Facebook Has Been Showing Military Gear Ads Next To Insurrection Posts

[Jason]

Facebook Has Been Showing Military Gear Ads Next To Insurrection Posts

WWW.BUZZFEEDNEWS.COM

---

Earlier this week, Facebook employees warned that military product ads were being advertised against news about DC riots. The company did not act.

 

[b_m_b_m_b_m]

[legend]

[SuperSpreader]

tHe AlGoRiThM

[Littleronin]

I mean they know their audience. 

[Mr.Vic20]

3 minutes ago, Littleronin said:

I mean they know their audience. 

Bingo. Social media is all about greasing the economic skids, not providing self aware counter measures to extremism. 

[Spawn_of_Apathy]

They never miss an opportunity to do and encourage the wrong thing for the sake of ad revenue. 

[skillzdadirecta]

I spend as little time on Facebook as possible... fuck it. They are an incompetently evil and irresponsible company DIRECTLY responisble for the fucked up social discourse we are currently experiencing. Fuck that company. The damage they've done to our parent's generation ALONE is unforgivable.

[osxmatt]

Guns were a mistake. Facebook was a mistake. America was a mistake.

 

I'm just having one of them dystopian days.

[5timechamp]

meh, algorithm and all that..

if its popping up on facebook then it means they were searching for it elsewhere

[Joestradamus]

sigh... i am trying to learn new skills so I can move to a different job.

[cusideabelincoln]

Facebook will show me an ad of something that it hears on TV through my phone and I've never searched for.

[legend]

18 minutes ago, cusideabelincoln said:

Facebook will show me an ad of something that it hears on TV through my phone and I've never searched for.

 

Facebook is awful, but I'm pretty skeptical that they're doing something like that, if nothing else but because of the technical challenge of what that entails. There's other reasons you seeing and ad in two places different places for the first time can be correlated.

[Commissar SFLUFAN]

Step 1:  Nationalize.

Step 2:  Purge all data, destroy all hardware.

Step 3:  There is none.

[osxmatt]

2 minutes ago, Commissar SFLUFAN said:

Step 1:  Nationalize.

Step 2:  Purge all data, destroy all hardware.

Step 3:  There is none.

 

Underpants?

[SuperSpreader]

20 minutes ago, legend said:

 

Facebook is awful, but I'm pretty skeptical that they're doing something like that, if nothing else but because of the technical challenge of what that entails. There's other reasons you seeing and ad in two places different places for the first time can be correlated.

Nah they are 100% listening. I went to a concert and after was talking to my wife in an Uber how rude crowdgoers are now, we check our phones and it's the No1 suggested articles on both our phones. Neither had ever searched that, ever. 

[Jason]

29 minutes ago, legend said:

 

Facebook is awful, but I'm pretty skeptical that they're doing something like that, if nothing else but because of the technical challenge of what that entails. There's other reasons you seeing and ad in two places different places for the first time can be correlated.

 

At least on Android the Facebook app is known for absolutely guzzling battery, so it wouldn't really shock me if they're actually doing shit like that, it would explain why the battery drain is so high. 

[Ghost_MH]

15 minutes ago, SuperSpreader said:

Nah they are 100% listening. I went to a concert and after was talking to my wife in an Uber how rude crowdgoers are now, we check our phones and it's the No1 suggested articles on both our phones. Neither had ever searched that, ever. 

 

I don't know if they're listening, but they're certainly tracking your GPS. If you're at concert, Google or Apple know it. Facebook will also know it is you have the app installed.

 

The best way to control Facebook is to simply pass laws that regulate what they can do with your data and what days they can collect. This can be done at the state level, even.

 

Nearly 1.6 million Illinois Facebook users to get about $350 each in privacy settlement

WWW.CHICAGOTRIBUNE.COM

---

Nearly 1.6 million Illinois Facebook users will get about $350 each in a landmark biometric privacy settlement, pending final approval from a California federal judge.

 

[legend]

I would not be surprised if they tap into phone functions they shouldn't. I would be surprised if they're actively listening all the time to parse for comments that would result in good advertisements to display to you. The amount of compute and algorithmic challenge to do that for everyone and to select what is relevant for ads is simply enormous, even for Facebook. So much so that it's deeply hard to imagine that any increase in ad revenue pays for it.

[SuperSpreader]

Oh i activated the max security settings on my fb account two years ago, i deleted it and my account

[Ghost_MH]

2 minutes ago, legend said:

I would not be surprised if they tap into phone functions they shouldn't. I would be surprised if they're actively listening all the time to parse for comments that would result in good advertisements to display to you. The amount of compute and algorithmic challenge to do that for everyone and to select what is relevant for ads is simply enormous, even for Facebook. So much so that it's deeply hard to imagine that any increase in ad revenue pays for it.

 

Yeah, that's why I don't think they're listening. Doing something with the GPS data they periodically poll? Absolutely. That one is easy to parse too. They know there's a concert at x address and x number of people's phones are polling from this address.

[cusideabelincoln]

35 minutes ago, legend said:

 

Facebook is awful, but I'm pretty skeptical that they're doing something like that, if nothing else but because of the technical challenge of what that entails. There's other reasons you seeing and ad in two places different places for the first time can be correlated.

I was playing the original NES Super Mario Bros. at a friend's house.  We talked about it and had that iconic music of the first level on blast.  Then we take a break, I check my Facebook, and on my second scroll pops up an ad with a picture of the original Super Mario Bros. NES cartridge.  First time in years I've played or searched for the game.  If it isn't facebook, then at least google us collecting this data and sharing it with facebook.

[Anathema-]

If they can identify the insurrection posts to monetize them they could instead do anything else. 

[legend]

Just now, cusideabelincoln said:

I was playing the original NES Super Mario Bros. at a friend's house.  We talked about it and had that iconic music of the first level on blast.  Then we take a break, I check my Facebook, and on my second scroll pops up an ad with a picture of the original Super Mario Bros. NES cartridge.  First time in years I've played or searched for the game.  If it isn't facebook, then at least google us collecting this data and sharing it with facebook.

 

This doesn't really address my objection to it  Life is filled with bizarre seeming coincidences even when there aren't reasons for them to be correlated. It would be more unlikely that you never experienced any and people are really good at noticing when they occur. Even with Mario, there is probably still better correlated information for why that kind of ad is something they would display to you, and every once in awhile a hit will feel oddly timely. 

 

Simply put, I just don't think remarking on the times you've observed something weird like that is good evidence when faced against the reasons it isn't likely to be realistic.

 

Of course,  there are plenty of other reasons we can declare Facebook a villain  

[Jason]

13 minutes ago, legend said:

I would not be surprised if they tap into phone functions they shouldn't. I would be surprised if they're actively listening all the time to parse for comments that would result in good advertisements to display to you. The amount of compute and algorithmic challenge to do that for everyone and to select what is relevant for ads is simply enormous, even for Facebook. So much so that it's deeply hard to imagine that any increase in ad revenue pays for it.

 

Is it feasible to have individual phones doing the parsing instead of pushing the audio back to a central server to do it?

[Jason]

Just now, legend said:

 

This doesn't really address my objection to it  Life is filled with bizarre seeming coincidences even when there aren't reasons for them to be correlated. It would be more unlikely that you never experienced any and people are really good at noticing when they occur. Even with Mario, there is probably still better correlated information for why that kind of ad is something they would display to you, and every once in awhile a hit will feel oddly timely. 

 

Simply put, I just don't think remarking on the times you've observed something weird like that is good evidence when faced against the reasons it isn't likely to be realistic.

 

Of course,  there are plenty of other reasons we can declare Facebook a villain  

 

I know the plural of ancedoate isn't data, but there are plenty of similar reports out there.

[legend]

3 minutes ago, Jason said:

 

Is it feasible to have individual phones doing the parsing instead of pushing the audio back to a central server to do it?

 

I'm going to go with no *especially* not without it being extremely transparent with hard evidence being produced of its existence.

 

2 minutes ago, Jason said:

 

I know the plural of ancedoate isn't data, but there are plenty of similar reports out there.

 

There are plenty of reports of weird observations in evidence of divine miracles too  

[Anathema-]

3 minutes ago, Jason said:

 

I know the plural of ancedoate isn't data, but there are plenty of similar reports out there.

 

I mean, I know people who have experienced it but it could just be another Seattle Windshield Pitting Epidemic. 

[legend]

Let me also say this: if you think it's listening all the time and displaying ads right afterwards, it should be really easy for you to test this hypothesis and we'd also have really strong evidence of its existence already as a consequence.

[Joe]

3 minutes ago, legend said:

Let me also say this: if you think it's listening all the time and displaying ads right afterwards, it should be really easy for you to test this hypothesis and we'd also have really strong evidence of its existence already as a consequence.

 

What are we setting p at?

[legend]

6 minutes ago, Joe said:

 

What are we setting p at?

 

P-values are the devil  But I think we can probably work out what sufficient demonstrations of control would be for anyone that want's to articulate what they think the scope of the effect is.

[cusideabelincoln]

30 minutes ago, legend said:

 

This doesn't really address my objection to it  Life is filled with bizarre seeming coincidences even when there aren't reasons for them to be correlated. It would be more unlikely that you never experienced any and people are really good at noticing when they occur. Even with Mario, there is probably still better correlated information for why that kind of ad is something they would display to you, and every once in awhile a hit will feel oddly timely. 

 

Simply put, I just don't think remarking on the times you've observed something weird like that is good evidence when faced against the reasons it isn't likely to be realistic.

 

Of course,  there are plenty of other reasons we can declare Facebook a villain  

 

I don't completely disagree with that hypothesis because it is impossible for our memories to track and record information in that manner, as no doubt there are countless times irrelevant ads go unrecognized.

 

But this instance was especially odd because it wasn't explicitly an ad.  But in the ad space there was a picture of the Mario NES cartridge I had just played.  Very odd to show a cartridge that isn't even made anymore

[LazyPiranha]

There’s no way they’re constantly listening in.  It’s observation bias and other algorithms combining with other data.  Shit like this has been going on for years.  When my wife got pregnant with our first son, we started getting baby shit in the mail randomly before we told anyone.  Stores have flagged certain purchases that new parents have in common before they even buy diapers and other shit, and once they get flagged you start getting it.  
 

The signal to noise ratio on listening in on everything would render it useless.  Your search data at the very least contains stuff you were actively interested in even in a negative context which is infinitely more valuable than a word you heard in passing.  
 

It’s same shirt as anything else in life.  I was randomly listening to old episodes of podcasts I like and there was one about the Spice Girls.  Randomly throughout the week, the Spice Girls came up a couple of times.  It’s obviously a coincidence but I never would have thought anything of it had I not listened to that podcast by chance and had them on the brain.  When your ads have no personal connection you never notice them, they just ping your brain when there seems to be a reason.

[Guest]

I heard an explanation of how some of this occurs that seemed very plausible, while maybe not answering every strange instance.

 

They said because of GPS and friends lists and such, these companies know when you are in proximity to somebody you know. The probability that you and said friend had conversations about something that one of you had recently searched is actually very high, so there is a good chance that while you did not search for that thing, your friend had recently done so and the system served you ads on the decent chance you and friend talked about that thing.

 

This made sense to me of a particular situation a few years ago where I had a conversation with an older lady who was talking about getting her husband to the VA hospital and within 10 minutes of saying by to her I was flooded with VA benefits ads. I initially had the “they must be listening” thought, but that actually made a lot of sense and is obviously far more practical than analyzing audio recordings.

[Anathema-]

I mean if it were a thing it would happen a lot more often.