Hacker News new | past | comments | ask | show | jobs | submit login
Instagram ads Facebook won't show you (signal.org)
1186 points by HieronymusBosch 2 days ago | hide | past | favorite | 175 comments





See, the trick here is to put all the personal data onto an oddly specific t-shirt that's for sale.

Available now, size "Large" t-shirt for sale!

"I'm a proud dad living in Seattle who attended the University of Washington and once went on a trip to Central America for a while. I have a dog and like to read and occasionally complain about politics."

I'd buy this if I saw it, if just for the lulz and nihilistic outlook when it comes to privacy.

In fact... if I were a privacy-focused company, I'd 100% do this as a marketing stunt.


OMG. This is just brilliant on so many levels. It's not just subversive against FB's rules but also can be the start of some kind of art project as a statement about what tech companies know about us. But the project itself gives us a choice in what is shown to others whereas FB doesn't give us that choice when selling the info to advertisers. It'll probably get shut down real quick by FB though once they know about this.

Edited: as my sibling comment mentioned, I too would buy one of these shirts.


A crude version of this exists: https://www.reddit.com/r/TargetedShirts/

Except this is Tshirt manufacturers coming up with generic slogans on Tshirt ads that correspond to your interests.


The scale of 'my boyfriend has anger issues' shirts is really dark...

Dog mom's got anger issues, too... it's kinda fb's bread & butter

I would buy one that describes the total opposite of me and watch the confusion on people's faces.

I so wish this was a thing but I doubt Facebook would allow using their graph for this kind of marketing.


This has existed since 2014, in embryonic form.

The exploit was that FB put your name in the public response from the Graph API. I remember discovering that, saying that might be abusable, and moved on with my life.

About a year later, I started getting oddly specific t-shirts like all <last names> do <x good>.


Ridiculously targeted t-shirt ads on Facebook are actually a thing, but probably not for privacy awareness purposes: https://thehustle.co/who-makes-those-insanely-specific-t-shi...

This article got dark very fast

Even without seeing the result, it was entirely predictable what was going to happen when someone let a computer generate phrases from word lists. The guy who went out of business generated 700 different combos algorithmically and then did not proofread them. That seems absurd. 700 isn't really that many, a few hours of manual labor to cull the bad ones, and his business may have made money.

I think other people have made money successfully with the few hours of manual labor version. At least, I think I've seen their customized products continue to exist and be available...

Where did you get the number 700? The article says they had 22 million shirts

From the article:

"In the end, he generated about 700 variations of the phrase on t-shirts, and put them up on Amazon."


700 "Keep Calm and ___" shirts, they also generated 22 million "___ a ___ who ___ and __ " style ones for their catalog and listed 550k of them on Amazon.

With four options he may have only had to check a few hundred words. 68^4 ~= 22 million

All of the examples shown in the article were of the "keep calm and ___" variety. With 550K more in the catalog, I imagine there were plenty of wild ones.


I really really like those cell phone cases.

They made me laugh out loud for a few minutes. Did anybody actually buy those? There has to be someone with the right sense of humor.

Even better: phrase it in relation to the person reading the shirt, not the person wearing it:

"You are seeing this shirt because [advertiser] wants to reach people who are friends with individuals who are [age], interested in [topic], located in or near [city], who wear [size] shirts.


As a FB marketing expert, this wouldn't work. Yes we could target you, but the market is too small and not worth the effort. The minimum audience worth targeting

As someone else pointed out, personal attributes are prohibited from ads.

My feeling is, communicating a compelling data collection story, even strictly positivist things like how much data is collected, let alone normative ones like we should collect less data or prohibit collecting it - you're not going to tell that story with some neat hack inside the system.


You can get very close, I had an ad on Facebook the other day with transliterations of hindi/urdu words on it. When I clicked to see why, Facebook thought I was interested in a Pakistani newspaper (I think I'd read some articles on it on my phone, I guess the trackers picked that up). I guarantee you 99% of readers of the newspaper where I am (or probably anywhere) are of Pakistani descent. The ad wasn't telling me my race/national origin, but it was clear Facebook knew.

I remember seeing a t-shirt ad that said something along the lines of "March people are X"... and I'm born in March. It just made me that I contributed to Zuck's data hoard.

Relevant facebook group: Shirts marketed to extremely specific demographics

https://www.facebook.com/groups/smtesdthemain


It hurts to admit, but I would probably actually buy one of these

You aren't alone - I know of a company that bootstrapped themselves by abusing the graph API to sell t-shirts.

Considering this is the top post on HN, it's only a matter of time before someone offers these for sale and has a Show HN post about it. I just hope they have a coupon code for HNers!

I wonder if this tactics could be used to increase click rate. Sound like a good idea to grab someones attention and maybe reduce spending by targeting niches.

Ooooh I like the idea of wearing those types of shirts ironically

I'm probably not going to, but I do like the idea


this.. would have worked and I would've believe the shirt was made for me..

A pretty cool piece.

Too bad they have to use text to make their point. It would essentially reach zero people due to rules (https://www.facebook.com/business/learn/lessons/how-to-adher...). Then there's personal attributes (https://m.facebook.com/policies/ads/prohibited_content/perso...). Then ads that do not sell products/services follow murky rules, and talking about Facebook itself is usually prohibited. (edited from: because the rule they're actually breaking is the "No Text" rule in Facebook ad creatives.)

Is there non-symbolic imagery that they could have used to say the same thing?

Perhaps they should have retained someone with this kind of creative experience.

Looking critically, the most narrow and serious obstacle to advocate for privacy is storytelling.


> "No Text" rule in Facebook ad creatives.

Where can this rule be found? That seems like a really odd rule, and https://www.facebook.com/business/help/388369961318508?id=12... says the opposite:

"Avoid too much text on the image itself. We've found that images with less than 20% text perform better, though there is no limit on the amount of text that can exist in your ad image."


I'm so glad they changed this rule from a blanket prohibition, since it precluded my company from effectively advertising on FB. We have developed a novel way to display text on screen that makes it easier and faster to read, and improves accessibility for people with dyslexia and ADHD.

With a screenshot of text, it's super easy to understand how our tech works and whether it is useful for a particular person. But FB wouldn't let us boost posts that had images of our product in use.

I'd be curious to hear from others how much FB penalizes advertisers in terms of cost/reach if they have lots of text.


The story goes that one of the founders hated text in ads, and decreed that it should be banned.

They later converted text in ads into a penalty, so you pay more (as an advertiser) for it.


The policy recently changed and now officially allowable. They used to even have a tool to test the old 20% rule [1]

But still their algo doesn't promote heavy text as much. It's a black box; ads aren't as simple as highest bidder, FB takes engagement on a per user basis into account and evidently heavy text lowers.

Which is ironic given they added those text only posts & the takeover of Memes which at least for me drove my personal browsing to IG only. And now that's starting to go the same way. maybe meme-ification is humanly inherent lol. I'm starting to unfollow friends who only post those inspirational memes and crap. I just want photos, inspiring content from creators/athletes that are relevant to me, and family updates.

https://www.facebook.com/business/help/980593475366490?id=12...


FB uses a rather obscure and proprietary internal scoring system. You may be able to publish an ad campaign with lots of text once or twice, but it will make your ads account score go down and at some point it will likely be disabled - temporarily first, after one or two more incidents forever.

The recommended amount of text on the image is less than 20%. Their system also doesn't like text because of potential "circumventions of policy" with putting text in images. Weird fonts that are not machine-readable will likely get the account banned fast.


This actually matches with my experience. My product is text-related and I ran a couple ads with allowable amounts of text (back when there were limits). The ads performed just fine, but FB decided to stop running them, claiming they were low-quality.

Do you have a source for this? I find it very hard to believe considering it's in conflict with the explicit unambiguous wording in their policy.

This actually violates the “personal attributes” rule. In my experience, this rule is enforced quite strictly — although you can still see the same targeting criteria under the “why did I see this ad” feature.

https://m.facebook.com/policies/ads/prohibited_content/perso...


This rule reads as “don’t tell the target about our assumptions of them”. Isn’t that kind of the point of Signals post?

I'm assuming the rule is around because FB knows it's targeting isn't 100% accurate and it might ruffle feathers if an ad claims you're something you're not. I don't see that specifically as nefarious.

Especially because FB tells me why I am being shown an ad already.


In the early days of COVID I was curious what the difference between types of viruses. e.g. how herpes and HIV hides in cells compared to viruses like influenza and SARS-CoV-2.

One of those sites clearly had Facebook integration because now Facebook is sure that I'm an HIV-positive gay man, with ads that correspond. It is one thing to get the ads but it would be a bit more overt if there was a text ad declaring that I was an HIV positive gay man.


People take offense at being told that there’s a machine labeling them, and would demand the right to revise or remove the machine-assigned labels. Facebook prevents that outcome by preventing disclosure of targeting characteristics.

See also this Show HN from /new for purging your complete set of labels, and ask why Facebook doesn’t build this into the site UI. https://news.ycombinator.com/item?id=27045374

(The link is broken, but apparently there's a page you can single-remove specific interests from, if you can find it.)


Wow, that item link points to a literally blank page. The site template is there but there isn't even a title header.

Sadly even the web archive's copy of the page has the empty version :(


Yes it is. Signals ads might still get approved if they took out the 'you' voice. But that dilutes the awareness they are bringing.

They could just turn it into questions ("Are you...?")

The rule doesn't exists anymore. Ads with a lot of text simply get penalized when it comes to distribution.

> because the rule they're actually breaking is the "No Text" rule in Facebook ad creatives

maybe breaking the rule was intentional, to make this article work


I don't think the point here was to actually run these ads. They don't look like they'd drive a lot of conversion, and there's a big risk that quite a few would actually be inaccurate in practice - Facebook's algorithms, while they know a lot about you, can also be quite wrong.

But you can't see that in an article like this, and it's far more likely to reach the right people.


The no text rule has not been enforced to disable an adaccount for years now. It merely de-ranks you. Source: Worked in advertising.

Though you can't have text in the image, you could put the target filters as the post's text and use the image to grab the attention.

Signal should sell this exact design on a targeted t-shirt and then advertise that.

This is both an ad for Signal and an ad for Facebook Ad's ability to target depending on who is reading it.

What I mean is that for your average consumer, they'll read this and be horrified that Facebook is using the information they voluntarily gave Facebook to make money. But someone who is buying ads will read this same thing, be impressed by just how tightly Facebook can target, and put $10K into an Ad Account to try it out.

As to me, I use Facebook, I am willing to see ads within Facebook using the information I share with Facebook but where I draw the line is Facebook "leaking" into my wider web browsing history (either tracking me, or using my non-Facebook browsing to advertise to me on Facebook). Therefore, I use Mozilla's Facebook Container extension and blacklist Facebook/Instagram's "Share" tracking buttons.

I also access Facebook from a mobile browser rather than app and use Signal instead of Facebook Messager, to limit Facebook's ability to track my location and other phone meta-data.


It's absolutely transparent then that, without intervention, companies act in ways that are against individual's and society's best interests in order to make more money.

With that evident fact, we to face the reality, however uncomfortable, that manufacturing desire at this scale has become unambiguously unethical.


Have we also reached a point where no intervention will fix the problem?

No

Any ad person in 2021 who isn't aware of this must be a sophomore in college still.

This claim seems to be predicated on some unrealistic standard wherein all businesses have dedicated professional in SEO and or advertisement management (or outsourcing to media management companies who do).

That exists of course, but sole proprietors, small businesses, and medium businesses also purchase online ads in high volume and therefore people from other domains are commonly buying ads (or more importantly for this discussion: deciding where to spend limited ad dollars).


Isn't it clear the moment you buy your first ad on Facebook that there are extremely specific categories you can target, i.e. to any "person buying ads"?

Yes, non-ad people doing ad work may not be aware of it. Anyone who is an "ad person" who is unaware of the degree to which internet ads can be targeted is not an "ad person", though.

Years ago I added a widget to the user interface on wikipedia for logged in users so that people were able to add geography specific notification to tell editors about meetups that were coming up in their area.

It turned out that if the message displayed was too specific, like "Upcoming meetup in your area: [New York meetup]" people got rather angry about the privacy invasion.

So instead the instructions for setting the messages had to tell the authors to instead say stuff like "Find out about upcoming meetups!" -- which of course was only displayed if there actually was an upcoming meetup near where you geolocated.

Of course, regardless of if any message is displayed the site could guess your geography based on your IP address. The exposure of private information was nearly identical-- actually arguably worse because someone might mention that they're currently seeing a notice without realizing that this fact leaked their geography... but the more generic messages didn't generate complaints.

(and WP policy effectively makes it impossible to edit via Tor, even for established users in good standing)

Sometimes it seems people care a lot more about enjoying the illusion of privacy than they care about actually having privacy.


>Sometimes it seems people care a lot more about enjoying the illusion of privacy than they care about actually having privacy.

Hiding the fact that you are spying on someone enough to make them stop openly complaining about it doesn't make them ok with you spying on them.


> The ad would simply display some of the information collected about the viewer which the advertising platform uses. Facebook was not into that idea.

Genius! But it’s unclear to me if the examples in the blog were actual ads shown to users before their account was blocked, or the campaign never got off the ground at all. If it’s the latter, the blog should make it clearer otherwise it makes it look like those were real ads


Those are real targetable attributes under Facebook. They would be real ads if approved.

I highly doubt they would ever be approved.

Right because they're ads critical of facebook designed to scare facebook users into not using facebook.

Why would facebook green light ads that are detrimental to them?


They're explaining something facebook does without any bias at all. Facebook should be proud to show off their technology! If it's "detrimental" then the problem isn't the ad, so rejecting the ad for that reason is scummy.

It's an ad for a competitor. Why would they do that?

In theory blocking ads from competitors is a violation of anti-trust laws, but when you have more money than god you can just pay the fine.

Well that's a completely different argument. If you want to talk about the part where it makes them look bad, consider an organization similar to the EFF doing this campaign.

I like whoever is running Signal's blog.

It’s definitely the executives. You don’t get to say something like “a Cellebrite fell off a truck” if you’re anyone besides the CEO.

Anyone with basic legal training would definitely say the same. Anyone reading too much hacking/legal/mafia fiction may say the same. Anyone from France could say the same, given that expression [0] entered common language back in the 70-80's when expropriating logistics trucks to redistribute wealth to poorer neighborhoods was very common practice.

[0] https://en.wiktionary.org/wiki/tomb%C3%A9_du_camion


It's not about the commonly-known phrase, it's about the fact that it was used in an official blog post.

And even cute staged pictures of the Cellebrite in the road. Nobody other than the CEO even would think of doing something so provocative in a space with national security interests and regulators involved.

Moxie wrote the Cellebrite article.

There are several writers, you can see their handle and pic at the top of the article. In this case, Jun Harada.

The editors and ghost writers aren't named.

Facebook already has a way in-app for you to see why you were targeted with an ad (on any ad click the 3-dot menu -> "Why am I seeing this ad?"). The tool will tell you things like whether you were retargeted vs targeted using lookalike audiences, targeted based off of your age, gender, location, interests, etc.

I don't think this is the "slam dunk" the author intends it to be, but I'm sure it will resonate with the Woke™ hackernews crowd regardless.


Facebook's "Why am I seeing this ad?" info is deliberately incomplete and does not reveal exactly how the advertiser targeted the ad.

https://qz.com/1245941/why-am-i-seeing-this-ad-explanations-...


Hit a paywall so I can't comment on the article, but just from the hero image the current implementation has more detail than what's being shown there. It looks like this article was last updated in 2018 though, so its claims might be dated?

Believe it or not, sometimes you want to raise awareness about something that would otherwise be hidden behind a kebab menu and rendered entirely clinical.

I don't think this is the sound criticism you intend it to be.


Believe it or not, I agree! But raising awareness and acknowledging that the information is readily available through the UI is not mutually exclusive.

After reading updates from both sides, I don't think Signal intended to deceive readers with this blog post, it just sounds like they jumped the gun. It seems like whoever authored the post doesn't understand Facebook's advertising tools all that well.

https://twitter.com/joeosborne/status/1389770672172318720


Ha, sorry, my comment was way too sarcastic. And yes, that looks to be what happened here.

guess what this guy does for a living...

:) is not taking things at face value a profession?

This isn't shocking at all. If anything, it makes me want to make a business account so I can see first hand what targeting criteria would be available to me.

I'm not an FB user, but I might as well be, since I have an Instagram account that I mindlessly scroll from time to time.


> If anything, it makes me want to make a business account so I can see first hand what targeting criteria would be available to me.

Here are some examples: https://imgur.com/a/7YVH3ch

There are likely thousands more, that's just the browsing section.


Damnit. That is very granular.

You don't need a business account but you do need a personal account (or rather a personal account can be a business acct).

The targeting is relatively specific. Think of a (general) category and it'll be available. I use it for bars and restaurants at a local level, and the granularity is quite nice vs other media (target people who like bars or restaurants, are into music, are into drinks, are into cocktails, are into concerts currently in the city or who are traveling to this city)

It really helps a small budget go a long a way, assuming the stats they give are correct.


> assuming the stats they give are correct.

And herein lies the rub.


I'd be surprised if there weren't different tiers of advertising accounts just to prevent normal people from having access to this. For example, you have to spend at least $10k before you get the next level of targeted information.

So you're saying Tom Cruise can probably target ads better than I can?

In all seriousness, I wouldn't doubt that either - spend more, target more. But based on the screenshots someone else posted up a bit, it's pretty granular (I presume) by default.


Even without a business account, you can click the 'why are you showing me this ad' button on FB and it'll give you similar text (although sometimes it's more vague depending on how the ad is targeted)

If you can't beat them, join them

This doesn't seem very effective to me. 99% of people who see an ad like that will not care. It's already common knowledge that Zuck's gonna take your data.

"Facebook knows I'm a single teacher in Moscow who likes soccer? ... So what?

And that's before taking into account that the labels FB/etc. put on you are often incorrect, further diluting the perceived seriousness of this privacy leak.


The point of ads isn't always to get a reaction. Clicks are the grand goal of course but mere impressions are valued by marketing standards as well. The fact that the target audience sees the ad, even if only passing by while scrolling is widely considered a success by marketing standards. And equally needless to say but I'll say it anyway; that's the very point of Zuck leeching, so the ads will find the target audience.

Agree with you there that this isn't much of a privacy leak as the average user mostly knows what's going on. I'd guess the article wasn't really meant to point out a threat to privacy, maybe more on the lines of "FB doesn't want to share it's methods of using the information it gathers". Shocking...


>99% of people who see an ad like that will not care. You're making quite the assumption here mate.

Flip the number. If your ad for literally anything had a conversion rate of 1/100 you would be over the moon. I think it's quite the overestimation that even 1/100 would do anything on seeing this ad.

It's about how it's framed. People don't know how to feel until you tell them.

Take for example the obesity epidemic. Obesity is a factor in 20% of all US deaths. People know it is killing them, their friends and their families and don't do anything.


I agree that these are extremely general categories that could be reproduced by scraping a person's public LinkedIn.

What's trickier is when the ads make assumptions about your taste based on the Facebook groups you participate in, and the websites you visit outside Facebook. Those are still connected to you via the Facebook beacons (share widgets) embedded in practically every website.


Exactly my thought. The "So what.

What do I care if Facebook shows me ads for the things I browsed on Amazon or Etsy. I often discover fun stuff directly from those ads for websites I wasn't even familiar with. On the contrary (and I could be wrong to do so) but I trust some website when I have seen its ad on Facebook, as I know it has been vetted by Facebook to not be some fraud.


You're missing an /s for the last bit, right? Like I'm actually with you that really specific targeting is actually great for product discovery and I wish I could harness it for myself not via ads sometimes but there is so much obvious crap and scams on FB/IG ads that it's not even funny. I have more faith that I'll actually get the product choosing a random AliExpress seller than a random Facebook ad.

Plus, it will likely lead me to think "How does this company that advertises on Facebook know so much about me? What kind of shady information practices do they have?"

Signal hitting hard as always. I wonder if they are trying to get people to stop talking about their cryptocurrency hiccup.

I've run ads on FB before, but this is an incredibly simple article to share with my non-technically minded friends and family as to why these services collect too much data. We need more of these simple and concise posts to share outside of the tech-bubble we live in.

This is a great pr stunt.

There’s a similar story about a guy who sets up an add targeted to his wife or fiancé or something.

Later Facebook apparently made it so whatever group your targeting has a minimum size.


I remember a story told on reddit about a guy who targeted his roommate with incredibly specific ads until it freaked him out.


Their campaign execution is brilliant...love it.

That said, nothing about this is new. Whether Facebook, Google, or any of the other countless (yes, thousands) of players in the AdTech ecosystem, this kind of targeting can be done with ease and for pennies per user.

The deprecation of third-party data, cookies, and cross-domain tracking couldn't happen soon enough. It's not a perfect solution but it's certainly a step in the right direction.


For any that haven't seen it- a great essay on advertising and how it relates to cancer: http://jacek.zlydach.pl/blog/2019-07-31-ads-as-cancer.html

You can export the data Facebook has on you and see what could be used to populate these ads for you.

https://www.facebook.com/help/212802592074644


This isn't some attack on Signal specifically the personal ads policy has already existed and it was enforced against us once with a somewhat similar idea for voter locality. It would have been approved likely if took out the 'you' voice examples in the policy below.

Though I understand the point of this as marketing & article to educate on what data FB does have and the ads look really cool!

https://www.facebook.com/policies/ads/prohibited_content/per...


Looking at the 'Suggested for you' posts in Facebook and Instagram, I am really puzzled about the way data is used.

Whereas they definitely have access to a lot of (my) data, they are still unable to correctly know my gender.

And speaking of rules for ads. I recently noticed that most of the ads are pure spam and even fake products are offered. It is extremely obvious what this is about.


Ha, when the Ledger Nano database was leaked a few months ago it published the address data of roughly 300k users, including their email. Given the fact that you can upload e-mail adresses to Facebook for more directed targetting (really nice feature lol), I thought it might be fitting to advertise a 5$ wrench offer to each of these users and if they might be interested in one. Really weird actually that it's OK to upload other peoples contacts to such services without any checks whatsoever.

You can use that feature to do some incredibly specific targeting to mess with people - as in, make some incredibly targeted ads that will only show up for a single person.

Brilliant ad campaign. So scummy that facebook allows advertisers these insights, but immediately shuts it down when you try to inform their users.

I wonder what the reasoning for disabling the account is. It would be extremely funny if the T&Cs said you can’t expose FB as creepy. :D

For personalized ads, Google bans [1] ads that "Imply knowledge of personally identifiable or sensitive information".

Facebook has something similar [2] but much narrower. "Gender Identity" is among the categories though ("LGBTQ adoption" in the first ad), as are medical conditions ("pregnancy exercises" in the second ad).

[1] https://support.google.com/adspolicy/answer/143465 [2] https://www.facebook.com/policies/ads/prohibited_content/per...


"Imply knowledge of personally identifiable or sensitive information" That's golden. It's perfectly okay to track users and collect their personally identifiable or sensitive information, and to serve them targeted ads based on it, but you must never let them know you're doing so.

Claiming that "LGBTQ" is a claim of gender identity is a stretch.

And, wait, does that rule apply to marketing manly products for men? If they're using "gender identity" as a euphemism for "non-cis gender identity" that's kind of awful. It reminds me of the people that say they don't have pronouns.


Someone mentioned that it’s the “No Text” if anyone else was wondering

That rule no longer exists, so it isn't that.

It's far more likely to be the "no personal attributes" [0] rule.

[0] https://facebook.com/policies/ads/prohibited_content/persona...


Which contradicts Facebook's written policy, so I doubt it: https://news.ycombinator.com/item?id=27040800

(although Instagram could have different rules)


It's probably the personal data clause. Something like you may not call out personal characteristics of an individual in your ad.

Signal has been putting out 10/10 content lately, I hope they keep it up and in the mean time I will continue to use their services!

Is there a way I can see this for myself? Ie who facebook thinks I am? Is there something similar for google or other ad networks?

You can get an extremely limited look on the ad preferences page. "Categories used to reach you" and "audience-based advertising". Nothing quite as slick as this little instagram hack.

https://www.facebook.com/adpreferences/ad_settings


Facebook is terrible.

However, Signal has its own failings as well. From what I understand, it:

* Refuses to federate.

* Hostile to independent clients.

* Run as a one-man show.

That's not Facebook-bad, but it's sad that Signal is consistently exhibiting this attitude, meaning that it can't be a good basis for personal instant messaging going into the future.


They have good reasons for doing so. Watch this talk if you actually want to learn why https://www.youtube.com/watch?v=Nj3YFprqAr8

If you don't, use something else. But using Signal is for sure a lot better than using Facebook Messenger or WhatsApp.


In what way are those failings?

In that it is similar to how Facebook operates:

* Facebook refuses to federate.

* Facebook is hostile to independent clients.

* Facebook is run as a one-man show.

And everyone agrees Facebook is Bad(TM).


I did say Facebook is terrible...

Sam Lavigne did a similar project in 2017, generating targeted video ads on twitter.

https://lav.io/projects/the-infinite-campaign/


This is brilliant and I love it.

Is it just me though, that for how targeted FB and Google try to be, they're usually still way off?

I mean sure they'll know things like relationship status that I directly input, but the inferences are often very incorrect.


Some more fun reading on Facebook targeting from 2014: https://news.ycombinator.com/item?id=17112559

I didn't know there was this level of detail in FB ads, I'll have to start targeting more specific audiences for my ads then. Very useful info.

Facebook business model will not be banned as they do what services always wanted to have - a large network of wilful informants. Keep feeding them data.

Would be cool to sign in with Facebook and then generate the ad set for yourself that they WOULD have made. I think it’d be quite telling.

cant this be fooled though? I believe there is a site where you could select a "profile" and it would open up these links in your browser to give off the impression of that profile.

For example, if you selecting "wealthy" it would just open up a bunch of urls to expensive brands, luxury items, private charter jets. I am not entirely sure if it actually worked or not.


And then there is me who has no idea about ads. NoScript, PiHole and uBlock Origin tend to do that so ¯\_(ツ)_/¯

I wonder how your post and posts like yours would be received if you said you have no idea about the purchase button or pricing of apps in app stores because you had ways to get around app and in app purchases.

Probably not well. At least usually not well. Depends on the day and the audience of course.


Comparing apples and oranges. I doubt HN posts lately are full of purchase buttons / pricing of apps having a negative effect on the surveillance we're constantly under.

It is apples and oranges in that they are both fruits with many common traits while also being different. In the end, what difference does it make to the creator what your reasons are? You’re depriving the creator of money.

My hypothetical Apple or Google App Store pirates could have a surveillance issue claim as well. Would that really change how angry many people would be with the behavior and the difference in response vs ad blocking?


That doesn’t mean that you aren’t being micro-targeted by ads. It just means you don’t see to whom your personal information is being sold.

Considering that:

1 - using this setup means considerably less info is exiting my systems to feed the beast

and 2 - exactly because I don't see them they're wasting their time on me

I'd say mission accomplished. Wanna guess in how many months the current state of ads/ad R&D would drop to 0 if everyone would do the same like me?

Here is a funny story, from this year. crackwatch dot com (the site, not the reddit) went down. why? because pirates that flocked there with thousands, all had this setup and they got exactly zero ad revenue from their users. Of course they went with a nice corporate BS "we could use more aggressive ads to raise revenue for servers but we respect you guys too much for something like that bla bla..." - translation: we know is not going to work so we just going to cut our losses.


I’m going to echo the sibling commenter. You somehow tried to negate my hypothetical App Store pirates by saying it’s a different situation without commenting on actual substance.

Now you’re admitting your set up is awful for creators and would screw things up for genuine people if everyone did it. Just because whatever creator example may not be a good site or person doesn’t mean that’s how everyone is.

I’m sure you wouldn’t want everyone taking whatever service or product your livelihood depends on.


So you acknowledge that if everyone acted like you, it would destroy most of the sites we use every day?

And they are building signal for peddling their ecoin?

I believe they're making a point about privacy and ad targeting that's somehow orthogonal to their cryptocurrency.

It is not orthogonal, they are clearly positioning their own cryptocurrency-via-chat as preserving privacy and are signaling that they are better than Facebook in that respect.

Signal is a non-profit that was primarily funded by a WhatsApp co-founder who left Facebook because they didn't like the plans to add Ads to WhatsApp.

From what I've heard, that was a zero interest loan that needs to be slowly paid back.

I don’t see how people believe the founders of Whatsapp didn’t know what was going to happen. They are getting a free pass and even more than that - praise. No one gives you $19 billion dollars for a product..only to barely monetize it and do it inconsistently with how the rest of the related company operates.

The amount of money that was left on the table in their early departure leads me to think there was at least some naivety on the way in to that deal.

Reduce the amount of text, theyll let you run it.

Seriously how is advertisement legal?

If you truly believe in free market then surely you must agree advertisement is a mass manipulation technique that should be illegal as anti-competitive technique (reinforces dominant positions).

If you're an anarchist/socialist then surely you've read or seen some talks by Noam Chomsky about "Manufacturing consent" and by now you want to burn down every TV station, bank and police station you can think of.

Even if you don't mind printed ads, if you're just a little bit concerned about privacy, you must be out of your mind that certain data obtained about you may be used against you and your loved ones

Who's left to defend that kind of degrading practice? How can we put enough social pressure on these people so they stop and develop healthier activities than to hijack our brains remotely?


Advertisement is good because it lets me know who is supplying goods and services I may desire or need.

There are a lot of methods allowed in advertising that are basically fraud, and a lot of methods to deliver advertising that are basically stalking. That doesn't mean that people shouldn't be allowed to post what they have for sale.


> goods and services I may desire or need

Well the thing is the amount of resources spent on advertisement is usually inversely proportional to the amount of resources spent on "goods and services i may desire or need".

I'm not saying it should be illegal to post about what you do. But advertisement is paying people to relay your message they couldn't care less about. It's not the same as people recommending products/services, although the frontier has become blurrier now with all those "sponsored" articles/videos.


Is the only way you out about things you want through ads? I don't even remember the last time I made a purchase because of an advertisement.

This can't possibly be true. Why do you think Apple is so popular among tech circles? It's not because everyone collectively decides to read their privacy policies. Apple has to actively advertise they don't collect data.

The simple act of making a spectacle of their products announcements are advertisements. Not clickable ads in a website but an ad regardless. They could quietly release products every year by updating their website but that wouldn't sell nearly as much...


I don't know if it's just me but I can't stop thinking about the day Facebook & Co will get bored with selling those stupid ads and will use all their powerful datasets to do more dangerous, scary things.

Will then everybody think "Oh, I really didn't see this coming..."?


Or perhaps they'll set their algorithms free to figure out how to better monetize all the info Facebook has on people, and the machines will figure out that blackmail is a possibility: maybe FB could sell ads that would let you pay not to have certain information shared.

Or someone else will do something with their data.

The irony that Signal itself has an Instagram account...

https://www.instagram.com/signal_app/


Could be seen as ironic, but I frame it more in the light of being critical of something, yet also understanding the value it can bring.

I may be critical of industrial farming, large corporate environmental policies and/or Facebook, but I still might buy corn, own a Toyota and use Facebook to keep up with friends overseas.

I think it can be super hard to take an ideological position at the expense of functionality (obviously depends on how strongly you hold your views, and what the cost of forgoing engagement with that company/person is).


> I may be critical of industrial farming, large corporate environmental policies and/or Facebook, but I still might buy corn, own a Toyota and use Facebook to keep up with friends overseas.

This could just another example of inconsistency of thought, or frivolousness of opinion.


;)

When it comes to guerrilla warfare, sometimes you have to use your opponent's own tools against them.

[not related but] I don't understand why people keep telling that signal app which uses a central server and phone number as identification and verification is secure and safe to use.

the future is decentralized


Because your threat model is not everyone's threat model.

did you read their home page?

""State-of-the-art end-to-end encryption (powered by the open source Signal Protocol) keeps your conversations secure. We can't read your messages or listen to your calls, and no one else can either. Privacy isn’t an optional mode — it’s just the way that Signal works. Every message, every call, every time.""

Have you observed them get sued for records and be unable to deliver?


And it can only be used on insecure platforms, iOS and Android.



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

Search: