it's highly possible you're estimating them inaccurately.
The hidden variable is the previa marketing budget. They have budget right now for a billboard and for online ads at the same time and they are focusing on your geography
They don't need to read your mind. The systems that push ads to you are also owned by the companies that are pushing contents and shaping how you see the internet. It's like how magicians incept ideas into your head then use it like they can read your mind.
One thing that is rarely discussed in relation to this urban myth is that it is widely believe and yet it doesn't actually effect the behavior of most believers. That is one of the major reasons why most companies and many governments don't care about privacy, the public doesn't really care about it either.
People will of course choose privacy over no privacy with all else being equal, but privacy is the first thing sacrificed when push comes to shove. If the average person is given the choice of having everything said within earshot of their phone being recorded and sent to Facebook or giving up Instagram, they'll happily choose Instagram and forsake their privacy.
If privacy advocates want to start turning the tide in this battle, the first step needs to be convincing the average person why privacy is important on a personal and tangible level. No arguments about future totalitarian regimes or hypothetical ideals. Abstract concepts like that rarely motivate people who have so many more practical political concerns. It needs to be something that is more important to people than having access to Instagram. And I have absolutely no idea how that could be accomplished which makes me concerned that we're already too late.
If people really do believe that their phones are spying on them all the time to show them ads that means that people are basically surrendering to an imagined surveillance state. They shut up and accept it, because they'd rather keep Facebook/Instagram installed than fight back.
I find that really depressing. I want people to have more agency than that.
We need people to understand the imagined v.s. the actual privacy threats, so they can push for better standards. If they believe in and submit to the conspiracy theories good luck getting anyone to campaign for actual meaningful improvements to the problems that are real.
If you believe that google is recording all your audio, uninstalling instagram is not going to cut it. I think such a person would have to go back down to a really dumb phone to have any confidence at that point.
People have only vibes, they think that if they paid with cash it would proobably be more private than a credit card, but what data is being sold and to whom for what uses? Is that even the case or are there regulations? If I constantly make cash withdrawals at the bank am I actually inviting extra scrutiny by looking like a money launderer? If I install this browser add-on maybe it sells all my data. But I'm also using chrome literally made by the ad company, and that youtuber told me if I don't use a VPN I'm constantly being tracked anyways...
If you just have a giant morass of confusing information about every digital decision, and a lot of annoying first steps you would take are likely to be no-ops, you just don't engage. People are defeated by ambiguity and lack of attention span, same reason lobbying works and people were constantly being poisoned by food & drug additives before the modern era.
I do think that this is a very "lightly held" belief. It's something people kind-of believe, they'll tell this to each other, but it doesn't affect any behavior - not because it's not important IMO, but because people mostly don't really, deep down, believe it.
And I do wonder if convincing people this isn't happening will have the opposite effect than we intend. Instead of being more aware of what actual privacy violations are, it'll just make people write off the whole idea of companies invading their privacy. Idk.
In my experience the people who believe conspiracy theories like this tend to have a fatalistic attitude towards them. They're not making any specific claims about specific apps listening in on their conversations. Rather it's more a vague sense that "they" (whoever "they" are) are always listening and that there's nothing that can be done about it without giving up on modern technology entirely.
I went back to my family for this christmas, and argued about privacy, I think it was about free email services again, and stuff likes this. And again - I just cannot understand it - the response was:
Yeah but I don't care, I have nothing to hide, let them have my data.
It's a slap in the face of me, trying to meticulously remove all internet access of programs and devices that don't need it, and moving from all free and not-privacy-friendly services to mostly paid and private-friendly ones - probably "losing" a big chunk of my lifetime doing this. I feel paranoid sometimes, when I hear this argument.
You are not losing your lifetime, you are building the future. If other people do not want to be on board it is a loss. I have spent years trying to teach people how to build their own websites and get away from the large walled gardens. Many will not come on board, but the community that does do it have built truly wonderful things. I think of the Neal Stepehson article "In the Beginning was the Command Line" where he compares Microsloths Family Sedan to the Free Tanks of Linux; there are all these Linux geeks yelling "come over here, we have free tanks" yet most people pay for the Family Sedan.
It sucks, yet I drive a Free Tank, and now help people online take care of theirs.
The other day I was talking with one from the future generation about what type of computer they would want and they said "A Fedora gaming desktop". And that made it all alright :)
I know someone who worked on smart TV software. They explicitly added audio fingerprinting via integrated microphone so they could determine which programs you were watching for ad profiling.
When in the history of online ads have advertisers not used available data?
Automatic content recognition is one of the freakiest kinds of tracking I've heard of. Smart TVs become data collection hubs for the home if they get connected to the internet. Some have microphones for voice recognition. If you go into the privacy policies for them, you'll find that your audio gets sent off to third parties (albeit only when in use).
Ha, I’m imagining a sort of thriller movie villain reveal where people are intently saying silly things into their phones, while the camera slowly pans over to the smart TV in the corner, which is actually listening.
I had this happen to me recently. I was speaking to a friend about a subject that suddenly started to appear in ads after I hung up. It was spooky. But I think it has more to do with the activities related to the people we know. I'm sure because of Android, Gmail and search they have a good understanding of our general association group. I bet my friend did a search on what we were talking about and Google was able to determine that others in the group would be interested in the subject. They don't need the microphone input. They have better (more direct?) ways to get the information.
What you described is basically how it works. You all get finger printed on the same network, finger printed on google searches, finger printed at the routers, and then associated. 6 guys get together for beer and someone brings up wood chips, 1 guy later googles BBQs, you're all getting BBQ ads for the next week and a half, the 5 other guys think the phone is listening to them because they don't know one of the guys actually googled or triggered something related to BBQs. I once knew my roommate was pregnant before anyone else because I got a bunch of ads for baby shit all of a sudden.
I actually believe the alternative theories are scarier than the phone actually listening to you. The one thing I keep hearing over and over again from people is that “I talked to so-and-so about this“ and now I’m getting a bunch of ads about it. What’s scary is that it knew you two were in the same room that one of you had a recent interest and that there was a likelihood you discussed it. That seems worse to me than just listening the conversations.
> ...they would need to be feeding those snippets in almost real time into a system which forwards them onto advertising partners who then feed that information into targeting networks...
I think this is totally mistaken. An ad seller which also wants to respect privacy keeps this data in-house and does the ad targeting themselves. The advertiser never needs to see personal information for this kind of market to give people ads related to overheard conversations.
Also, couldn't a lot of it be done on device? Say, have a set of key phrases to listen for, and if the device hears them, then mark that the user should be targeted by certain kinds of ads.
I don't have any evidence either way that that happens, but it seems like a more practical way to accomplish it.
I mean, many advertisers go through the Google display network without having Google target their ads (AKA run their bidding algorithms). Typically they go through other middlemen agencies, who indeed would need at least some of this data to make use of it, tho perhaps in some derivative form.
Still, you’re right that Google could be keeping it all for themselves and feeding it to their black box targeting services. I really strongly doubt that’s happening with incidental assistant snippets much less intentionally-eavesdropped recordings, but it is more plausible than this makes it seem.
The reason I think this isn't happening (other than they have told us it's not happening and technically seems very difficult to do this long without researchers/people figuring it out), is that in my life I've spent millions and millions of dollars on ad budget, we spent so much with twitter they made us custom hoodie, nobody tried to sell this function to me, it's not an enterprise level option, I've seen those, and it's not a button in the ad platform, so how is the key word getting into an ad network getting attributed to my ad profile and my ad served? There is a lot of rich ad enhancement you can buy at the enterprise level, but I've never seen anything like that mentioned even remotely. There are A LOT of ways I can target you, some of them pretty creepy, but using convo snippets from the phone? I can't even think how that would work.
Or, for that matter: if this functionality exists in phones, surely there were some people involved in developing the software and infrastructure to support it. It seems hard to imagine that none of them would have spoken up, even anonymously, to confirm its existence.
Like if an arduino can reliably carry 1 wakeword, an iphone could carry what? 64 at minimum?
Instead of waking up recording the conversation and sending it to the stasi, it could just toggle a bool in your secret advertising profile, then sync that up at random intervals.
They dont need context, they just need "Biscuit" "Nappies" "Birthday".
Not to say it isn't all made up. It probably is. I just dont think its technically difficult to achieve. It is probably all psychology. But sitting there with a packet capture going as you talk about dogfood isnt necessarily a great test to confirm the negative.
But if you're listening for 64, or even 1024 wake words, you won't get to the level of granularity people are claiming to observe, like in the other comment here where they tested "pool fencing". If you were putting together a list of 64 words, would you really include "pool fencing"?
The wakewords could be simply "buy", "cheap", "sale", "price" etc - and then collect the exact words around it. I bet the pool fencing conversation followed with the exchange of ideas where to get equipment.
There are certainly ways to make it more viable, but I hope that the systems up to billions of people use are sufficiently independently audited so that such features would quickly be found.
We search for everything we talk about these days…our location services are always on. The times I’ve noticed this I can usually trace it back to a friend had been searching for something and then we (our phones) were in the same place. Even the guy from the lawsuit with the doctor… his phone knows he was at that doctor and that doctor offers that procedure. Seems possible without listening.
Or more likely: Confirmation bias. I also occasionally get Nigerian memes on IG despite never having searched it. But unlike you I don't have the Nigerian film google search history that would lead me to conclude it was targeted.
Put another way, if you googled a Swedish film then saw Swedish memes you would think that was targeted. Then upon viewing the Nigerian memes you would have attributed it to a random video rather than anything targeted.
This is gaslighting. Even if it’s a coincidence, people have numerous examples of situations where they or anyone with them did not search for a thing. I’m not saying the only possible explanation is that the phone is spying on you - but I’m saying, don’t gaslight people and say “you probably searched about it and forgot”.
This doesn’t seem that hard to test? Put a phone on WiFi, run it through a traffic sniffer, and see if it is constantly sending little audio packets to listener1486.facebook.com or whatever.
Or heck, open up a phone and stick a probe on the mic lead. See when it is getting power or not.
There are smarter ways to do this without constantly sending all the audio. Wakewords + surrounding snippets for example. In fact if the audio recognition is done on-device then only the text data could be sent which would be indistinguishable from telemetry.
> 404 Media previously reported Cox Media Group (CMG) was advertising a service that claimed to target ads based on what potential customers said near device microphones. Now, here is the pitch deck CMG sent to prospective companies. Google has kicked CMG off its Partner Program in response.
The alternative is worse: more passive forms of targeting are so good that advertisers don't need to record your every moment in order to make a large number of people genuinely believe that they are.
It's relatively comforting to think they'd be so brazen as to care about your every word, as opposed to the fact that they own so much information about you that they can predict your actions and thoughts better than you can.
Reuters journalist Jon Stempel repeated the plaintiffs' paranoid and unsubstantiated allegations but did not include a single statement from Apple, just a vacant "did not immediately respond". If only there were years of statements made under oath in the courts to draw from.
Siri uses a pseudonymous identifier when communicating with Apple's services. The identifier is not linked to your Apple ID. Therefore Siri does not have access to your iCloud data. When you ask Siri to "call mom", it constructs a search query which is then executed on-device against your contacts database.
More of these requests are served entirely on-device. Just ask: how is that consistent with the idea that they are doing ad targeting based on Siri requests?
You can request a data export for your Apple account and see that there's no Siri data included in it. If they are caught lying, EU regulators will have a field day.
Apple describes their ad targeting here. They list the—frankly boring—signals they use for ad personalization, which does not include anything you say to Siri or have in your personal iCloud data. https://www.apple.com/legal/privacy/data/en/apple-advertisin...
No one is apologizing for Apple's nontransparent collection of Siri audio for QA purposes, which is what they settled over and made opt-in.
Neither do I. There’s no technical way it can be happening without it being easily spotted, not to mention how many would have to conspire.
However I am still waiting for a comprehensive explanation for how it is actually happening.
There there are many ways to correlate people and their interests, but I wonder how deep it goes. Basic geolocation and public interaction metadata: Sure. But I’ve also heard people believe that it ought to be possible to spot closeness via Wifi and Bluetooth.
This is such a culturally relevant topic that there must be some serious knowledge about it somewhere.
As for my own tests, I’ve made it a point to mention that I’m considering spending money on a well advertised-for product that I haven’t yet searched for online every time this topic comes up. No ads so far. In the next phase I’ll progressively tell more people to start looking it up.
The argument I use to convince people is to ask them how long your battery lasts when you’re running any apps, I’ll then ask how long do you think your battery would last if the apps were listening all day long.
I can’t tell if my friends are convinced or if they’re too polite (or disinterested) to argue.
It's hard to tell I can imagine some motivated individuals could utilize all sorts of packaging systems and embed them in third-party applications and so on, and extract pertinent information using this type of surveillance, and then sell this data to data brokers which would sell it to the large ad networks. I mean there's lots of ways to transcribe even most of the whisper models can run all the way down to 150 megabyte file not to mention the quantization versions of these models. I have something that I run on my computer for my server not throughout my house that does real Time transcription and whatnot but I use it for my own purposes, so you know someone who makes money off advertising or even selling insights about people, would certainly find ways to do this. I mean it's simply not regulated is it?
As of 2019 (my last contact with the industry), there was no real-time onboarding solution that could do this. Or if it existed it wasn't good enough to attract any attention. People were saying this in 2014 or so. I remember around 2017 thinking it would be pretty cool (repurpose wake-word detection to do some pre-determined segments or something) but no-one had it working. There's all sorts of stuff going on behind the scenes in ad tech and if https://news.ycombinator.com/threads?id=manigandham would probably have something useful to share if he was still around.
But it wasn't real. Even back then people would publish "sources" saying it was being done, but I'm telling you that if it worked well people would be doing it and we were in the middle of it all and couldn't find anyone who was.
There were smart-devices that acted as beacons to report people to ICE and stuff like that and they used personal data to tie it together. So there was crazy stuff out there (none tied to the respectable ad tech industry HN knows as the "privacy-violators" and so on). But this specific thing wasn't there.
Ask HN - let's say hypothetically, our phones are as bad as some think. Face camera taking covert pics, regular screenshots sent back home, microphone recording every conversation, every message, email, and bit of internet traffic scraped and harvested for ads and anything else - how would you change your habits if you knew this was happening?
Have you? I can't be bothered to cite but all of the above can and does happen in certain contexts - the covert face camera is admittedly a stretch, but all your personal photos are usually whisked away to the cloud without warning, which is just as bad (see the recent apple photos news)
The danger with this kind of opinion is that you start out doing some healthy critical thinking, and research. You form an opinion, based on good intentions. Then you spend all your time on explaining your conclusion to others. At that time, people who have the same opinion flock to you, and you don't have time to do more research.
This is the problem with experts and politicians. You can make money on either position, but once you have started stating your opinion, it's unlikely you'll do active research to disprove your opinion, as many have already concluded [1].
How about a pitch deck for a company [1] claiming it can offer microphone eavesdropping based ad targeting [2]? Maybe that was still aspirational though.
TVs and cars have microphones now and are privacy nightmares. Car companies have patents [3] for in-car keyword based ad targeting. Without legalisation, it's really only a matter of time.
Go to a public library without any electronic devices and pick up a print copy of a magazine with some ranked list of products or brands. Select a handful that you don't think would be targeted towards you. For example, beauty products, childcare products, cruises marketed towards seniors, etc.
Over the course of a few weeks, use Siri to send text messages mentioning these brands, add them to your Reminders list, etc. Just don't allow it to redirect you to any apps or web searches, since that would invalidate the experiment.
Browse Apple News daily, which uses Apple's internal ad service. If you see any ads related to at least 2 of your canary brands, you have plausible evidence that your Siri interactions have been used for ad targeting.
From a potential culpability or liability perspective, the question of targeted advertising is irrelevant. Intercepting private communications without consent, absent an applicable exception, is still against the law, e.g., the Wiretap Act and/or the California Invasion of Proivacy Act, ads or no ads. Targeted ads would not be used as evidence that private communications have been intercepted.
Many years ago, Google was sued for wiretapping and lost. See Joffe v Google
More recently Apple was sued for eavesdropping without consent. They just settled for $95 million. See Lopez v Apple
My wife has a very new (<2yo) iPhone. We live in public housing, no luxuries. We were talking about getting her some slippes, and I made a joke about her not getting any Gucci slippers. The Gucci ad appeared on her phone within 30 minutes.
I sh_t you not. She was a bit gobsmacked. I was neither surprised nor disturbed.
That's real anecdata, my friends. It could have been coincidence, and while we have nothing to hide, it's not an optimal situation, whatever the cause.
Besides, we ain't buying no Gucci slippers, no way, no how.
Gucci is classified as an "aspirational luxury" brand. In other words, luxury for poor people. It makes total sense that it would be pushed to people in public housing. If you get recommendations for European brands you've never heard of selling $5000 t-shirts, then I'd be worried.
I'm not worried, and thanks for suggesting that advertising would be a viable motive given our geolocation. Good thing Apple doesn't also have means and opportunity, too. Silly me!
I guess we should all just blindly trust Apple. Thanks, friend.
My wife and I, just being goofy, have an inside joke featuring a relatively uncommon name that we ocasionally yell out, like a few times a week. Two months ago we got junk mail addressed to UNCOMMON NAME + our last name.
I don’t think its a coincidence. Something is listening. Its kind of messed up that I am a normally rational, skeptical minded person, fairly knowledgable about information security and in 2024 I can’t even draw any clear lines between a nutty conspiracy theory and reality.
I’m sorry but I don’t find this article basically just saying “yeah but what are the chances Apple would do that?” persuasive at all.
I think it's normal to have a hard time delineating between conspiracy and genuine concern when it comes to this. We're literally carrying around powerful computers with exceptionally reliable connectivity combined with high-resolution cameras and high-fidelity microphones that are frequently used for voice recognition. Not to mention that for 50% of US smartphones, the software is designed by the most prolific advertiser in the world.
I had an experience a few years ago where I had talked about a fairly niche product (I can't recall exactly what it was) and the next day I started seeing ads for that product all over the place. I commented about it to two of my coworkers that day, how I had been skeptical about the conspiracy that our phones were listening to us for marketing purposes but that this felt eerie. What shocked me was their response: They had seen the same ad all over the place. Since then I've had a hard time deciding what to be more concerned about: That my phone might be listening, or that I might have been subtly influenced into thinking about this thing; that my whole experience was actually a result of being susceptible to marketing directed at my demographic
>My wife and I, just being goofy, have an inside joke featuring a relatively uncommon name that we ocasionally yell out, like a few times a week. Two months ago we got junk mail addressed to UNCOMMON NAME + our last name.
Are you sure it's not because of an opsec fail? eg. you used your nickname when registering for some service, and that made it into some sort of mailing list? What you're seemingly implying (ie. that there's some sort of secret listening system that can figure out your nickname, tie it back to your address, and then send spam to that) makes little sense. Your name + address is already readily available. It's in the public records. You freely hand it over to random websites (eg. for online shopping). There's zero benefits in anyone making such a system to figure out people's names using surreptitious listening.
I am open to other explanations. I did not use the name as an online pseudonym, thats why I found it so odd.
I am not sure about there being no benefit either. We moved here very recently and scraping local property records would take time and not be easily automated. So what if some data aggregator still had a blank under the name field for our address and needed to fill it in so they could address letters since we more or less automatically throw away “current resident” letters. I just don’t know, yes its far fetched but I don’t see any other explanations as much more parsimonious.
There is something I've always been wondering about companies' privacy policy. When Google says that it does not send user audio data to their server, does that prevent them from sending say whisper embedding of the audio itself? Because technically it would not be audio. It would still include a very rich representation of what the user said. Rich enough to be used to sell things.
I find it completely plausible that either Google, Facebook or some other installed app is listening to your conversations to develop marketing profiles.
What I find amazing is that people don't use ad blockers! I wouldn't be able to tell you if they are harvesting my voice data, because I haven't seen an ad in years! It is trivial to block ads, why do so many people choose to see them?
What happens is that you in fact lend part of your brain to all the data that you see on your screen. You don't realize it but your subconscious mind sees senses a lot of stuff your conscious mind is not aware about. Until you see it after you just talked about it thinking it was your idea and panic.
What you actually talk about with people (excluding maybe the proffesional stuff) is mundane stuff you see on your screens, newspapers, billboards, etc.
I think it’s wild that your counter argument is basically just gaslighting everyone. There are tons of (possibly coincidental) examples of this type of thing happening. Even if it’s a coincidence, it’s not explained away by “maybe everyone just forgot or is not paying attention”.
The assumption is that it's either widespread or nonexistent, but I could see it being just a rare instance that gets A/B tested, and maybe by now the companies have decided it's not worth it. But IME all kinds of not quite kosher stuff gets A/B tested, and that would line up with anecdotes where it's too specific to be coincidence, but also uncommon and only affecting certain people.
I have successfully argued this for at least a certain class of devices. Namely, it practically can't happen _for small devices_. e.g., a TV remote with a voice feature activated by a button simply won't have enough _power_ to constantly be listening, uploading, or processing.
Harder to convince folks that _nothing_ does this, but takes an edge off their more paranoid tendencies.
I personally don't believe any argument that says it's not feasible.
Transcribing all spoken text and sending it home, sure, not feasible.
What if we have 256 keywords, or 65536 keywords, maybe preconfigured for particular products or product classes. Some basic linear predictive coding mechanism ( you know, what powered those '80s chips Stephen Hawking style, speak and spell, etc) - very very low computational overhead. When the word is triggered, queue a message back home at the next reasonable opportunity - user id, timestamp, word. It will only take a couple bytes. It can be slipped in anywhere and obfuscated by any means by nature of being so small, data-wise, even as a watermark of some sort. By using a timestamp and waiting until the next opportunity, maybe minutes, hours, or days away, no time correlation detection is possible either.
People say big tech is ethical, fine. Maybe some ad company is sponsoring some free app or game for the phone, and slipping this in there. Now the developer can pay their rent and food costs. Maybe the ad company is then selling that data back to big tech who washes their hands of any wrongdoing. Maybe it's all legal because the fine print of the EULA allows for this.
Seems to me though this can be figured out empirically, just have a voice play something like "need to buy adult diapers" or "new tires" etc next to a device, enumerate every device, look for ads on whatever very specific topic, minding along the way to tell nobody and never enter it in any internet-connected keyboard.
For sure! I mean, my Google pixel has "Now Playing" which is able to passively listen to the microphone for songs it knows, and displays them on the oled lock screen.
So, we already know that it is
A. completely feasible for a smartphone to do this.
B. At least a subset of smartphones have always-on microphones.
Maybe not a remote control... But why would you put it in a remote anyways when everyone has a phone?
What if the activation of the microphone for marketing purposes occurred not systematically but randomly, making it harder to detect such functionality? And what if this functionality were activated with code loaded remotely at the moment it needs to be used?
I've noticed people often research products, services, medical issues/ailments, etc. using Google (or similar). When they start seeing related ads on social media, they tend to attribute more to having discussed the topic out loud, forgetting about the searches they made from the same device or IP address.
Another option not mentioned is that the thing they’re talking about might have been in their subconscious because of an ad they’d recently seen but didn’t notice. The topic/concept/product eventually bubbles up to consciousness, they talk about it, then start noticing the ads that were already there.
listen to me when i say this. we will start killing if we don't have same access to that tech. why kill monster? well I'm dead anyway as a slave to their machine.
people better demand equal access or start slitting throats right now.
The post's arguments can basically be summed up in three categories:
1) "Doing this wouldn't be technically feasible or would require a technical effort wildly out of proportion"
2) "There are lots of psychological biases that lead people to believe something like this happened even if it didn't actually happen"
3) "Apple is such a nice and honest company, they would never do such a thing..."
As for 1), there is enough technical discussion in this thread to disprove that point. But just as a reminder: Google build an always-on song recognition service into android, free of charge, without any obvious monetization, just because they can. OpenAI released Whisper last year as open source, a highly precise audio transcription model. By now lots of variants for on-device use exist.
All that tech doesn't just exist, it's not even seen as a moat. It's already being commodified.
As for 2), yes of course cognitive biases are always a thing. The problem is that you cannot use them to disprove something. They constitute an absence of evidence, not evidence of absence.
As for 3), yeah no comment here. Except maybe, remember Snowden. "No one would do such a thing" has already been spectacularly wrong in the past.
I honestly couldn't care less if that tech existed back then or not, I care about what's possible today.
Between 2017 and today, there were massive changes, both in technological development but also in mindset: Surveillance capitalism became much more normalized and generally accepted as a standard part of business.
So if you argue that it was a baseless conspiracy theory back then, therefore it automatically must still be a baseless conspiracy theory today, that argument is flawed.
The author admits the fine for breaking this rule is basically a non-issue compared to the potential earnings. Why wouldn’t the most valuable company on earth see this as a huge, easy win to record and advertise based on those recordings?
Under GDPR don't they risk 10% of worldwide revenue? (Plus for a company like apple trying to be seen as the defender of privacy, the PR and reputational hit)
OK, but query me this batman. Lately, I have been inundated by ads for anything to do with Kayaks. I do not own a Kayak. I do not want to own a Kayak and have never even spoken the word as far as I can remember. But for some reason, I am seeing ads for a Kayak carrier, a Kayak launcher, roof rack, you name it. It is so weird. I cannot figure out why suddenly these idiots think I am remotely interested in Kayaking. Kayaking, it is just so random. But way back in the 2000s, a friend of mine worked for a web site dev company. Her job was to visit with clients and take pictures of whatever they were hawking, write some copy and the dudes back in her office would turn it into a web page. She would occasionally ask me to come along and stare at whatever crap they were selling and act like I cared about it while she snapped some photos. I joked that meant technically I was a model. But then one time she was doing a web page for a marketing company that did focus groups. That was back before we had smart phones tracking your every move. I posed in a fake focus group with some other people and after that, the dude that ran the place asked me if I wanted to be in real focus group and I agreed. They seemed to think there were certain connections that made no sense. For example, they were sure that they could market adult diapers to nerds that play video games who were to absorbed in their game to get off their ass to take a simple dump. Apparently that is actually a thing? Then I went to this one focus group where they were asking if you own a dog and buy a lot of peanut butter. Read between the lines. So gross. So, I am guessing, some marketing weirdo thinks there is some weird connection between something else in my life an Kayaks. It would be interesting to know what that is. But, I think, a lot of times, they might be right about those weird connections that make you feel like they are reading your mind. Just in this case, whatever it is, maybe that I watch a lot of documentaries and that means I like Kayaking? I guess they only have to be right some percentage of the time. Just not in this case.
Depending on your age and if you have any vaguely outdoorsy interests, you probably are due for a kayak interest. A bunch of guys I know in the late 30s early 40s are into kayaks right now. Kayaks and smoking meat.
Hmmm, I live alone and in a pretty remote area (on an island). I keep my wifi pretty secure, so, that seems pretty unlikely. But who knows, anything is possible I suppose. Maybe because I live on an island.
> As one report put it, “Amazon can acquire more comprehensive data on people’s living habits....” The company wants to sell real-world services such as house cleaning, plumbing, and restaurant delivery, but according to some insiders, the vision is more far-reaching: an omniscient voice that knows all experience and anticipates all action! Already, forward-looking Amazon patents include the development of a “voice-sniffer algorithm” integrated into any device and able to respond to hot words such as “bought,” “dislike,” or “love” with product and service offers.
The Age of Surveillance Capitalism (2018) - Zuboff, Shoshana
Given that this is form 2018, the rise of voice assistants since then, all the market forces at play and the reward highly outweighing regulatory penalty payments there is little that's supporting that users are not being recorded. The TOS phrasses like 'service improvement', 'may share with third parties', and 'user experience' are common place and are not particularly concise. Simply dismissing anekdotes and honing in on Apple Inc., the only company which does not want/need to sell data, is not convincing.
I think people tend to shoot the messenger, when it comes to privacy discussions. Getting told one is being spyed on by a divice, which is so intimt and close to one slef is embarassing. Of course things are more like a spectrum and differ from case to case.
That is EXACTLY what I said in my article: "Convincing people of this is basically impossible. It doesn’t matter how good your argument is, if someone has ever seen an ad that relates to their previous voice conversation they are likely convinced and there’s nothing you can do to talk them out of it"
To all the people expressing doubt and disbelief, refer to every whistleblower story which was derided as tinfoil hat conspiracy until blown up. If we still haven't learnt that money will make corporations (and people) do despicable things, only we are to blame ourselves.
What I worry about sometimes is malware sniffing for hardware wallet seed phrases being practiced out loud by users when they think they're alone. I always tell people if they keep crypto on a hardware wallet - NEVER speak your seed phrase aloud. This could also be spied on by laser microphone surveillance.
I had a brand of wine in a closed, we were discussing with my life our wedding. Immediately this one brand of wine popped up on my wife user feed, wall. That happens from time to time, so, yeah... I know that it works like this for sure.
Many mobile apps use an adtech SDK which does on-device audio keyword spotting.
The company making this SDK is super secretive, so not much information about it.
I saw a whole PR release for adtech oriented keyword spotting tech like this one time, I wish I'd saved it... I can't remember the name of that particular company.
But thinking from the first principles, do you really think that all this phonetic keyword spotting[1] IP developed for defense tech in the late 90s and early 2000s was abandoned once recording entire phone conversations and doing full speech to text on them became technologically possible?
I still remember lots of startups who did this stuff openly a decade ago, before Cambridge Analytica scandal[2]. After that it became impossible to get funding or get acquired, so the few that stayed in this field became very secretive.
oddly enough, I compare this conspiracy theory with clout-chasing word-of-god christians on social media: there's an element of "i don't understand how it works, therefore the simplest answer must be correct"
if everyday people knew just how much data companies have on them and their habits, they'd be absolutely horrified. There's so much personal attribution data that most companies that know how to market properly are going to get you at some point. You may think you're careful with what you share and what is being recorded, but if you are friends or relatives of someone who does, or if you live in the same street as someone who does, they data on you.
I used to work for these kind of companies, and have a lot of friends and family who currently do, and the stuff they track and the patterns they have never fail to surprise me and make me laugh.
Sure, if they could, they would absolutely listen in to you, that would make their jobs a hell of a lot easier, but they don't. But the data they have on you so goddamn accurate, it's easy to assume they must be.
What evidence would anyone accept that 1+1 is 3? That's a genuine response. The cases are so clear cut (words picked up from conversations that would never fit on whatever profile they have on me; up to specific mentioned brands and things completely alien to be) and so numerous. I still can't believe that anyone even debates it. It's blatant as hell and all of my peers notice it, too
Theres a reason why they put a microphone in every single Internet connected device they can, even when it serves little purpose. My remote has a mic ffs, not to mention oddly specific ads after certain conversations.
I have nothing to do with ad tech, or selling mobile phones or anything like that.
I don't like seeing people fall for obvious conspiracy theories, and I have used my own deductive reasoning to decide that this is an untrue conspiracy theory.
Every time I read some technical description about why this isn't happening, the technical description seems convincing.
However...
A friend tested the theory a few years ago. He doesn't own a swimming pool, doesn't want to, and has never expressed any desire to. He put his and his wife's phone on the table and said to the wife (loudly), "Why don't we look into pool fencing?". She agreed with him. Shortly after, on both of their phones, on a particular social network, they were inundated with ads for....pool fencing.
Think about what this implies. If your phone is listening, it’s listening all the time, right? So like 12-18 hours of continuous audio every day. That’s a lot of ad triggers. Way too many to actually be served with ads during your browsing time, which is a strict subset of your total audible proximity to your phone (plus ad inventory is a strict subset of what you view on your phone).
So how does the phone + ad networks decide which words to prioritize to trigger which ads when?
So for this anecdote to be true, not only would the phone have to be listening, but the targeting algorithm would need to decide to actively exclude all the other audible triggers from that time period, and fill your limited ad impression inventory with the one phrase you were intentionally testing.
How would it do that? Especially if this is indeed an outlier one-off topic of conversation that you cover in a single sentence. There would not be contextual clues (like repetition over time) that might indicate you are actually “in market” for a pool fence.
To me this is the problem with these anecdotal tests. You understood that that was an important phrase in the context of ad targeting. But how did the automated ad system know it should serve you ads on that topic, and not one of the many other advertisable topics you talk about over the course of several days? Or that your phone hears over several days?
1) App stores the trailing two minutes of speech in memory.
2) If the app detects a consumption-related trigger word, the related conversation is flagged for transmission to the server.
3) Flagged audio block is converted to text. Consumption related verbs ("buy", "purchase", etc) are identified. The syntax of the sentence clearly indicates which noun is the target of a given consumption-related verb ("new car", "pool fencing")
4) Serve related ads
22 replies →
Your voice is unique and can be fingerprinted to ID you (see Alexa devices). Add in things like positive sentiment analysis, changes in vocal inflection/intonation and context surrounding spoken products like purchase inference/intent and you can probably triangulate a threshold for showing products with high likelihood of purchasing intent.
Really smart people have been working on these things at Google for decades and that’s barely scratching the surface of this nuanced discussion. CPU/GPU has only gotten faster and smaller with more RAM available and better power management across the board for mobile devices.
Anything is possible if there is money to be made and it’s not explicitly illegal or better they can pay the fines after making their 100x ROI.
8 replies →
My phone can listen all day every day. It listens for "hey google" and it can listen and passively tell you songs that are playing. It's not outside the realm of possibility to do their audio fingerprinting on keywords and what not. The advertising potential makes it extremely juicy
13 replies →
The system knows to serve you ads about the new topic because it's new. You're already getting ads for the stuff you're normally talking about. The new topic stands out easily.
It doesn't have to be your phone. Could be your TV or any other device.
Most importantly there's just patterns of behavior. Companies are absolutely desperate for every scrap of data they can get on you. Why would they not capture audio from your mic?
You’re so right. We should just trust the computers in our pockets, hands, and nightstands 24/7/365 running proprietary operating systems, firmware, and sensor suites phoning home as much targeting data as they can possibly collect — but not that! What could they possibly gain from harvesting that?
1 reply →
It's not a strict 12-18 hour window. Instead, it depends on the time frame between specific vocal or conversational cues / signal vs. noise.
[dead]
>So how does the phone + ad networks decide which words to prioritize to trigger which ads when?
The same way they analyze your email and web searches. Basically, statistics.
>To me this is the problem with these anecdotal tests. You understood that that was an important phrase in the context of ad targeting. But how did the automated ad system know it should serve you ads on that topic, and not one of the many other advertisable topics you talk about over the course of several days? Or that your phone hears over several days?
Buddy, so many people have witnessed this happening for at least 10 years and even done experiments at this point that it's common knowledge. I know for a fact that one of my friends now has a phone that is especially receptive to hearing me say things around it, because our conversation topics ALWAYS come up in my searches, ads, and feeds shortly after. Think about that. Someone else's phone sends data to a cloud that I never gave permission to. It then puts that together with data from MY phone about where I was (perhaps even the devices chirping at each other!). The aggregation happens within a week then I see relevant ads. I've seen this happen dozens of times. It's no coincidence.
As far as the article, I'm not even going to read it. It's got to be stupid. We know from leaks, reverse-engineering, and personal experience that this spying is going on. I question the source of this article, but I suppose we should never underestimate the lengths someone will go to in order to feel that they are smarter than the rest of us with our eyes open.
24 replies →
"How would it do that? "
AI.
That is entire premise of 'Nexis' from Yuval Harari.
Individualized bot driven surveillance .
https://www.theregister.com/2024/09/16/oracle_ai_mass_survei...
""Ellison declares Oracle all-in on AI mass surveillance, says it'll keep everyone in line
Cops to citizens will be 'on their best behavior because we're constantly recording and reporting'""
7 replies →
Why did they pick a swimming pool? Did they see people in their area installing pools? I think that's often people's best guess, is that the "random" thing people use to test this actually isn't random and subconsciously they already had this topic seeded to them.
Something similar -- while on a family visit at my parents' house, my brother was talking about his upcoming Hawaii trip, Specifically he was going over a snorkeling adventure he signed up for.
For the next week or so, I got many ads on my phone about underwater packages for Hawaii, along with ads for various snorkeling and swimming gear. Now I had never researched any of that on my phone, however obviously my brother has. And the ad trackers saw that both my phone and his had communicated out over the same IP address (my parents wifi) on other random internet connections, so that is probably why they were then targeting my tracker cookie with ads that would be related to his tracker cookie. (This is all technically "easy" for the trackers to do, and seems logical that they would, because "why not").
On an unrelated note, I was making a peanut butter sandwich, started browsing some sites, and started getting ads for Skippy peanut butter. My phone must have smelled the peanut butter in the air.
6 replies →
I know my iPhone isn’t listening to me. And I know about my friend’s activity influencing the ads I get served, and my demographic, and location, and all of that. And my random idea for a test word being predictable in a shocking way.
But, recently I started thinking about the average user, who will install anything and approve any permissions requested without reading it. And imperfect App Store reviews approving a Trojan horse accidentally.
Am I positive someone hasn’t inadvertently allowed mic access to a malicious party? I wonder if that person’s phone may, in fact, be listening to them.
1 reply →
using multiple high value advertising targets would be best. like specific brands of mattresses or industrial equipment.
No, they deliberately chose a topic they had absolutely no interest in, to try and avoid confirmation bias. It’s not impossible that what you describe is actually what happened to an extent though, a lot of the recommendations and ads on FB do seem to have a “what people around you / in your network like” factor.
1 reply →
Confirmation bias makes it hard to extract much from these types of anecdotes. On a daily basis you might be talking about dozens of products. If your lookback period is a few days, that could 100s of products, and you'll get spooky coincidences pop up from time to time from pure chance alone.
And, if not you, your friends and family. “I know a guy who had it happen to him” is almost as bad for confirmation bias as “it happened to me”.
Technilogical causes are much more likely than accidental causes for such effects to appear, in today's world.
Occam's Razor and the answer to the question, "What kinds of companies are at work in the environment?" push that probability in a specific way, because the motives and means are definitely there. Do you think they are the kinds of companies that would waste such an opportunity?
Their Chief Councel's recommendation depends on how slimy they are, right?
What would happen if they got caught? Slap on the wrist would be all, if that, no?
12 replies →
Something similar happened to me with backpack zippers. It convinced me the phone is listening and serving me ads despite the technical explanations that it isn't.
I was walking to work and my backpack zipper broke getting off the elevator. When I got to my cube I set my phone on the desk and said to my coworker, "damn, my backpack zipper just broke!" 45 minutes later I was in a meeting and checked my phone and backpack zipper ads appear. I had never googled backpack zippers before, never seen backpack zipper ads. Literally the only proceeding thing before getting these was was telling my coworker that mine had just broken.
Maybe your coworker searched for backpack zipper?
1 reply →
But this is just selection bias. If a hundred people do that and one gets an ad, it’s proof. Nevermind the 99 others who never saw a thing and wouldn’t bother posting.
The only way to test this would be to have your anecdote together with the complete marketing profiles of your friend and his wife. If such a profile could even be compiled in principle, from it we would be able to tell whether your friend or his wife had generated any non-audio pool-related signals, or whether they had seen other pool-related ads recently. Also, it'd be nice to know how often people in their marketing categories receive ads for pool fencing. Could be an astonishing coincidence.
It’s definitely a difficult one to test in a scientific way. But they 100% had no interest in pool fencing, living long-term in a rental townhouse. They chose the phrase specifically to be something they had no interest or search history in.
1 reply →
The most striking example happened to me while watching a documentary about siberian cats.
We were watching it in Italian, our main language, and I wanted to know more about it, as I typed "g", the first result was "gatto siberiano", exactly the cat I was looking for. Way too specific.
Another time as my girlfriend said she was interested how much a specific model of a watch a friend of him costed, the very same happened, as I typed the first few letters the very watch brand and model appeared.
Since then, I just don't care about how much technical description I can read, nothing's gonna convince me of it being a coincidence.
There’s a breadth of possibilities between coincidence and the phone actively listening to everything via a microphone.
I’m fascinated that this urban legend persists among tech people because it’s so easy to disprove.
Did you know that you can set up a proxy from your phone and capture all traffic from it? It would be so trivial to find the traffic from your phone. There are ways to MITM and inspect the traffic, too.
There are also many people doing static reverse engineering of phone apps looking for security vulnerabilities. To believe this urban legend, you’d also have to believe that none of them have ever encountered this hidden voice analysis code.
If we ignore that, you know there are OS-level security controls on apps, right? iOS and Android don’t make it easy for apps to use the microphone constantly and run in the background to process it.
Finally, if we ignore all of that, how can anyone believe that these companies are recording conversations but none of their employees have ever chosen to blow the whistle? We’ve seen numerous FAANG “whistleblowers” come through with everything down to trivial or baseless complaints, but nobody has blown the whistle on these supposed widespread spying programs?
The whole urban legend is preposterous to anyone who has any experience with apps or phone security, let alone common traffic analysis or reverse engineering tools. I don’t understand why the myth is so persistent among even some technical people.
I'm not sure if the legend is true or not. But this argument doesn't really disprove it. The devices don't need to send full audio recordings. They are powerful enough these days that they can do a cheap on-device audio analysis and tagging, and then upload the (very small) tags. It doesn't need to be Siri quality analysis because it doesn't matter if the analysis is incomplete or sometimes inaccurate. They would just be scanning for certain keywords.
As for whistleblowing... Is there really that much to whistleblow about it? We already know that ad-based companies like Google are collecting our data every chance they get, because they make billions of dollars from it. They're scraping our emails, studying our GPS location, paying attention to who we are in proximity with, etc. The level of surveillance is incredible and people don't really care. It wouldn't be headline news to find out that they are taking advantage of yet another side channel.
>Did you know that you can set up a proxy from your phone and capture all traffic from it?
The phone knows about your proxy. There are phones - actual brands - that were caught on sending secret telemetry to their manufacturer, but only when not listened - definitely only on mobile data, no wifi, and I assume with cert pinning.
I know a person who was researching this and they needed a Faraday cage and a BTS to conduct experiments. So it's not exactly trivial.
The difference is that these were small Chinese brands that were not even that popular in my country - and still someone researched this. Imagine how much research Android and Iphone get, and there's not a single proof of and wrongdoing. Now that is unlikely.
This is just flat earth for technophiles. They don't really want to know the truth, they just want to enjoy their fantasy of living a conspiracy theory.
It is interesting how people always come up with anecdotes like this but none of them try repeating the experiment multiple times.
You might think the pool fencing example might be an extreme coincidence, but far weirder things happen every day. And what made your friend consider pool fencing as an example if they don't like pools? Maybe something they saw recently gave them the idea? Hmm...
Had this happen. “Airport tier tar” was the phrase someone said near me. Saw ads on Instagram for such a niche thing the next morning. Not only did I see ads they were insanely local. I have never needed to buy tar.
Then theres the time a friend told me about a very specific brand of Ramen, I opened up Facebook, and there it was, very first ad.
There is a video of Zuck denying they "recording peoples microphone" -- but how he said it with a smirk I took him to mean "we do on-device transcription and only send back keywords"
You see thousands of ads of every type every day and ignore them. Now you’re doing a test and consciously looking for ads related to pools. Of course you’re going to find something.
its called a noise gate: basic audio gear that triggers a function based on SPL(sound pressure level),which would be a reliable way to trigger a capture event and....the rest, without listening to everything.Change in tempo and pitch could also be good for an "event trigger". the start of digital audio goes back to the 1980's and the full suite of capabilities is trivial for any phone, as they are integrated extensivly to cancel background noise anyway. And with so called digital voice assistants running, I cant be surprised. My main point would be that ,everyone is convinced that there phones are spying on them, its one more thing to make them flinch and grimace, argueing about it will only draw deeper lines.And that, is where we are.
Now tell it over a campfire and maybe shine a flashlight under your face.
The party told you to recount the evidence of your eyes and ears as a campfire ghost story. It was their final, most essential command.
So...... the listening isn't very good? Because recommending a swimming pool simply based on the single word pool is just terrible.
Either they have the most technically impressive spying system that can't do anything right or it's just not happening and people are making connections where there isn't really any.
I’m not sure what point you’re attempting to make here, but they chose the phrase “pool fencing” and were rapidly inundated with ads for pool fencing, which, in isolation, would suggest the listening is extremely accurate.
we haven't done an experiment like that but I've had family know what medium certain topics were expressed and discussed over and those topics that landed in the ads that startled with their topic intersection were topics discussed only verbally vs typed into some search field or connected to some other web interaction etc.
That's kind of the smoking gun when you can create a disjoint set of topics and a disjoint set of mediums of communication delivery and see what shows up in the ad space from those discussion topics strictly expressed verbally.
But why did he pick pools? What if he lives in an upper middle class suburban neighborhood where everybody has pools? And what if he slowed his scrolling just a little too much on a pool ad on Instagram? What if he actually, kinda does think about getting a pool?
Who knows.
I'm just saying, the technical, ethical, and legal implications of creating an ad network that surreptitiously slurps up audio 24 hours a day in violation of the claimed terms of service without anybody leaking anything about it is a conspiracy that seems less likely than people just being more predictable than they would like to believe.
I have similar anecdotes about this and Yamaha guitars.
Whatever made him use pool fencing as his random example is probably also why the ad showed up. Maybe it's the season for that stuff, he saw other ads earlier, or other friends talked about it. He may not consciously remember that, but it could make him more likely to think of it again later. In other words he talked about it because of the ads, not the other way around.
Sounds convincing!
...unless there were actually several thousand people who performed this experiment, got a negative result, and therefore don't remember it or post anything about it.
If true then it's easy enough to produce video evidence of this.
> Every time I read some technical description about why this isn't happening, the technical description seems convincing.
Having knowledge of the technical limitations and challenges myself, I used to be on board for this argument, but now less so.
All of the technical arguments against the listening seem to ignore the "Ok, <DEVICE>" or "Hello, <DEVICE>" initiating phrases for the voluntary surveillance devices people put in their rooms, and offer only a worst case defense ~"how could they process everything everyone is saying?!"
Why is it such a stretch to imagine these devices grab Direct Objects and Subjects and store those singular items for ad keywording?
We have cookies and know how they work, why is it difficult to extrapolate?
simonw is a breathless proselytizer of LLMs and likely is suffering from "a man's salary depending on misunderstanding" and all that.
It bears repeating, "these corpos are raising billions and hiring former alphabet heads to their boards for reasons other than just making you a better programming assistant."
> Why is it such a stretch to imagine these devices grab Direct Objects and Subjects and store those singular items for ad keywording?
How about because Apple say they don't do that, and can and do get sued if they say things like that which are not true?
(Sadly I make basically no money at all from my "breathless proselytizing" of LLMs. I hope to fix that this year, someone should pay me for this stuff!
You know I've written more negative things about LLMs than almost anyone else, right? 121 posts tagged AI ethics right here: https://simonwillison.net/tags/ai+ethics/ )
1 reply →
> the technical description seems convincing.
Isn't the technical hurdle just changing the "Alexa" or "Siri" wake word into a keyword for an ad campaign?
Siri doesn't let you do that, it doesn't give you audio it does pick up, and it doesn't even listen for the wake word with the main CPU.
1 reply →
A family member who seems intelligent and sane occasionally tells me stories of her experiences with ghosts.
...and yet no one seems to ever be able to replicate these results.
My experience working at one of the companies that gets accused of this a lot is that many colleagues wish we were as evil as claimed because it would be so much easier do their jobs that way than struggling through the reality of it which is endless red tape over the tiniest issues that have even the slightest proximity to privacy. So I've been a bit skeptical too.
Exactly. The big companies are scared of lawsuits and trying to get approval for something like that would be a nonstarter. As a matter of fact the device folks at the same company would be working hard to kill such an idea in its infancy because it’s already an uphill battle to sell always-listening or always-watching devices to consumers because of the creepiness factor.
And people also are terrible at math. Modern ML (regression & neural nets) are ridiculously good at predicting stuff you might be interested in, particularly when rich data sources like browsing and e-commerce histories are available; the decision to show the ad to you at some point almost certainly was made long before any audio-to-marketing pipeline could act on it.
Aren’t TikTok and Huawei easy counterpoints to this?
1 reply →
After shadow profiles, cambridge analytica, prism, etc, I don't think those companies are all that scared of privacy violation lawsuits.
Yup - having worked at Google Display Ads (arguably the epicenter of such talk), I personally only ever witnessed people walking the walk, privacy-wise. The threats to our privacy are quite public and not at all illegal; IMO data brokers and 3P browser trackers are at the top of the list, but all of Google’s known ills are there too (location tracking, exchange monopolization, allowing predatory advertisers, gestures broadly at chrome, etc etc etc).
They don’t need to be listening to us, and wouldn’t know how to even begin hiding it if they were. Something like that would require tons of compute and thousands of conspirators risking massive backlash, all to prop up a relatively tiny part of their business.
> Convincing people of this is basically impossible
Absolutely correct IME, btw. This is one of those things a smart engineer learns not to argue online, or at the Christmas dinner table for that matter. People tend to stand their ground on this one and move quickly to accusations of bias and naïveté…
I'm not trying to change your mind, but this response (from another user) was flagged, so I'm providing a pull quote.
> A marketing firm called Cox Media Group has recently revealed that it is listening to user conversations via their smartphones through its so-called "Active Listening" Software. With this, the company will push advertisements that users will see on certain platforms based on the heard conversations as unveiled by a report.
https://www.techtimes.com/articles/307372/20240904/cox-media...
Is techtimes.com junk?
14 replies →
Well said. I've worked in adtech and this aligns with my experience. Alphabet probably wouldn't even make that much more money compared to its current ad program. There's no shortage of supply in display advertising.
1 reply →
[dead]
Citation for Apple?
1 reply →
They don't need to: they can read your mind.
My proof? Yesterday I was driving home and I saw an old Toyota Previa minivan and thought to myself "Oh, a Previa, you don't see those very often these days." When I got home, I started scrolling through my Google News and it showed me an advert for the new Previa.
I agree with Simon: you basically aren't going to convince someone that their phone doesn't listen to you and serve you adverts based on it, because they've run into instances where it seems like it.
I think people are prone to believing that their phone is listening to them out of an instinct to believe they are a lot more unique than they actually are. A clash between western individualism and hyper-efficient consumerism.
That's not me taking a stance on whether it's true, though. There would be a lot of fame in it for a whistleblower, but on the other hand if Google or Meta figured out a way to do it with a low chance of it being proven, why wouldn't they?
Seriously, a whistleblower in tech? In 2025?
“Be part of the richest portion of the middle class and never have to worry about money again orrrrr mysteriously die 6 months from now leaving no identifiable impact”
> out of an instinct to believe they are a lot more unique than they actually are.
This is trivializing people in my opinion. The non-trivializing interpretation would be that for whatever reason people are skeptical that a black box that they don't understand very well, don't know how to audit, and don't know how to exert low level control over is doing things that they don't want it to do.
When framed that way it is immediately clear that this is an incredibly reasonable stance to take. The default assumption should always be that a third party who has a vested interest is pursuing it.
As an example. It is paranoia to assume that a 1970s era vehicle with almost no electronics in it is reporting on me to the manufacturer. It is willful ignorance bordering on delusion to assume that a vehicle manufactured in 2020 is not.
I think you're right on the money here. Most people are surprisingly predictable, and if your algorithm based on whatever works for at least 30% of people at least once a year, which is very low bar if you think about it, the population at large will notice
This is due to ad blindness.
The Previa ads were always there but that was the first time you noticed it because it was on your mind.
Alternatively, the Previas were always there, and they developed the memory of seeing it after having been exposed to the ad.
(To be clear, I'm not being facetious -- this is a viable possibility)
Baader-meinhof phenomenon
Had something similar happen to me. I wanted a mid sized ladder. Didn't search for it or even tell anyone that I was looking for a ladder but suddenly started getting a bunch of ads for ladders. Eventually I figured out that when I went hiking a few days earlier, I forgot to turn off my Fitbit after I got done. On the way home, I stopped at Walmart and looked at the ladders they had. I could see on my Fitbit account my path through the state park and through the Walmart, including the section with the ladders that I dwelled in front of for a while. That was enough to trigger some set of parameters somewhere and get me ladder advertisements.
> suddenly started getting a bunch of ads for ladders
The trouble is, what does "suddenly" and "a bunch of" mean? This doesn't sound very specific, let alone scientifically rigorous. How many ladder ads were you getting previously, and how would you know? Exactly when did the rate of such ads increase? Unless you have these details, it's highly possible you're estimating them inaccurately.
Or maybe there is a hidden variable that:
(A) Explains why you noticed the Privia. (B) Explains why you searched for something that might have triggered a Privia ad.
To a layman this might be indistinguishable from magic. So in some way maybe the phone really can read your mind :)
The hidden variable is the previa marketing budget. They have budget right now for a billboard and for online ads at the same time and they are focusing on your geography
They don't need to read your mind. The systems that push ads to you are also owned by the companies that are pushing contents and shaping how you see the internet. It's like how magicians incept ideas into your head then use it like they can read your mind.
Partially mind reading, but mostly mind control. You didn't think of the Previa out of nowhere. They led you to that thought.
For more consumption.
Try reading Qualityland, by Marc-Uwe Kling. It's SF (funny SF) about this exact premise.
One thing that is rarely discussed in relation to this urban myth is that it is widely believe and yet it doesn't actually effect the behavior of most believers. That is one of the major reasons why most companies and many governments don't care about privacy, the public doesn't really care about it either.
People will of course choose privacy over no privacy with all else being equal, but privacy is the first thing sacrificed when push comes to shove. If the average person is given the choice of having everything said within earshot of their phone being recorded and sent to Facebook or giving up Instagram, they'll happily choose Instagram and forsake their privacy.
If privacy advocates want to start turning the tide in this battle, the first step needs to be convincing the average person why privacy is important on a personal and tangible level. No arguments about future totalitarian regimes or hypothetical ideals. Abstract concepts like that rarely motivate people who have so many more practical political concerns. It needs to be something that is more important to people than having access to Instagram. And I have absolutely no idea how that could be accomplished which makes me concerned that we're already too late.
That's the main reason I care about this so much.
If people really do believe that their phones are spying on them all the time to show them ads that means that people are basically surrendering to an imagined surveillance state. They shut up and accept it, because they'd rather keep Facebook/Instagram installed than fight back.
I find that really depressing. I want people to have more agency than that.
We need people to understand the imagined v.s. the actual privacy threats, so they can push for better standards. If they believe in and submit to the conspiracy theories good luck getting anyone to campaign for actual meaningful improvements to the problems that are real.
If you believe that google is recording all your audio, uninstalling instagram is not going to cut it. I think such a person would have to go back down to a really dumb phone to have any confidence at that point.
People have only vibes, they think that if they paid with cash it would proobably be more private than a credit card, but what data is being sold and to whom for what uses? Is that even the case or are there regulations? If I constantly make cash withdrawals at the bank am I actually inviting extra scrutiny by looking like a money launderer? If I install this browser add-on maybe it sells all my data. But I'm also using chrome literally made by the ad company, and that youtuber told me if I don't use a VPN I'm constantly being tracked anyways...
If you just have a giant morass of confusing information about every digital decision, and a lot of annoying first steps you would take are likely to be no-ops, you just don't engage. People are defeated by ambiguity and lack of attention span, same reason lobbying works and people were constantly being poisoned by food & drug additives before the modern era.
That's a very good reason to talk about this.
I do think that this is a very "lightly held" belief. It's something people kind-of believe, they'll tell this to each other, but it doesn't affect any behavior - not because it's not important IMO, but because people mostly don't really, deep down, believe it.
And I do wonder if convincing people this isn't happening will have the opposite effect than we intend. Instead of being more aware of what actual privacy violations are, it'll just make people write off the whole idea of companies invading their privacy. Idk.
In my experience the people who believe conspiracy theories like this tend to have a fatalistic attitude towards them. They're not making any specific claims about specific apps listening in on their conversations. Rather it's more a vague sense that "they" (whoever "they" are) are always listening and that there's nothing that can be done about it without giving up on modern technology entirely.
The people I know who believe this myth go into their settings and turn off microphone access on an app-by-app basis.
I went back to my family for this christmas, and argued about privacy, I think it was about free email services again, and stuff likes this. And again - I just cannot understand it - the response was:
Yeah but I don't care, I have nothing to hide, let them have my data.
It's a slap in the face of me, trying to meticulously remove all internet access of programs and devices that don't need it, and moving from all free and not-privacy-friendly services to mostly paid and private-friendly ones - probably "losing" a big chunk of my lifetime doing this. I feel paranoid sometimes, when I hear this argument.
You are not losing your lifetime, you are building the future. If other people do not want to be on board it is a loss. I have spent years trying to teach people how to build their own websites and get away from the large walled gardens. Many will not come on board, but the community that does do it have built truly wonderful things. I think of the Neal Stepehson article "In the Beginning was the Command Line" where he compares Microsloths Family Sedan to the Free Tanks of Linux; there are all these Linux geeks yelling "come over here, we have free tanks" yet most people pay for the Family Sedan.
It sucks, yet I drive a Free Tank, and now help people online take care of theirs. The other day I was talking with one from the future generation about what type of computer they would want and they said "A Fedora gaming desktop". And that made it all alright :)
I know someone who worked on smart TV software. They explicitly added audio fingerprinting via integrated microphone so they could determine which programs you were watching for ad profiling.
When in the history of online ads have advertisers not used available data?
Automatic content recognition is one of the freakiest kinds of tracking I've heard of. Smart TVs become data collection hubs for the home if they get connected to the internet. Some have microphones for voice recognition. If you go into the privacy policies for them, you'll find that your audio gets sent off to third parties (albeit only when in use).
I completely disconnected my TV from the internet, but it still prompts me to connect it to Wi-Fi, to agree to policies, etc. This page is full of horribles: https://www.consumerreports.org/electronics/privacy/how-to-t...
Ha, I’m imagining a sort of thriller movie villain reveal where people are intently saying silly things into their phones, while the camera slowly pans over to the smart TV in the corner, which is actually listening.
I don't get this, why would they use a microphone when they already have the original quality audio stream going to the speakers?
They don’t always have the audio stream, e.g. when someone is using an audio receiver.
I had this happen to me recently. I was speaking to a friend about a subject that suddenly started to appear in ads after I hung up. It was spooky. But I think it has more to do with the activities related to the people we know. I'm sure because of Android, Gmail and search they have a good understanding of our general association group. I bet my friend did a search on what we were talking about and Google was able to determine that others in the group would be interested in the subject. They don't need the microphone input. They have better (more direct?) ways to get the information.
What you described is basically how it works. You all get finger printed on the same network, finger printed on google searches, finger printed at the routers, and then associated. 6 guys get together for beer and someone brings up wood chips, 1 guy later googles BBQs, you're all getting BBQ ads for the next week and a half, the 5 other guys think the phone is listening to them because they don't know one of the guys actually googled or triggered something related to BBQs. I once knew my roommate was pregnant before anyone else because I got a bunch of ads for baby shit all of a sudden.
I actually believe the alternative theories are scarier than the phone actually listening to you. The one thing I keep hearing over and over again from people is that “I talked to so-and-so about this“ and now I’m getting a bunch of ads about it. What’s scary is that it knew you two were in the same room that one of you had a recent interest and that there was a likelihood you discussed it. That seems worse to me than just listening the conversations.
>That seems worse to me than just listening the conversations.
And that's the one that's actually happening. Listening to your conversations isn't technically possible and is easy to disprove.
If this parasitic nonsense technology ever turns against us Stasi-style we're gonna wish the ads back.
> ...they would need to be feeding those snippets in almost real time into a system which forwards them onto advertising partners who then feed that information into targeting networks...
I think this is totally mistaken. An ad seller which also wants to respect privacy keeps this data in-house and does the ad targeting themselves. The advertiser never needs to see personal information for this kind of market to give people ads related to overheard conversations.
We're specifically talking about Apple here, based on Siri wake words. Do you think Apple are running this kind ad of targeting?
My post is limited to pointing out what I think is a flaw in the reasoning. I am not taking a stance on the larger question.
why are you limiting your argument to wake words? It's comparatively trivial to serve ads based upon what's spoken after the wake word.
And as we know from the lawsuit, it seems that there's been a lot of data gathered accidentally too.
12 replies →
Also, couldn't a lot of it be done on device? Say, have a set of key phrases to listen for, and if the device hears them, then mark that the user should be targeted by certain kinds of ads.
I don't have any evidence either way that that happens, but it seems like a more practical way to accomplish it.
I mean, many advertisers go through the Google display network without having Google target their ads (AKA run their bidding algorithms). Typically they go through other middlemen agencies, who indeed would need at least some of this data to make use of it, tho perhaps in some derivative form.
Still, you’re right that Google could be keeping it all for themselves and feeding it to their black box targeting services. I really strongly doubt that’s happening with incidental assistant snippets much less intentionally-eavesdropped recordings, but it is more plausible than this makes it seem.
The reason I think this isn't happening (other than they have told us it's not happening and technically seems very difficult to do this long without researchers/people figuring it out), is that in my life I've spent millions and millions of dollars on ad budget, we spent so much with twitter they made us custom hoodie, nobody tried to sell this function to me, it's not an enterprise level option, I've seen those, and it's not a button in the ad platform, so how is the key word getting into an ad network getting attributed to my ad profile and my ad served? There is a lot of rich ad enhancement you can buy at the enterprise level, but I've never seen anything like that mentioned even remotely. There are A LOT of ways I can target you, some of them pretty creepy, but using convo snippets from the phone? I can't even think how that would work.
Or, for that matter: if this functionality exists in phones, surely there were some people involved in developing the software and infrastructure to support it. It seems hard to imagine that none of them would have spoken up, even anonymously, to confirm its existence.
I dunno, seems trivial to me.
Like if an arduino can reliably carry 1 wakeword, an iphone could carry what? 64 at minimum?
Instead of waking up recording the conversation and sending it to the stasi, it could just toggle a bool in your secret advertising profile, then sync that up at random intervals.
They dont need context, they just need "Biscuit" "Nappies" "Birthday".
Not to say it isn't all made up. It probably is. I just dont think its technically difficult to achieve. It is probably all psychology. But sitting there with a packet capture going as you talk about dogfood isnt necessarily a great test to confirm the negative.
But if you're listening for 64, or even 1024 wake words, you won't get to the level of granularity people are claiming to observe, like in the other comment here where they tested "pool fencing". If you were putting together a list of 64 words, would you really include "pool fencing"?
The wakewords could be simply "buy", "cheap", "sale", "price" etc - and then collect the exact words around it. I bet the pool fencing conversation followed with the exchange of ideas where to get equipment.
1 reply →
Its an extreme low side estimate.
And the chance that Pool OR Fencing is one is pretty high.
1 reply →
There are certainly ways to make it more viable, but I hope that the systems up to billions of people use are sufficiently independently audited so that such features would quickly be found.
We search for everything we talk about these days…our location services are always on. The times I’ve noticed this I can usually trace it back to a friend had been searching for something and then we (our phones) were in the same place. Even the guy from the lawsuit with the doctor… his phone knows he was at that doctor and that doctor offers that procedure. Seems possible without listening.
Location being on in general doesn't mean the search query has access to it. Safari makes you approve each individual location access by a website.
Yes, but I think that's the part that feels... unsettling.
I watched a Nigerian film on the seat back display on the plane the other day. I looked it up on my laptop when I got home because I didn't finish it.
The next day in my Instagram explore page I had a Nigerian meme...
Yes, I get it -- could be frequency illusion or some IP address/cookie shenanigans. Still feels weird.
Or more likely: Confirmation bias. I also occasionally get Nigerian memes on IG despite never having searched it. But unlike you I don't have the Nigerian film google search history that would lead me to conclude it was targeted.
Put another way, if you googled a Swedish film then saw Swedish memes you would think that was targeted. Then upon viewing the Nigerian memes you would have attributed it to a random video rather than anything targeted.
“ I looked it up on my laptop when I got home”
That is the real mechanism for triggering the ads and memes, not that “your phone is listening to you.”
At least feel creepy about the real cause and not some imagined monster under the bed.
I keep location services off, but realize that’s probably not completely hiding my location.
This is gaslighting. Even if it’s a coincidence, people have numerous examples of situations where they or anyone with them did not search for a thing. I’m not saying the only possible explanation is that the phone is spying on you - but I’m saying, don’t gaslight people and say “you probably searched about it and forgot”.
But what if their friend searched for it rather than them, and they were associated due to their location?
This doesn’t seem that hard to test? Put a phone on WiFi, run it through a traffic sniffer, and see if it is constantly sending little audio packets to listener1486.facebook.com or whatever.
Or heck, open up a phone and stick a probe on the mic lead. See when it is getting power or not.
There are smarter ways to do this without constantly sending all the audio. Wakewords + surrounding snippets for example. In fact if the audio recognition is done on-device then only the text data could be sent which would be indistinguishable from telemetry.
> 404 Media previously reported Cox Media Group (CMG) was advertising a service that claimed to target ads based on what potential customers said near device microphones. Now, here is the pitch deck CMG sent to prospective companies. Google has kicked CMG off its Partner Program in response.
https://www.404media.co/heres-the-pitch-deck-for-active-list...
The alternative is worse: more passive forms of targeting are so good that advertisers don't need to record your every moment in order to make a large number of people genuinely believe that they are.
It's relatively comforting to think they'd be so brazen as to care about your every word, as opposed to the fact that they own so much information about you that they can predict your actions and thoughts better than you can.
Reuters journalist Jon Stempel repeated the plaintiffs' paranoid and unsubstantiated allegations but did not include a single statement from Apple, just a vacant "did not immediately respond". If only there were years of statements made under oath in the courts to draw from.
Siri uses a pseudonymous identifier when communicating with Apple's services. The identifier is not linked to your Apple ID. Therefore Siri does not have access to your iCloud data. When you ask Siri to "call mom", it constructs a search query which is then executed on-device against your contacts database.
More of these requests are served entirely on-device. Just ask: how is that consistent with the idea that they are doing ad targeting based on Siri requests?
You can request a data export for your Apple account and see that there's no Siri data included in it. If they are caught lying, EU regulators will have a field day.
Apple describes their ad targeting here. They list the—frankly boring—signals they use for ad personalization, which does not include anything you say to Siri or have in your personal iCloud data. https://www.apple.com/legal/privacy/data/en/apple-advertisin...
No one is apologizing for Apple's nontransparent collection of Siri audio for QA purposes, which is what they settled over and made opt-in.
Neither do I. There’s no technical way it can be happening without it being easily spotted, not to mention how many would have to conspire.
However I am still waiting for a comprehensive explanation for how it is actually happening.
There there are many ways to correlate people and their interests, but I wonder how deep it goes. Basic geolocation and public interaction metadata: Sure. But I’ve also heard people believe that it ought to be possible to spot closeness via Wifi and Bluetooth.
This is such a culturally relevant topic that there must be some serious knowledge about it somewhere.
As for my own tests, I’ve made it a point to mention that I’m considering spending money on a well advertised-for product that I haven’t yet searched for online every time this topic comes up. No ads so far. In the next phase I’ll progressively tell more people to start looking it up.
Well clearly HN is listening to my phone because I literally had this exact debate several times last week!
The argument I use to convince people is to ask them how long your battery lasts when you’re running any apps, I’ll then ask how long do you think your battery would last if the apps were listening all day long.
I can’t tell if my friends are convinced or if they’re too polite (or disinterested) to argue.
How would you tell the difference if the apps are listening all the time without you knowing it?
It's hard to tell I can imagine some motivated individuals could utilize all sorts of packaging systems and embed them in third-party applications and so on, and extract pertinent information using this type of surveillance, and then sell this data to data brokers which would sell it to the large ad networks. I mean there's lots of ways to transcribe even most of the whisper models can run all the way down to 150 megabyte file not to mention the quantization versions of these models. I have something that I run on my computer for my server not throughout my house that does real Time transcription and whatnot but I use it for my own purposes, so you know someone who makes money off advertising or even selling insights about people, would certainly find ways to do this. I mean it's simply not regulated is it?
https://huggingface.co/spaces/jilangdi/whisper-web
I used to not believe in confirmation bias, but ever since I started really paying attention to it, it’s suddenly everywhere!
As of 2019 (my last contact with the industry), there was no real-time onboarding solution that could do this. Or if it existed it wasn't good enough to attract any attention. People were saying this in 2014 or so. I remember around 2017 thinking it would be pretty cool (repurpose wake-word detection to do some pre-determined segments or something) but no-one had it working. There's all sorts of stuff going on behind the scenes in ad tech and if https://news.ycombinator.com/threads?id=manigandham would probably have something useful to share if he was still around.
But it wasn't real. Even back then people would publish "sources" saying it was being done, but I'm telling you that if it worked well people would be doing it and we were in the middle of it all and couldn't find anyone who was.
There were smart-devices that acted as beacons to report people to ICE and stuff like that and they used personal data to tie it together. So there was crazy stuff out there (none tied to the respectable ad tech industry HN knows as the "privacy-violators" and so on). But this specific thing wasn't there.
One option is that people you talk to look for the stuff to discuss.
E.g. person A talks about bouldering with person B. Person A then continues with their day while person B googles bouldering terms.
A and B then receive both bouldering ads. Person By because of the googling and person A because of proximity to person B.
Person A then assumes the phone spies what they say, but it doesn't. Location is enough.
Ask HN - let's say hypothetically, our phones are as bad as some think. Face camera taking covert pics, regular screenshots sent back home, microphone recording every conversation, every message, email, and bit of internet traffic scraped and harvested for ads and anything else - how would you change your habits if you knew this was happening?
Answer - you wouldn't.
I'd write to my representatives about it.
Have you? I can't be bothered to cite but all of the above can and does happen in certain contexts - the covert face camera is admittedly a stretch, but all your personal photos are usually whisked away to the cloud without warning, which is just as bad (see the recent apple photos news)
1 reply →
The danger with this kind of opinion is that you start out doing some healthy critical thinking, and research. You form an opinion, based on good intentions. Then you spend all your time on explaining your conclusion to others. At that time, people who have the same opinion flock to you, and you don't have time to do more research.
This is the problem with experts and politicians. You can make money on either position, but once you have started stating your opinion, it's unlikely you'll do active research to disprove your opinion, as many have already concluded [1].
I'm happy that Simon only sees this as a hobby.
[1] https://quoteinvestigator.com/2017/11/30/salary/
Related:
Apple will pay $95M to settle Siri privacy lawsuit
https://news.ycombinator.com/item?id=42579009
Has he read this: https://www.404media.co/mindsift-brags-about-using-smart-dev...
Have you read the early stories and discussions that make that look unlikely?
https://arstechnica.com/gadgets/2023/12/no-a-marketing-firm-...
I think the article would be improved if he added a bit about what evidence he would accept.
Show me some sniffed network traffic demonstrating an app sending that data back to a server.
Discovery in a lawsuit.
Credible insider leaks confirming this happens and how.
How about a pitch deck for a company [1] claiming it can offer microphone eavesdropping based ad targeting [2]? Maybe that was still aspirational though.
TVs and cars have microphones now and are privacy nightmares. Car companies have patents [3] for in-car keyword based ad targeting. Without legalisation, it's really only a matter of time.
[1] https://www.fastcompany.com/90999277/cox-cmg-active-listenin...
https://www.404media.co/heres-the-pitch-deck-for-active-list...
[2] https://www.documentcloud.org/documents/24224884-how-voice-d...
[3] https://www.pcmag.com/news/ford-patents-in-car-advertising-s...
3 replies →
Cox Media Group selling a product that uses captured audio to target ads https://www.404media.co/heres-the-pitch-deck-for-active-list...
1 reply →
How about Apple paying 95 million dollars in fines for it? https://edition.cnn.com/2025/01/02/business/apple-siri-priva...
5 replies →
Here's what I would accept:
Go to a public library without any electronic devices and pick up a print copy of a magazine with some ranked list of products or brands. Select a handful that you don't think would be targeted towards you. For example, beauty products, childcare products, cruises marketed towards seniors, etc.
Over the course of a few weeks, use Siri to send text messages mentioning these brands, add them to your Reminders list, etc. Just don't allow it to redirect you to any apps or web searches, since that would invalidate the experiment.
Browse Apple News daily, which uses Apple's internal ad service. If you see any ads related to at least 2 of your canary brands, you have plausible evidence that your Siri interactions have been used for ad targeting.
[dead]
From a potential culpability or liability perspective, the question of targeted advertising is irrelevant. Intercepting private communications without consent, absent an applicable exception, is still against the law, e.g., the Wiretap Act and/or the California Invasion of Proivacy Act, ads or no ads. Targeted ads would not be used as evidence that private communications have been intercepted.
Many years ago, Google was sued for wiretapping and lost. See Joffe v Google
More recently Apple was sued for eavesdropping without consent. They just settled for $95 million. See Lopez v Apple
My wife has a very new (<2yo) iPhone. We live in public housing, no luxuries. We were talking about getting her some slippes, and I made a joke about her not getting any Gucci slippers. The Gucci ad appeared on her phone within 30 minutes.
I sh_t you not. She was a bit gobsmacked. I was neither surprised nor disturbed.
That's real anecdata, my friends. It could have been coincidence, and while we have nothing to hide, it's not an optimal situation, whatever the cause.
Besides, we ain't buying no Gucci slippers, no way, no how.
>We live in public housing, no luxuries.
>Gucci
Gucci is classified as an "aspirational luxury" brand. In other words, luxury for poor people. It makes total sense that it would be pushed to people in public housing. If you get recommendations for European brands you've never heard of selling $5000 t-shirts, then I'd be worried.
I'm not worried, and thanks for suggesting that advertising would be a viable motive given our geolocation. Good thing Apple doesn't also have means and opportunity, too. Silly me!
I guess we should all just blindly trust Apple. Thanks, friend.
1 reply →
What social media apps does she have installed? Just curious
None.
My wife and I, just being goofy, have an inside joke featuring a relatively uncommon name that we ocasionally yell out, like a few times a week. Two months ago we got junk mail addressed to UNCOMMON NAME + our last name.
I don’t think its a coincidence. Something is listening. Its kind of messed up that I am a normally rational, skeptical minded person, fairly knowledgable about information security and in 2024 I can’t even draw any clear lines between a nutty conspiracy theory and reality.
I’m sorry but I don’t find this article basically just saying “yeah but what are the chances Apple would do that?” persuasive at all.
I think it's normal to have a hard time delineating between conspiracy and genuine concern when it comes to this. We're literally carrying around powerful computers with exceptionally reliable connectivity combined with high-resolution cameras and high-fidelity microphones that are frequently used for voice recognition. Not to mention that for 50% of US smartphones, the software is designed by the most prolific advertiser in the world.
I had an experience a few years ago where I had talked about a fairly niche product (I can't recall exactly what it was) and the next day I started seeing ads for that product all over the place. I commented about it to two of my coworkers that day, how I had been skeptical about the conspiracy that our phones were listening to us for marketing purposes but that this felt eerie. What shocked me was their response: They had seen the same ad all over the place. Since then I've had a hard time deciding what to be more concerned about: That my phone might be listening, or that I might have been subtly influenced into thinking about this thing; that my whole experience was actually a result of being susceptible to marketing directed at my demographic
>My wife and I, just being goofy, have an inside joke featuring a relatively uncommon name that we ocasionally yell out, like a few times a week. Two months ago we got junk mail addressed to UNCOMMON NAME + our last name.
Are you sure it's not because of an opsec fail? eg. you used your nickname when registering for some service, and that made it into some sort of mailing list? What you're seemingly implying (ie. that there's some sort of secret listening system that can figure out your nickname, tie it back to your address, and then send spam to that) makes little sense. Your name + address is already readily available. It's in the public records. You freely hand it over to random websites (eg. for online shopping). There's zero benefits in anyone making such a system to figure out people's names using surreptitious listening.
I am open to other explanations. I did not use the name as an online pseudonym, thats why I found it so odd.
I am not sure about there being no benefit either. We moved here very recently and scraping local property records would take time and not be easily automated. So what if some data aggregator still had a blank under the name field for our address and needed to fill it in so they could address letters since we more or less automatically throw away “current resident” letters. I just don’t know, yes its far fetched but I don’t see any other explanations as much more parsimonious.
[dead]
I suspect that it is emergent behavior that even the ad companies themselves aren't aware of.
I have definitely experienced it myself, and have no doubts that it happens.
There is something I've always been wondering about companies' privacy policy. When Google says that it does not send user audio data to their server, does that prevent them from sending say whisper embedding of the audio itself? Because technically it would not be audio. It would still include a very rich representation of what the user said. Rich enough to be used to sell things.
I find it completely plausible that either Google, Facebook or some other installed app is listening to your conversations to develop marketing profiles.
What I find amazing is that people don't use ad blockers! I wouldn't be able to tell you if they are harvesting my voice data, because I haven't seen an ad in years! It is trivial to block ads, why do so many people choose to see them?
What happens is that you in fact lend part of your brain to all the data that you see on your screen. You don't realize it but your subconscious mind sees senses a lot of stuff your conscious mind is not aware about. Until you see it after you just talked about it thinking it was your idea and panic.
What you actually talk about with people (excluding maybe the proffesional stuff) is mundane stuff you see on your screens, newspapers, billboards, etc.
people surf the net so automatically that they forget they are doing it and blame Alexa for snooping.
I do recall reading the Alex engineers insisting on a physical off switch while the opposite was true for the Google Home product.
I think it’s wild that your counter argument is basically just gaslighting everyone. There are tons of (possibly coincidental) examples of this type of thing happening. Even if it’s a coincidence, it’s not explained away by “maybe everyone just forgot or is not paying attention”.
It's not gaslighting. I believe people see the ads for things they talk about in the same room as a smart speaker.
I also believe people very rarely only speak about a thing -- without there being any digital footprint at all of them engaging with it.
I have literally seen this play out with family and traced it back to make a point.
The assumption is that it's either widespread or nonexistent, but I could see it being just a rare instance that gets A/B tested, and maybe by now the companies have decided it's not worth it. But IME all kinds of not quite kosher stuff gets A/B tested, and that would line up with anecdotes where it's too specific to be coincidence, but also uncommon and only affecting certain people.
I have successfully argued this for at least a certain class of devices. Namely, it practically can't happen _for small devices_. e.g., a TV remote with a voice feature activated by a button simply won't have enough _power_ to constantly be listening, uploading, or processing.
Harder to convince folks that _nothing_ does this, but takes an edge off their more paranoid tendencies.
I personally don't believe any argument that says it's not feasible.
Transcribing all spoken text and sending it home, sure, not feasible.
What if we have 256 keywords, or 65536 keywords, maybe preconfigured for particular products or product classes. Some basic linear predictive coding mechanism ( you know, what powered those '80s chips Stephen Hawking style, speak and spell, etc) - very very low computational overhead. When the word is triggered, queue a message back home at the next reasonable opportunity - user id, timestamp, word. It will only take a couple bytes. It can be slipped in anywhere and obfuscated by any means by nature of being so small, data-wise, even as a watermark of some sort. By using a timestamp and waiting until the next opportunity, maybe minutes, hours, or days away, no time correlation detection is possible either.
People say big tech is ethical, fine. Maybe some ad company is sponsoring some free app or game for the phone, and slipping this in there. Now the developer can pay their rent and food costs. Maybe the ad company is then selling that data back to big tech who washes their hands of any wrongdoing. Maybe it's all legal because the fine print of the EULA allows for this.
Seems to me though this can be figured out empirically, just have a voice play something like "need to buy adult diapers" or "new tires" etc next to a device, enumerate every device, look for ads on whatever very specific topic, minding along the way to tell nobody and never enter it in any internet-connected keyboard.
For sure! I mean, my Google pixel has "Now Playing" which is able to passively listen to the microphone for songs it knows, and displays them on the oled lock screen.
So, we already know that it is
A. completely feasible for a smartphone to do this.
B. At least a subset of smartphones have always-on microphones.
Maybe not a remote control... But why would you put it in a remote anyways when everyone has a phone?
1 reply →
What if the activation of the microphone for marketing purposes occurred not systematically but randomly, making it harder to detect such functionality? And what if this functionality were activated with code loaded remotely at the moment it needs to be used?
I've noticed people often research products, services, medical issues/ailments, etc. using Google (or similar). When they start seeing related ads on social media, they tend to attribute more to having discussed the topic out loud, forgetting about the searches they made from the same device or IP address.
Another option not mentioned is that the thing they’re talking about might have been in their subconscious because of an ad they’d recently seen but didn’t notice. The topic/concept/product eventually bubbles up to consciousness, they talk about it, then start noticing the ads that were already there.
Maybe the algorithms simply concluded that you might interested in things your friends are interested in.. that's part of why you're friends?
If any of your friends searched for something online, and you're connected to them in a social network, they might show you the same ad by proximity.
listen to me when i say this. we will start killing if we don't have same access to that tech. why kill monster? well I'm dead anyway as a slave to their machine.
people better demand equal access or start slitting throats right now.
The post's arguments can basically be summed up in three categories:
1) "Doing this wouldn't be technically feasible or would require a technical effort wildly out of proportion"
2) "There are lots of psychological biases that lead people to believe something like this happened even if it didn't actually happen"
3) "Apple is such a nice and honest company, they would never do such a thing..."
As for 1), there is enough technical discussion in this thread to disprove that point. But just as a reminder: Google build an always-on song recognition service into android, free of charge, without any obvious monetization, just because they can. OpenAI released Whisper last year as open source, a highly precise audio transcription model. By now lots of variants for on-device use exist.
All that tech doesn't just exist, it's not even seen as a moat. It's already being commodified.
As for 2), yes of course cognitive biases are always a thing. The problem is that you cannot use them to disprove something. They constitute an absence of evidence, not evidence of absence.
As for 3), yeah no comment here. Except maybe, remember Snowden. "No one would do such a thing" has already been spectacularly wrong in the past.
Did that tech exist prior to 2017 when this conspiracy theory was already widely circulating?
I honestly couldn't care less if that tech existed back then or not, I care about what's possible today.
Between 2017 and today, there were massive changes, both in technological development but also in mindset: Surveillance capitalism became much more normalized and generally accepted as a standard part of business.
So if you argue that it was a baseless conspiracy theory back then, therefore it automatically must still be a baseless conspiracy theory today, that argument is flawed.
1 reply →
The author admits the fine for breaking this rule is basically a non-issue compared to the potential earnings. Why wouldn’t the most valuable company on earth see this as a huge, easy win to record and advertise based on those recordings?
Under GDPR don't they risk 10% of worldwide revenue? (Plus for a company like apple trying to be seen as the defender of privacy, the PR and reputational hit)
OK, but query me this batman. Lately, I have been inundated by ads for anything to do with Kayaks. I do not own a Kayak. I do not want to own a Kayak and have never even spoken the word as far as I can remember. But for some reason, I am seeing ads for a Kayak carrier, a Kayak launcher, roof rack, you name it. It is so weird. I cannot figure out why suddenly these idiots think I am remotely interested in Kayaking. Kayaking, it is just so random. But way back in the 2000s, a friend of mine worked for a web site dev company. Her job was to visit with clients and take pictures of whatever they were hawking, write some copy and the dudes back in her office would turn it into a web page. She would occasionally ask me to come along and stare at whatever crap they were selling and act like I cared about it while she snapped some photos. I joked that meant technically I was a model. But then one time she was doing a web page for a marketing company that did focus groups. That was back before we had smart phones tracking your every move. I posed in a fake focus group with some other people and after that, the dude that ran the place asked me if I wanted to be in real focus group and I agreed. They seemed to think there were certain connections that made no sense. For example, they were sure that they could market adult diapers to nerds that play video games who were to absorbed in their game to get off their ass to take a simple dump. Apparently that is actually a thing? Then I went to this one focus group where they were asking if you own a dog and buy a lot of peanut butter. Read between the lines. So gross. So, I am guessing, some marketing weirdo thinks there is some weird connection between something else in my life an Kayaks. It would be interesting to know what that is. But, I think, a lot of times, they might be right about those weird connections that make you feel like they are reading your mind. Just in this case, whatever it is, maybe that I watch a lot of documentaries and that means I like Kayaking? I guess they only have to be right some percentage of the time. Just not in this case.
Depending on your age and if you have any vaguely outdoorsy interests, you probably are due for a kayak interest. A bunch of guys I know in the late 30s early 40s are into kayaks right now. Kayaks and smoking meat.
> So, I am guessing, some marketing weirdo thinks there is some weird connection between something else in my life an Kayaks.
That would explain the intensity, but I’d still be surprised to see any patterns not found by ML/statistics.
It seems to me that somebody in your life has gotten into Kayaks. Take it as a challenge to find out who it is.
someone probably googled “kayak” from your wifi?
Hmmm, I live alone and in a pretty remote area (on an island). I keep my wifi pretty secure, so, that seems pretty unlikely. But who knows, anything is possible I suppose. Maybe because I live on an island.
3 replies →
Or they googled fitness or outdoors and live near water
Edit: yep
2 replies →
It's because the big companies have the simulation's/global consciousness' hacked endpoints as uplinks into their infrastructure. lol
> As one report put it, “Amazon can acquire more comprehensive data on people’s living habits....” The company wants to sell real-world services such as house cleaning, plumbing, and restaurant delivery, but according to some insiders, the vision is more far-reaching: an omniscient voice that knows all experience and anticipates all action! Already, forward-looking Amazon patents include the development of a “voice-sniffer algorithm” integrated into any device and able to respond to hot words such as “bought,” “dislike,” or “love” with product and service offers.
The Age of Surveillance Capitalism (2018) - Zuboff, Shoshana
Given that this is form 2018, the rise of voice assistants since then, all the market forces at play and the reward highly outweighing regulatory penalty payments there is little that's supporting that users are not being recorded. The TOS phrasses like 'service improvement', 'may share with third parties', and 'user experience' are common place and are not particularly concise. Simply dismissing anekdotes and honing in on Apple Inc., the only company which does not want/need to sell data, is not convincing.
I think people tend to shoot the messenger, when it comes to privacy discussions. Getting told one is being spyed on by a divice, which is so intimt and close to one slef is embarassing. Of course things are more like a spectrum and differ from case to case.
You can think whatever you want. When ad companies claim they have that capability and when I see evidence with my own eyes that’s enough for me.
That is EXACTLY what I said in my article: "Convincing people of this is basically impossible. It doesn’t matter how good your argument is, if someone has ever seen an ad that relates to their previous voice conversation they are likely convinced and there’s nothing you can do to talk them out of it"
To all the people expressing doubt and disbelief, refer to every whistleblower story which was derided as tinfoil hat conspiracy until blown up. If we still haven't learnt that money will make corporations (and people) do despicable things, only we are to blame ourselves.
What I worry about sometimes is malware sniffing for hardware wallet seed phrases being practiced out loud by users when they think they're alone. I always tell people if they keep crypto on a hardware wallet - NEVER speak your seed phrase aloud. This could also be spied on by laser microphone surveillance.
I had a brand of wine in a closed, we were discussing with my life our wedding. Immediately this one brand of wine popped up on my wife user feed, wall. That happens from time to time, so, yeah... I know that it works like this for sure.
Many mobile apps use an adtech SDK which does on-device audio keyword spotting. The company making this SDK is super secretive, so not much information about it.
If "many mobile apps" use it surely some docs or other details have leaked out somewhere? Or at least a name?
I saw a whole PR release for adtech oriented keyword spotting tech like this one time, I wish I'd saved it... I can't remember the name of that particular company.
But read through this doc:
https://picovoice.ai/docs/faq/porcupine/
I won't disclose names.
But thinking from the first principles, do you really think that all this phonetic keyword spotting[1] IP developed for defense tech in the late 90s and early 2000s was abandoned once recording entire phone conversations and doing full speech to text on them became technologically possible?
I still remember lots of startups who did this stuff openly a decade ago, before Cambridge Analytica scandal[2]. After that it became impossible to get funding or get acquired, so the few that stayed in this field became very secretive.
—-
[1] https://en.m.wikipedia.org/wiki/Keyword_spotting
[2] https://en.m.wikipedia.org/wiki/Facebook%E2%80%93Cambridge_A...
3 replies →
Nice try, Meta.
oddly enough, I compare this conspiracy theory with clout-chasing word-of-god christians on social media: there's an element of "i don't understand how it works, therefore the simplest answer must be correct"
if everyday people knew just how much data companies have on them and their habits, they'd be absolutely horrified. There's so much personal attribution data that most companies that know how to market properly are going to get you at some point. You may think you're careful with what you share and what is being recorded, but if you are friends or relatives of someone who does, or if you live in the same street as someone who does, they data on you.
I used to work for these kind of companies, and have a lot of friends and family who currently do, and the stuff they track and the patterns they have never fail to surprise me and make me laugh.
Sure, if they could, they would absolutely listen in to you, that would make their jobs a hell of a lot easier, but they don't. But the data they have on you so goddamn accurate, it's easy to assume they must be.
They do. They absolutely do. Not a week goes by without a blatant case on my phone. I'd wager everything I own on this.
What evidence would you accept that they're not doing this in a way that would make this an actionable bet?
What evidence would anyone accept that 1+1 is 3? That's a genuine response. The cases are so clear cut (words picked up from conversations that would never fit on whatever profile they have on me; up to specific mentioned brands and things completely alien to be) and so numerous. I still can't believe that anyone even debates it. It's blatant as hell and all of my peers notice it, too
Theres a reason why they put a microphone in every single Internet connected device they can, even when it serves little purpose. My remote has a mic ffs, not to mention oddly specific ads after certain conversations.
I now believe it's happening based on this article.
We already know about TV piracy detection leading to bars showing soccer without a license being fined.
With what credibility does the author of this piece speak? Are they in all of the ad platforms on mobile/tablet?
I have nothing to do with ad tech, or selling mobile phones or anything like that.
I don't like seeing people fall for obvious conspiracy theories, and I have used my own deductive reasoning to decide that this is an untrue conspiracy theory.
[dead]
[dead]
[dead]
I set my iPhone language to Italian and now I’m getting non-stop ads for Olive Garden.
What is the correlation? Olive Garden has nothing to do with anything Italian.
I set mine to Spanish and now it's nothing but Taco Bell ads
[flagged]