Ask HN: Is mental illness an ad-targeting category?

[hnthrow_asdf]

Hi HN, hope this topic is ok. I'm looking for commentary or advice from readers familiar with ad systems.

A few years ago I got involved with the wrong wealthy, connected crowd and many weird/violent things happened. The story is strange, so people assume mental illness. I'm seeing a counselor, taking pills that don't seem to help except make me fat and sleepy, and doing my best to get back to dev work before my funds run out.

I wanted to talk about my internet experience, though. I used to get ads for food, clothes, consumer tech. Now I get strange ads which are really not helping my situation.

On Twitter, ad with only a few views about "losing a game, checkmate the cost is your soul".

On IG, which I use to look at cat pictures and my friends, I get ads for Freemasonic end-of-life care, occult ritual materials, bizarre poetry sometimes containing my name/usernames, artwork with the tagline "Maybe you aren't paranoid and people are following you ;)", going blind, car accidents, ALS, cancer, organ failure. Or, my personal favorite when I've managed to distract myself from my troubles for a few hours, paraphrased "DID SOMETHING EXTREMELY BAD HAPPEN TO YOU AND YOU CAN'T STOP THINKING ABOUT IT?". I usually see that one multiple times a day.

I've shown these ads to other people; I don't have any history of seeing physical things others don't. I try to scroll past them fast but usually still see what they say. I usually don't take screenshots because I don't want the pause to register as engagement.

I get that most people would just ignore this stuff. I would have in the past. I don't know how to explain the change in perspective after you've been in life-threatening situations, had your windows smashed, or had weird totems left in your yard, etc etc. It's not a stretch to think "maybe this IS relevant to me, or a message warning me: bad thing going to happen!". Regardless, I want to avoid interacting with content like this.

I have long been opted out of search history or ad personalization on G/FB/AMZN/Twitter/Instagram/Verizon, and have repeatedly clicked "hide ad - irrelevant or not interested". "Why am I seeing this ad?" says "(wide age group) in your area", yet I've asked and my peers don't get ads like this.

I think I'm privacy savvy somewhat. Logged out, incognito/Ublock/Ghostery. Few apps, all permissions denied. DDG. Unique usernames, lurking or throwaways. No prescription rewards. I only visit a few websites. I don't share my computer. No other devices listening in the house.

The only information leaks I can think of are: * credit card billed for telehealth * short text journal on dropbox * friends may have discussed me/topic in their chats * a friend asked on Messenger why I disappeared, I said something like "I might have schizophrenia. Some violent situations happened and I'm struggling now."

I know, Messenger wasn't E2E, but I didn't realize it might flag me so easily. I haven't done anything else to indicate my interest in this topic; it's not like I'm posting in "targeted person" forums, buying anything other than groceries, or watching horror movies.

Even if the explanation is just "machine learning black box parses chat, determines unlabeled cluster demonstrates increased engagement with thrilling, morbid, suggestive, edgy content", well, I explicitly rejected opting-in to that. The supposed "opt-outs" have little effect, at best hiding one specific ad but not actually updating to the point of "I don't want to see ads about PTSD, dismemberment, horror, paranoia, accidents, tragedy, eschatology, or the occult, please start showing me only normie ads again".

I'm not sure what to think at this point, short of going offline or quitting social media. I did that in the past but I am trying to keep my weak network alive. Anyway, I'm curious to hear if anyone has feedback or advice on this situation. Am I missing something, mistaken, doing something wrong? Thanks!

[mactyler]

"I have long been opted out of search history or ad personalization on G/FB/AMZN/Twitter/Instagram/Verizon, and have repeatedly clicked "hide ad - irrelevant or not interested". "Why am I seeing this ad?" says "(wide age group) in your area", yet I've asked and my peers don't get ads like this."

I think this could actually explain your situation, if you give ad feedback like that too much you might run out of things for "the algorithm" to reasonably show you and you're getting fallen back to some sort of odd pool of low quality or new account test ads of some sort that are pre mainstream since you give feedback often or hitting some sort of weird bug after a point where it tries to find something to show you and then grabs for these weird ones somehow less shown and therefor less reported by other people?

I don't know the right answer, but I feel like something in this area could totally explain what you are experiencing. If you want to test the theory maybe make a new account on a service and don't give feedback on anything and see if you end up with mostly normal stuff?

[hnthrow_asdf]

Good point. I meant that I give feedback to try to hide the weird/creepy ads, though. I've never clicked "stop showing me Ritz crackers or Crossfit ads" just because I'm not interested.

[rdtwo]

Here is what i suggest you do.

Go on discord and buy famrmed gmails with a 30 day proxy. They are 5-10$ A piece resellers use them a lot. Use that account for 30 days then dump it and get another. If you still see wierd shit it’s because you are searching wierd terms or because you are visiting wierd sites or perhaps you are suffering from a untreated mental health condition. The metal health stuff can cause heightened awareness of things like this that other folks gloss over.

[Traubenfuchs]

You do realize that serious mental illness would make you think that you really confirmed with friends that those ads are real even when they are not?

They sound so outlandish and insane (poems using your name) that I am sceptic.

You should screenshot and compile those ads in form of a blogpost or on twitter.

I am rather mentally ill and all I get are ads for those useless chatbot apps and meditation apps.

[hnthrow_asdf]

Sure, I mean anything's possible. If I was that out of it though I would expect consequences to arise - feedback from family/doctors, arguments, alienated relationships. In any case that wouldn't be something HN could help with, and I am taking the strong antipsychotics that are supposed to quell stuff like that.

It's been months since I saw the poetry specifically and I don't have pics of that, but here are a few ad pics I did save:

https://imgur.com/thmf8hU https://imgur.com/rbRRxmk

[Pigalowda]

Aren’t those recommended posts from your instagram feed?

Hovering over those and examining and studying them in anger/fear/horror is still engaging with the ad. Won’t the algo serve more similar content if you are spending time looking at it?

[hnthrow_asdf]

Like I explained with the initial post, 99% of the time I try to avoid exactly that - averting my eyes, scrolling and using peripheral vision to see when the ad's gone. Not taking screenshots or pausing because that's engagement. I've tried both tactics of pausing to click "Hide ad - irrelevant", or just scrolling past as fast as possible.

I did, twice apparently, stop and take screenshots because my therapist wanted proof. If I get an ad for pistachios or moisturizer these days I do the opposite - pause, click the ad.

[daniel-cussen]

Those are confessions that those responsible for these messages are targeting you. You haven't provided irrefutable proof they are real, but I chose to believe them.

They want to provoke you to claim there's a conspiracy, because in this cultural context that claim is never treated with respect, and forfeits the few real protections society provides against this.

Conspiracies abound, it's not remarkable. The Bible is full of them. History is full of them. Courts deal with them on a daily basis.

Just think, how many sports people carry out alone, and how many are team sports? In the same way, how many forms of harm can be carried out alone, and how many can be carried out in a group?

Read up on the law in this regard, it's a totally different story if you say "there's a conspiracy" or "they're out to get me" (never ever say that, that always backfires), or if you say "I found evidence of meeting of the minds" or "I'm seeing messages that are gloating that they are harming me". That's a totally different story. And you need to have higher status first, society won't help you with this normally. So work on your presentation as much as you can. I see your sentences are grammatical and the development of the logic is linear; this is good.

I actually would get off social networks, or limit it to the max. Try to get offline as much as possible. It's just really hard. But for instance you can start using well-litigated forms of communication, like land lines and the mail, and just paying what it costs for long-distance phone calls. Read an actual newspaper instead of news aggregators, not a whole lot of javascript and cookies on the print edition. All that encryption shit, like hey there's this great new app that covers the exact vulnerability you have, that sucks because then you're trusting the dude who made the app, too, and who knows what bugs he left in his code. Only the absolute best can be trusted, and not completely.

This stuff is expensive. It is all expensive. It's the cheapest thing to get everything for free and give up your data, basically (and especially in your case, I think you'll get a little laugh out of this despite your suffering if you follow my dark humor) you're selling your soul, lol.

One thing you can do that's cheap is when you're online, mislead. Don't always click on links you're actually interested in, search for things you don't give a shit about, be dishonest in your browsing basically.

It also sounds like your case isn't hardcore. It's persistent, and it's everywhere online, but it's just some guy or guys doing cheap, easy stuff that anyone can do with like four dollars of targeted ad spending, and two dollars of data bought off a broker. It looks like they're just measuring how well their psy-ops work by seeing how you react. They're not actually going to send someone to come get you where you live, that's expensive and dangerous and you're not that important to them.

I do some of this stuff as a hobby, I just have fun with security. You can claim the same thing, that you're just doing it as a passion.

[bryan_w]

If you've turned off ads personalization, on IG/meta then you're just getting the bottom of the barrel ads -- no conspiracy or anything, just you get thrown into the randoms pool and those are the ads that big bidders place in that pool.

Perhaps turning personalization back on but using it only in a Firefox container with only normie browsing might be a good middle ground.

[Daimanta]

People with tendencies of spending unhealthy amounts of money on games could be classified as having a mental illness. Instead they are indeed targeted and known as "whales".

[missedthecue]

In Facebook, you can target people based on groups they're a member of. In Google, you can target keywords which, of course, can include keywords related to mental health.

[auganov]

It's trivial to create a lookalike audience based off your site's traffic, even if there's no pre programmed category. You just put up some content that you know for a fact will attract a given group, promote it to a wide audience, and then create a lookalike audience based on people who clicked the link (or some more specific conversion metric).

That said, the examples of ads you've posted seem more like attempts to get some attention by acting strange. Perhaps you're being shown low-bidders and that simply exposes you to total randoms rather than the usual polished ads. I experimented a bit with rock bottom bids and the engagement would often be from bizarre looking accounts. I'm guessing it may go both ways.

Even if advertisers are targeting you on this, they still need to outbid others. You don't need to completely change the data they have on you. You only need one new signal that attracts higher bids from some other category of advertisers. Merely changing your IP address or mobile device might be enough.

Also if you really saw your name or username in these ads then why worry about being targeted for a history of mental illness related stuff. If that's true then you've got a much bigger problem.

[blessingdusk]

Yes. Take it from the guy who helped design the very software that targets you. Christopher Wylie worked for Cambridge Analytica and I recall reading from his book that psychographics are definitely taken into account when targeting people with political advertisements; I am sure that this applies to ads for consumer goods. In an excerpt in the book, the author discusses how there are 5 psychographic "categories" that are taken into account when selecting which ads to show to people.

"militarism" — guns and shooting, martial arts, crossbows, and knives "violent occultism" — drugs, black magic, paganism "intellectual activities" — singing and making music, foreign travel, the environment "credulousness" — the paranormal, flying saucers "wholesome interests" — camping, gardening, hill-walking

Individuals are also rated on openness, conscientiousness, extraversion, agreeableness and neuroticism.

The idea is that with the massive amounts of data one produces from surfing the Web or even merely being online, the software can decide which ads to show to which type of person. Do they engage in "violent occultism" and "credulousness" such as unironically browsing and discussing ideas in QAnon forums? Well, they're probably likely to engage with and respond to political advertisements, so the algorithm feeds them political advertisements on the assumption that they will engage. (This is a cloudy example, but the idea remains. Data that is remotely useful will be used. This is Big Data.)

I think what is interesting to note is that in their research/experiments, they've specifically targeted the Incel community specifically for this reason. Demographically speaking, Incels are typically young adult males "who can't get laid" (verbatim) and are highly susceptible to targeted advertising. If you want to read more about this, there is a book written about how Cambridge Analytica helped Donald Trump win the 2016 US elections by Christopher Wylie. I highly recommend it.

[zenexer]

I love tinkering with various ad personalization settings and ad-blocking technologies just to see how it changes the ads I see.

Facebook (including Instagram) is my favorite. For the most part, when you tell it to stop personalizing your ads beyond basic information, it obliges, although it does try to make that process difficult.

If you disable most targeted advertising on Facebook/Instagram, you’re going to see generic ads targeting the average person in your region—and it’s enlightening to see what those ads are. As you move between regions, the ads change dramatically.

Where I live, the ads I get are pretty mundane. However, visiting family members less than two hours away yields some interesting results. They’re in a relatively poor, conservative part of an otherwise affluent and liberal state, so there’s a lot of inequality and frustration. Violence and drugs are rampant, and the police seem to love exacerbating the social problems in the region.

What sort of ads do I get there when I have personalization disabled? Pretty much exactly what you described, although a bit less cult-oriented and more on the blatantly-violent side.

Want to test this out for yourself? Head to Facebook and remove yourself from all advertising lists and disable any personalization that you can. Then, head to the Facebook Marketplace. Set the location radius relatively small and move the location around a bit; assuming they haven’t changed it, you’ll see ads targeted to the average demographic in that location. It works best when current events are causing a significant political divide; for example, there were some fascinating ads around January 6, 2021.

[hollowpython]

What do you think about using https://trackthis.link/ ? It opens 100 tabs of random stuff to confuse trackers.

(It will ask for confirmation first, so you can click on the link and read more about it)

[kant_adorn_arts]

> * credit card billed for telehealth * short text journal on dropbox * friends may have discussed me/topic in their chats * a friend asked on Messenger why I disappeared, I said something like "I might have schizophrenia. Some violent situations happened and I'm struggling now."

All are potential sources of 'ad revenue'. When I was having a long distance relationship, every site tried to sell me 'Russian Brides' etc, it was highly offensive but it went away - I also quit certain sites and mail/messenger and now always UBlock everything

[vba616]

There are a lot of (de)personalization settings for Google, and Facebook, and Microsoft.

If you haven't spent like an hour or two, you probably haven't found them all. There are screens, and screens, and patterns to discourage you.

And if you haven't checked them in a month or two, there might be new ones.

My belief is that the game isn't that unfair, that they do allow you to turn stuff off, but it's difficult enough to ensure statistically most people won't.

I believe that FB Messenger is plenty enough explanation for your experience, by itself. Why would they not be processing every word?

[type0]

Sure it is, just a recent one https://news.ycombinator.com/item?id=30121581 on "Suicide hotline shares data with for-profit spinoff, raising ethical questions" https://www.politico.com/news/2022/01/28/suicide-hotline-sil...

[dartersnek]

You say you use uBlock. Are you using it to block ads, or do you have it turned off on Twitter, Instagram, etc.? If you want to

> avoid interacting with content like this

an ad blocker is definitely a good option.

[hnthrow_asdf]

I use uBlock for my laptop browser, and Safari content blocking on my phone, and have looked into stuff at the router level. I don't disable ad-blocking for any site.

The issue is primarily with mobile apps, which don't support ad-blockers and serve ads from their domains. I just found out about Nitter the other day, and didn't realize that you could now use Instagram/DMs from a web browser instead of using the app, so I will try that going forward.

This possibly affects things like YouTube recommendations too, not ads - I use it logged out and mostly search for music but if right now I open the site my top recommendation is a meme video "no amount of therapy will make this scene ok". I don't see any cookies and am not sure how to reset whatever session id for a fresh start.

[PegasusxBeaver]

uninstall, remove files, reinstall. Alt. replace browser consider browsing in VMs consider fresh start hw/sw & account hygiene

[Crazyontap]

How about you turn ON all ad personalization and just search for cat food, funny jokes websites and a bunch of random noise.

I think there is a chrome extension that will do this, i.e search random shit to poison all trackers. Maybe it will help you defeat it.

Learned this in school, it far easier to turn a F into E, I to T, etc by overwriting it with a pen and making complete nonsense of things than trying to erase it completely, which is harder and leaves more traces.

[rdtwo]

Farmed gmails with aycd autosolve do this for you. Inject months of nose then use it and dump it. Just gotta use a proxy

[PegasusxBeaver]

You are not only one and you are not crazy. There ARE people behind this. Domestic or foreign. Can't say lots. Consider SmarterEveryDay 263 +. Rethink all online relationships Consider https://en.wikipedia.org/wiki/Internet_troll#Psychological_c...

[yuppie_scum]

Switch to DuckDuckGo and call me in the morning

[jdr23bc]

Sorry to hear that. It sounds like you are being served some weird targeted ads, but you also are also dealing with a stressful situation.

Taking a media break until your situation resolves or your targeting profile gets some less weird stuff in it might be good.

[Cr0s]

Like a another comment suggested, I would recommend using an adblocker. It might not be the answer you are looking for, but it seems like the fastest, most immediate solution.

[barrystaes]

Since using adblockers i see less ads, nearly none. And on unprotected devices and networks i see strange ads, unsurprisingly. I can tell you all are shockingly irrelevant and repulsive like the examples OP gives.

So full circle: if you avoid them you get less but also worse ads.

[hedora]

This is unfortunately too common. Abuse prescription drugs? Expect to constantly see ads for those drugs. For some reason ad targeting thinks I have an unpleasant disease or two that I don't.

I've had good luck enabling all the tracking and searching for high-end lifestyle stuff. The prospect of selling luxury goods to the ultra wealthy is jucier to the ad auctions than exploiting a measly few thousand a month for prescription meds, and the ads for that stuff is eye candy anyway.

After kicking it a few times like that, I re-enable the ad blockers.

On a related note, Trump famously targeted the mental illness category with conspiracy theories and libel. It probably won him the presidency.

If I had my way, the entire surveillance capitalism industry would be banned (including targeted ads).

[hnthrow_asdf]

I hadn't thought about skewing the ad auctions that way, thanks. I used to get ads for fancy river cruises, Peloton, high-end furniture cookware etc and will try searching for that stuff again, but am guessing the data brokers know I no longer earn an income. Failing that maybe pretending to be looking for payday loans, lottery, coupons might be "lucrative" enough.

[hedora]

The canonical example I've heard of from Googlers is "yoga pants", though that's out of date. Go with something that's trendy this year.

With any luck, they'll think you have a rich significant othet. I somehow accidentally convinced them I have a mistress with a taste for $600 machine washable silk pyjamas (there is apparently a company that specializes in only that!)

It helps to clear cookies, etc immediately before kicking the recommendation engine.

That seems to significantly lower the weight it puts on the prior. Instead of being a native Spanish speaker, I now know Vietnamese.

(Of course, all of these inferences about me are wrong.)

[rdtwo]

Aycd and kylin will search popular terms with your Google profile. It’s called farming usally too increase capcha score but it’s also useful to improve the noise floor to make it hard to target ads.

[fuzzfactor]

Well I'm not a consumer so anyone targeting me would be confirmed as somewhat misguided mentally.

[macrolocal]

This post has class action liability written all over it. :(

[YXNjaGVyZWdlbgo]

I got targeted on my bipolar research google account on google search and youtube.