I was totally skeptical when I first heard stories about people's feeds being curated to suit what they were talking about in front of their computers, but then a couple of things made me rethink my views.
I was in the car with my friend once (both our phones were in the car with us too), and we were driving and it started raining. His wiper blades were shoddy and doing a bad job of clearing his windscreen. We had a 5 minute discussion on cleaning blades, where was the best place to get a replacement, etc. Neither of us picked up our phones, or did a search on 'wiper blades', nor asked Siri etc. about nearest places to buy.
But the next day, for the first time ever, I started seeing ads for wiper blades on my feed. They disappeared after a few days, but it was spooky, as it was for an item that I have never ever discussed, or seen on my feed before, or since.
Another anecdote. I was sitting working in my office the other day, when my son came in to talk to me about something. While chatting, he picked up a bass guitar that sits behind my desk and started playing the riff from 'Seinfeld'. I may have said something like "Oh, the Seinfeld theme - you know that was played on a synth and not an actual bass guitar?". Our conversation though, was about something else altogether.
A day later, on my feed, I see an ad about "How they played the Seinfeld theme" [0]. A totally specific ad like that, within 24 hours, totally unrelated to any searched I had done in my past. This one too disappeared within a couple of days.
> But the next day, for the first time ever, I started seeing ads for wiper blades on my feed. They disappeared after a few days, but it was spooky, as it was for an item that I have never ever discussed, or seen on my feed before, or since.
Could be something like this: Facebook knows from location data you were driving/riding together. Soon after your ride, your friend searches for wiper blades. Facebook then shows you wiper blade ads based on that knowledge.
The problems with this 'psychological explanation'
1. It is just as unscientific as saying that Facebook MUST be listening, without finding any evidence in the workings of phones, app source code, etc. Almost by definition, there is no experimental evidence for the Baader-Meinhof Phenomenon.
2. It posits that everyone who has experienced this is unsophisticated and ignorant about cognitive distortion. Clearly, some people who believe that Facebook is listening to them talk also believe that there were no planes on September 11th, etc. However, many many people today have a good understanding of psychology, are well aware of (basic) cognitive biases, have thought about these experiences. People also have experiences of unlikely events but which have well understood or easy-to-estimate probabilities, which didn't occur much historically. Lots of people can distinguish quite well between their own illusory biases and events which are coincidental, but not 'in the mind'. Some of these people seem to also believe their speech has found its way into advertisers' profiles.
> 2. It posits that everyone who has experienced this is unsophisticated and ignorant about cognitive distortion.
Not really, it just suggests people aren't perfect statisticians and perfect at recording information. Which we aren't.
> People also have experiences of unlikely events but which have well understood or easy-to-estimate probabilities
The probability of seeing a relevant advert is not something we can easily estimate.
> Lots of people can distinguish quite well between their own illusory biases and events which are coincidental,
I would be shocked if this was even a reasonable percentage of people. I am certainly not a perfect innate statistician with perfect recall, and I think most other people are not either. Humans are just bad at this stuff.
> which didn't occur much historically.
I'd be shocked at anyone with a very solid recall of all the adverts they see.
Without strong evidence, this seems like a perfectly reasonable base view. It'd be a complex, expensive, power and data hungry way of targeting adverts, extremely controversial and possibly illegal. There's no evidence they're actually doing it, no whistleblowers, and many of their adverts are just terribly targeted.
People here are bringing up examples like a young couple being advertised a pregnancy test. Hardly requires invasive monitoring of a billion people to achieve.
I would disagree with that phenomenon. In direct contrast, my awareness of these ads kicked into overdrive after seeing them for the first time, so I was hyper aware and looking out for them afterwards, which is why I noticed they disappeared a day or two after popping up.
Since then, my feed has been a bland sea of adverts for mattresses, knives etc. that almost all of my friends complain about seeing.
I don't use ad blockers, due to my development work in web apps etc. I like my browsers to be as stock as possible, and I do actually at least glance at nearly all ads I see in FB etc. just to see how their targeting is being skewed by my search history or messaging online.
These were two quite distinct cases of ads that I've never seen before, which popped up within 24 hours of me talking about (but not doing anything else online related to them) and then disappearing soon after.
Sure, an anecdotal sample size of 2 could just be pure coincidence, but something about it still doesn't sit right with me.
It would be weird if that sort of coincidence never happened. Two incidents of this happening to you ever doesn’t sound like much. Maybe even a bit low.
If they really are using audio for ad targeting like they do with search history, then it should be easy enough to prove it by experiment.
Sure. I am used to, nay, I _expect_ ads for software tools and music gear given my work and hobbies in my life, as I frequently talk about those things to my friend on FB and IRL all the time.
But wiper blades is so off the charts for me, that I was totally surprised to see about 3 ads for them the day after I had a discussion in the car. I don't even go to sites remotely associated with the automotive industry. Not even locality based, as it was a full 4 months out of the real 'wet season' we have later in the year. (NOTE: Not even in our wet season did I see random ads for wiper blades).
Having said that, I won't be surprised to see more wiper blade ads now that I have been talking about it on HN - who knows if they have linked up the data from online forums back to FB etc.? The handle I use here I tend to use on nearly all online forums I belong to.
> I don't use ad blockers, due to my development work in web apps etc. I like my browsers to be as stock as possible, and I do actually at least glance at nearly all ads I see in FB etc. just to see how their targeting is being skewed by my search history or messaging online.
You are doing yourself a disservice. Have two users (or at least two browsers), one for work, which you keep ad-blocker free, and one for life, which you shouldn't.
Just a note, not trying to pick your anecdote, but the experience of something seemingly occuring frequently out of nowhere is exactly the baader-meinhof phenomenon. You’d need to demonstrate a causal link to firmly disprove it, which would assumably be difficult in this case.
I think that it's more likely that the companies that manufacturer wiper blades run targeted ads in places where it's raining and that your son looked up the sheet music for the Seinfeld theme on his phone and Facebook thought you would like it because you are Facebook friends or because your GPS coordinates were next to each other.
Fair call - this reasoning I can relate to. I was the first spot of rain in a long dry season spell, so it stands to reason that drivers in the area were highly likely to have dried out, unserviceable wiper blades just when they needed it.
The link about friends search history and close GPS co-ordinates also makes sense.
I had a similar experience in June when I was on a car journey and we got a crack in the windscreen. My phone was out (was using it to navigate) but it wasn't touched at all. It was a Sunday, and I didn't bother researching how to get it repaired.
Later on when I checked my phone I started getting ads in the Facebook feed for windscreen repairs.
It’s useful to specify iOS or Android in the anecdotes. iOS will show a banner at the top of the screen when an app in the background is using the microphone. On Android, the last I’ve seen, this isn’t the case.
I was in the car with my friend once (both our phones were in the car with us too), and we were driving and it started raining. His wiper blades were shoddy and doing a bad job of clearing his windscreen. We had a 5 minute discussion on cleaning blades, where was the best place to get a replacement, etc. Neither of us picked up our phones, or did a search on 'wiper blades', nor asked Siri etc. about nearest places to buy.
But the next day, for the first time ever, I started seeing ads for wiper blades on my feed. They disappeared after a few days, but it was spooky, as it was for an item that I have never ever discussed, or seen on my feed before, or since.
Another anecdote. I was sitting working in my office the other day, when my son came in to talk to me about something. While chatting, he picked up a bass guitar that sits behind my desk and started playing the riff from 'Seinfeld'. I may have said something like "Oh, the Seinfeld theme - you know that was played on a synth and not an actual bass guitar?". Our conversation though, was about something else altogether.
A day later, on my feed, I see an ad about "How they played the Seinfeld theme" [0]. A totally specific ad like that, within 24 hours, totally unrelated to any searched I had done in my past. This one too disappeared within a couple of days.
[0] - https://twitter.com/dsabar/status/924763005014831104