Hacker News new | past | comments | ask | show | jobs | submit login
The Hidden-Pregnancy Experiment: Could I Hide My Pregnancy from My Phone? (newyorker.com)
73 points by fortran77 7 months ago | hide | past | favorite | 60 comments



As someone in marketing, I’ll point out that the technology has advanced quite a bit from the infamous Target case study.

The most powerful signals now are not demographics and search history, but location, transactions and automatic content recognition (ACR).

Transactions is pretty straightforward, your bank and/or payment networks sell your data to brokers at a merchant level. Many retailers sell your data at a SKU level.

Location is less visible. There are various levels of granularity of location tracking and reporting built into many SDKs that are included in just about any mobile app you download, and that data is sold.

Automated content recognition (ACR) is the craziest. The actual contents of your screen- the Instagram post, the TV show you are watching, etc (and its underlying transcript) are analyzed by the app or the TV, and sold often in real time to data brokers with such efficiency that you can be viewing the photo of the new Land Rover your friend just bought and posted, to your very next swipe being an ad for a Land Rover


Location and ACR may explain the ads I keep getting. But they're way off (because I'm cautious).

I happen to live in one of country with the highest GDP per capita (higher than the US) and I also spend lots of time in another country, in France, on a very famous place on the french riviera. That'd be the location part.

Then I love fast expensive sports cars. I just can't help it. Own two or three.

So what kind of ads do I get all the time? Ads for very very very rich people. Ads that want to sell me private jets flights or... Directly private jets. Yup: I'm not kidding, I keep getting ads where this insert many millions here private jets are for sale. Porsche cars: all the time.

Needless to say: I don't have anywhere what it'd take to fly private, let alone buy a private jet (even used).

But I find it funny: somehow the ad industry is thinking I'm completely ballin'.

Now of course I make sure to not only click on the ads but also pretend I'm deeply interested: for example I'll make sure to Google "Is a used Embraer better than a used Gulfstream?" : )


What is the point of your story? You click on ads and search for things targeted to billionaires and... get more ads targeted to billionaires.


"…your very next swipe being an ad for a Land Rover"

That automation has reached this degree of sophistication is why I claimed in my post that internet privacy is dead for the vast majority of people.

One cannot own a smartphone and maintain even a modicum of privacy unless one takes drastic measures such as those mentioned in my post. And the vast majority will never go that far and Big Tech is fully cognizant of the fact (acquiescence is so much easier than struggling to achieve one's privacy, it underpins Big Tech's huge success).


I essentially don’t see ads online. Do I have privacy or just an illusion of privacy?


If no one tells you your secrets but they all know them, do you have privacy?


Buuuut.... The flip side is that all these mechanisms are unreliable.

They might match your phone up with someone elses laptop and you'll see ads for stuff they searched.

They might get your location wrong by 1000 miles and suddenly all the ads are in French.

You might have clicked no to a cookie banner so now you get to see ads for ED medication despite definitely not having a use for it...

Advertising is all a numbers game, and they just aim to be mostly right most of the time.


How do you feel about working in marketing?


Let me guess: "I was only following orders."


All the more reason to use cash and throw the phone on airplane mode until you want to pull from the network. You might even get back to battery life measured in days territory of the dumb phone era making use of that setting more often.


God bless GDPR, I haven’t seen this technology yet and hopefully will never see. Does availability of such data/tech contribute to the discussion where it’s better to start the company - EU or USA?


You not seeing does not mean it's not used. Despite all my attempts at educating her on the matter, my wife still signs up for pretty much every newsletter and browses mostly from her Google Phone. This leads to things like having her discussing about Ikea furniture with a friend and less than 10 minutes later being shown an Ikea ad on the TV while watching YouTube.


I can be sure it is not used against me. I avoid of course using American services directly, only through their European entities.


> I can be sure it is not used against me.

Oh,the hubris on this one...

Can you say with 100% confidence that everyone you know and interact with has as good as a digital hygiene as you?

Can you say with 100% confidence that no company is acquiring data about you through shady brokerages?

Can you say with 100% confidence that no company had a data leakage (accidental or not) and went on without reporting it?


You are asking wrong questions. Of course, there exist data brokers with some information about me. It is not impossible that even a copy of one of my IDs leaked and one of my identities is currently actively exploited by robocallers. What’s important is that within the EU space most of that information is outdated, toxic and not actionable. There’s zero realtime surveillance and I do not receive any relevant product offers beyond reasonably expected behavior of marketplaces.


Outdated, how? I just gave an example of seeing targeted ads related to a conversation had minutes before.

Not actionable, how? GDPR hasn't stopped companies from collecting the data, it just added a bunch of roadblocks which are being overcome via dark patterns that make the majority of people consent to data sharing.


> Outdated, how?

Based on rare encounters with data processors outside of EU. They sell their products to me based on the knowledge that is at least 10 years old (and they use some old contacts for it).

> I just gave an example

Irrelevant. I‘m talking only about my experience, because I‘m actively exercising my rights under GDPR and similar privacy laws.

> GDPR hasn't stopped companies from collecting the data

GDPR made it illegal to collect it without unambiguous consent. If you want privacy, you can get it. As I can see from the inside sources, every big business in EU cares about compliance. Startups usually take a risk-based approach, but they don’t dare to sell the data and mostly compromise on security until they get bigger. Overall, here it is much safer and saner environment, if you don’t consent to marketing stuff. All you need is to read the label on the checkbox when you sign up.


> All you need is to read the label on the checkbox when you sign up.

Your first comment was arguing "tracking doesn't exist in the EU, because we have regulations for it". Now this here is a much weaker version of the argument, because it means "data tracking can be avoided if you are always vigilant and practice good Internet hygiene".

You might believe that you are protected, but what I am saying is that even those aware of tracking might end up getting a lot of their data collected via those that are not.


>Your first comment was arguing "tracking doesn't exist in the EU, because we have regulations for it"

It was not. It is false generalization of my words.

>those aware of tracking might end up getting a lot of their data collected via those that are not

We all leave pretty big footprint both online and offline, very often through a consented and legitimate process. It is a simple matter of data hygiene to make most of that footprint just useless pile of bits if you control the narrative and spread of information is constrained. Who cares if one of my phone numbers landed in a public database because I registered a company many years ago? I wrote a lot on this forum — do you think you can build a precise portrait of me with this, that or some other data? Do you really believe one can use it against me in a way I would not expect? A while ago I was working on a data mining project for some Western government, an information warfare they used to shape public opinion globally. I know quite well what data can be extracted, especially with modern technology. Without privacy laws there exists asymmetry between data industry and individuals, but the balance is shifting to the point, where you can actually manage what others know about you.


Discussing in person and the phone is spying? Or over some text messaging platform? (And if so, which platform?)


In person


I’m a non-Hispanic male in my mid-50’s and I occasionally get diaper ads on YouTube, or random ads in Spanish. I’m moderately privacy conscious, but I don’t do anything drastic to mask my activity. If the big tech companies are surveilling me so thoroughly, they’re not always doing a great job using that info.


People are very bad at understanding how large scale data collection works: namely, that it's entirely based on the law of large numbers. No one except under very specific circumstances cares who you are specifically, they care that they're right about that somewhat better then 50% of the time.

Accidentally misfiring an ad has no consequences if it doesn't happen in place of a better targeted ad and you don't reliably abandon the platform as a result.


This is why the idea you sometimes see floated that we should be able to sell our own personal data isn't really much of a solution. One person's data doesn't have much value on its own. If you were able to sell it, it'd be worth a few cents at most. Your data is only valuable as part of a large data set. I get the appeal of using capitalism to rein it its own excesses, but I really don't see anything short of strong regulation having a shot at regaining and maintaining our privacy.


I had a pretty long period where all my Google ads were in Spanish and a lot of them implied my demographic profile was set to think I'm a 50+ year old gay Filipino (I am none of those things, as far as I know). It just gets really off target sometimes.


People essentially bid on ad space, and when no one is particularly interested in bidding on you, you tend to get odd ads.

Company I worked for a decade back used to show charity ads when there was no relevant ad campaign.



I know there is the famous Target case where Target sent pregnancy related ads to the living-at-her-parents daughter. That's how the dad found out.

I remember a podcast about this. Maybe Reply All. But I cannot find it anymore.


It later turned out Target sent some of those ads to all families, blindly—it wasn’t causally related to the daughter’s pregnancy. Whoops.

But this way makes a better story.


Do you have a source for that? I've heard that claim before, and I can completely believe it, but I've never seen any real evidence for either side of the story, just a lot of guesses about how the system might have worked.


Even the original NY Times story was I heard from somebody who heard from somebody... Did the reporter talk to the angry dad? Who did they talk to?



I was expecting a concrete evidence on why the story is false instead of speculations arguing that it's unimportant. For me it feels like preaching to the choir, I already feel that AI profiling isn't as powerful as people screamed, and that one story is repeated too often, but I need some real argument to say so.



I learned about this from the audiobook version of The Power of Habit by Charles Duhigg. Maybe that is what you're thinking of?


Thank you so much! I owe you a <insert-drink-of-your-choice>!



The ads know that I'm a man in my 60s.

Lots of ED ads, lots of "How to retire" ads, those weird ads with retouched old men with six-pack abs, the stupid intermittent fasting ads, etc.

I do get the occasional run of Spanish-language ads. Not sure why.


My demographics have me in a few paid survey groups. They’ve finally quit putting “prenatal vitamins”, “diapers”, and “toys” in the middle of “what has your household purchased in the last six months” surveys. I guess they finally figured out it’s not happening.


>Surveillance encompasses both policing and caretaking, Hamacher notes. In practice, its polarized qualities—“beneficial and harmful, intimate and distanced”—intertwine. Baby monitors use technology developed for the military. Many contemporary models run on CCTV.

What does Jia think "CCTV" is, exactly?


tl;dr she didn’t send any obvious signals that she was pregnant, and didn’t get any ads. The headline only represents about four paragraphs from a very long article that rambles all over the place.


Then eventually she dropped the experiment, proving just how easy it is for one slip-up to tip them off:

> My modest experiment went surprisingly smoothly. Because I’d had my first child not long before, this time I didn’t need to buy anything, and I didn’t want to learn anything. I smooth-brained my way to three months, four months, five; no diaper ads. I called up a lawyer and data-privacy specialist named Dominique Shelton Leipzig to get her perspective. Globally, she told me, we generate 2.5 quintillion bytes—that’s eighteen zeroes—of data per day. “The short answer is, you probably haven’t hidden what you think you have,” she said. I told her about the rules I’d set for myself, that I didn’t have many apps and had bought nothing but prenatal vitamins, and that Instagram did not appear to have identified me as pregnant. She paused. “I’m amazed,” she told me. “If you didn’t see any ads, I think you might have succeeded.” I congratulated myself by instantly dropping the experiment and buying maternity pants; ads for baby carriers popped up on my Instagram within minutes.


I reckon privacy is a lost cause except for diehards like me. For by far the vast majority of people find the so-called freeware offered by Google et al just too convenient to resist.

For Google et al the tiny minority of us who've achieved some degree of privacy aren't worth worrying about, and we as a group aren't going to grow any larger for reasons that achieving a reasonable degree of privacy is a far too onerous a job for the majority of people, moreover most would consider their phones broken after such privacy modding.

For instance, the phone I'm typing on now hasn't yet been rooted but that largely doesn't matter, it's reasonably private (but not completely so) with the following tweaks: first, it has no Google account (probably the most important tweak of all), all apps except for a few F-droid ones have internet access disabled (their calls to the internet are diverted to a VPN nul location by a firewall), all Google apps are disabled including Google Play Services and especially Chrome, no non-F-droid app has access to background data, and as a precaution the disabled Google apps have all permissions denied/turned off, in effect they have access to nothing, neither the internet nor hardware. These tweaks send both the Play Services and the apps that rely on it into spasms and they keep bleeting notifications to turn Play Services back on or the apps won't work, this bleeting is most annoying but it too is easily remedied by disabling notifications from the offending apos. Finally, my F-droid browsers have JavaScript disabled (there's more but that'll do for now).

Wirh these tweaks the phone still works fine for me, calls and messaging all work OK, so too do GPS, WiFi, the internet along all phone sensors and my non-Gmail (POP) email account.

That's about the minimum one needs to do to achieve even a modicum of privacy on one's phone. That said, a phone so tweaked is essentially useless to the vast majority who expect Google, Facebook and similar apps to work (if you want privacy you just can't use them). That the vast majority cannot do without these Big Tech apps is why I believe privacy is essentially dead.

Incidentally, some may be interested to know that many (but not all) apps that complain about not working if Play Services are turned off actually do still work. What Google and developers don't tell you is that these apps use the Play Services to report your activities to Google and the world, it's a key reason why I nuke Play Services.


Usually microG makes the GMS complainers work while still voiding all the data they try to report.

IMHO unrooted android is starting to become actually viable, for example there are apps (like AdGuard) that use a VPN to MitM all your connections and filter out ads even if they attempt to hide inside HTTPS connections.

And with the `pm disable-user` trick you can disable apps that you can't normally uninstall (or disable). Which is great for OEM- or carrier-installed bloatware.


Also worth noting: it was her second child so she “didn’t need to buy anything or search for information”. And the experiment was terminated (it appears - it’s a bit unclear) after 5 months.

Another interesting factoid mentioned:

> identifying a single pregnant woman is as valuable to data brokers as knowing the age, gender, and location of more than two hundred non-pregnant people, because of how much stuff new parents tend to buy


I don't remember the exact list, but it was 4-6 life events that are each potentially worth $100+ per person to marketers, and I think that was in ~1996 dollars.

Iirc:

College graduation/ first real job, Wedding, first home purchase, first kid on the way, retirement filed-for, and ??? I'm forgetting something. Maybe out-of-town move?

Related and overlapping: I suggest searching for the late 90s article on Target basically telling a teen that she was pregnant via direct mailers before she even knew. Dad over reacts then has to eat his accusation when they (Target) were right.

Based on non-typical purchase of un-scented lotion and 2-3 other undisclosed items.

I'm sure it's only gotten worse since then.


Wow, other than retirement, I’m glad I passed all those milestones in the before times.


I wouldn't be so sure you did.

Idk how old you are, but part of the point is the 'before times' ended much earlier than most are aware.

The biggest lesson of the Target story (for the marketers) was don't let consumers know that the marketer knows so much about them.

Had they just sent diaper coupons mixed with generic ads, they might have been more successful. 'Congratulations You're Pregnant!' is what didn't work.


This is very typical of New Yorker pieces - I enjoyed the broader context of capital surveillance. We can’t editorialise titles here, so this is bound to happen.


Thanks for saving me the time, I was wondering if she will go into more detail, but the more I read the it went sideways.


[flagged]


Compare actual crime rates of today in the US to as recently as the 1970s, and even including the small spike that happened during pandemic lockdowns, you're better off in a 'dangerous' inner city now than you would have been in a 'safe' suburb then.

Of course, we still won't know for another generation or two if the drop over time is because we're doing anything societally better, or just the effects over time of banning leaded gasoline.


"Using the FBI data, the violent crime rate fell 49% between 1993 and 2022...FBI data also shows a 59% reduction in the U.S. property crime rate between 1993 and 2022"[1]

My take would be the job that the police aren't doing is traffic enforcement, and driving has gotten noticeably worse in the last decade.

1: https://www.pewresearch.org/short-reads/2024/04/24/what-the-...


I can’t tell if you believe the world is actually more dangerous now than it was in the past or if you are describing public perception.


If ring cameras haven’t made communities safer (doubt), they certainly have made it much easier to get video evidence of criminals. Pretty much any real life criminal case heavily relies on pulling surveillance data from private residences now.

That being said, fuck this surveillance state.


Isn’t crime down significantly, decade over decade?


A lot of the time police would like to do their damn job, but they have been crippled by more and more restrictions. And then even if they manage to still do their job, the justice system doesn't cooperate, letting criminals out on ridiculously low bails, lenient sentences, etc. All in the name of "disproportionately impacted".


This has such back to the fifties sentiment (leave door unlocked) and I resent that solely for the fact that women had far fewer rights back then.

Yet I feel for you as I myself still live in a country with a functioning government and a police force that is there to help citizens and not just there to protect the rich.

That said, your approach to porch package thieves is missing the point. You don’t steal if you are doing well. Tons of people barely getting by with little to no safety net. Catch a porch pirate and another steps in. But solving the root cause for theft is a political thing and given current political circumstances, caring about people is seen as weak by half of the electorate.


> You don’t steal if you are doing well. Tons of people barely getting by with little to no safety net. Catch a porch pirate and another steps in. But solving the root cause for theft is a political thing and given current political circumstances, caring about people is seen as weak by half of the electorate.

I don’t buy this. Theft wasn’t this bad in San Francisco just ten or twenty years ago. Theft also is worse in SF than it is in places like Alabama/Mississippi where they actually enforce laws.


There's a difference between stealing shit out of a grocery store to survive - but ffs there is no justification for police to let people openly fence stolen goods, looted en masse in flashmobs, on the streets.


> You don’t steal if you are doing well.

This is absurd. Rich people steal too. Off the top of my head Bernie Madoff proves this assertion wrong.

You’re perpetuating a harmful stereotype that associates poverty with crime.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: