Go over to friend's place and watch the ads they get, you'll get a good idea of what kind of health concerns or illnesses they may have.
So far, in situations where it wouldn't be rude to ask, I've been able to determine with pretty good accuracy that at least someone in the household has the advertised health concerns.
You can also get an idea of their financial situation, given what buckets advertisers put them in and what they're advertised, as well.
Similarly, advertisers know when you're at friend's location, or elsewhere, and may show ads tailored to your profile.
I wouldnt jump to conclusions too fast. I'm a very thin person, but I get a lot of ads for GLP-1. My doctors have always said I need to gain weight, so I can assure you I'm not searching for weight loss solutions. Nor am I diabetic.
GLP-1 drugs are such money makers that it's worth advertising them so that you 1) know they exist 2) might recommend them 3) might take them yourself at some point.
A month's supply of Wegovy/etc costs nearly $2k.
Same thing with the monoclonal antibody drugs, you can be looking at tens of thousands of dollars per month for a single patient.
You also see this with luxury brands, that advertise to segments that cannot afford their products, in order to build image of luxury and prestige.
That said, I asked my friends/family if they had <not embarrassing or too intrusive> condition, and more often than not, someone in the household had it.
To be fair, for GLP-1 drugs I see enough of them in untargeted ads (eg. billboards) that I just assume they're carpetbombing everyone rather than doing precise targeting.
A few months ago, during a boardgaming session, I used a friend's computer to play some music. I made fun of the AI porn ads he had playing whenever I visited a YouTube page.
I only later realized that he's not using an adblocker, so this might actually be tailored to his profile.
If you've been to a doctor's office, clinic or hospital lately, you've likely signed away the right for them to share your information with their partners during your initial intake.
Connect to their WiFi, or your WiFi is off and still scanning for (E/B)SSIDs and thus can triangulate your location (this is what Google location services does), same thing with Bluetooth, apps/OS will grab your GPS location, and your phone carrier is selling your live and historic location data to anyone who wants it, including your government.
It’s a meme that will never die, but there’s no proof it ever happened:
> This story doesn’t even show that Target tried to figure out whether the girl was pregnant. It just shows that she received a flyer that contained some maternity items and her weird dad freaked out and wanted to talk to the manager. There’s no way to know whether the flyer arrived as a result of some complex targeting algorithm that correctly deduced that the girl was pregnant because she bought a bunch of lotion, or whether they just happened to be having a sale on diapers that week and sent a flyer about it to all their customers.
> I wish we had an update on what the situation looks like today.
My wife and I spent 3 years in fertility treatments, which involves a lot of online activity similar to that of someone newly pregnant (buying pregnancy tests, researching symptoms, etc).
We were constantly bombarded with pregnancy related advertising, it really ramped up after the first year. Tons of "congratulations" cards, coupon books, "new mom" magazines, up to and including unsolicited shipments of formula and branded blankets.
So to answer your question, it's still happening, and it's disgusting. I strongly suspect Carrot Fertility sold our information because the peak of it all happened a couple months after I gained access to them through my employer.
(We did eventually succeed, our baby is nearing 6 months)
They have even better psychometric profiles on everyone now than they did previously. This is why Zuckerberg can confidently tell people during an interview that he knows they want at least 15 friends¹ & he is going to deliver those friends to them w/ his data centers.
I would guess that purely observational psychometrics completely fail to predict how people will respond when challenged or stressed. I think they're trading on fools gold.
Observational psychometrics over a long enough timeframe (e.g. social media profile lifetimes) probably include periods of challenge or stress, which may help the predictive behaviour.
"before her family knew" is a pretty low bar to clear, especially if the daughter was actively trying to hide the pregnancy (eg. by wearing baggy clothing). Moreover if we're taking the example of this specific story, where the women presumably knew she was pregnant (as opposed to the more sensational story of "Target figured out a women was pregnant before she even knew!!1!" that also makes the rounds), it's not hard to imagine how Target might be in a better position to infer her pregnancy without being galaxy brained or creepy. Take the examples given in the article:
>Take a fictional Target shopper named Jenny Ward, who is 23, lives in Atlanta and in March bought cocoa-butter lotion, a purse large enough to double as a diaper bag, zinc and magnesium supplements and a bright blue rug.
>The manager apologized and then called a few days later to apologize again.
>On the phone, though, the father was somewhat abashed. “I had a talk with my daughter,” he said. “It turns out there’s been some activities in my house I haven’t been completely aware of. She’s due in August. I owe you an apology.”
That quote, especially the "there’s been some activities in my house I haven’t been completely aware of" part makes me think she knew and was trying to hide it. There's also the expectation that if she really didn't know, that person writing the article would put that detail in, given how extra sensational it would make the story.
I continuously get Spanish advertisements despite being unable to speak Spanish. (The only foreign language media I visit on YouTube is Anime/Japanese and Germans stuff).
There are surely large numbers of dumb cases of being mispredicted out there in.
The miss rate is still wildly high based on the ads in my instagram feeds.
But the really-good hits are probably tough to notice. They won't stand out as "boy this is a stupid ad" and even if I just scroll past a well-targeted ad, it's probably doing its job of making that company/product a bit closer to the top of mind...
> Yeah AI is taking a lot of damage when the actual problem is capitalism
And I think it's fair to to throw flak in AI's direction, if what it does is make capitalism less tolerable by removing some of the "inefficiencies."
While apologists for capitalism have done a good job of pushing me towards wanting to burn it all down, I doubt that's in the cards any time soon and limits on AI technology are far more likely.
Seems ironic. Perhaps we should celebrate AI as the accelerator of capitalism’s ultimate unworkable demise
We aren’t getting regulations on AI. The military industrial complex includes the tech industry now. It’s an existential race to beat China.
The sad reality is that for all of our potential futures, we aren’t getting the Star Trek post scarcity utopia. Our onboard ship computers aren’t generating Earl Grey, hot, they are generating trillionaires on one hand and poverty on the other.
It annoys me that big-tech marketing has made most people believe that "personalised advertising" means they get ads which are more "useful" to them. I regularly see people opt in to personalised advertising because of this.
Personalised advertising is about collecting every detail about your life and using it to extract as much money as possible from you. AI advancements might be making this even more effective but it's been this way for a long time.
This is one of those places where it's worth disentangling the status quo from an optimal alternative.
Currently, every factlet you leak to one of these systems poisons them toward their profit (and almost unanimously against your best interests). Advertising draws your attention away from the products that would make your life better (cheaper, heathier, tastier, whatever) and toward profitable alternatives.
It doesn't have to be that way though. You physically don't have time to research every thing that exists, or even to hear about every possible product in passing. Supposing some of those would improve your life on average, is word-of-mouth really the most efficient way we can come up with to tell you about the things you do actually want to spend your money on? In theory, this is a great business -- customers want to spend money, companies want to sell things, and the information/discoverability asymmetry means that companies are inclined to get word of their products out there with customers _also_ wanting to hear about those products (if they're sufficiently personalized). If "advertising" were good enough, I'd pay money for it.
That only falls apart because of a lack of trust and ethical behavior. Instead of being treated like the information market it is, it's thrust onto individuals to try to prey on their weaknesses.
> It doesn't have to be that way though. You physically don't have time to research every thing that exists, or even to hear about every possible product in passing. Supposing some of those would improve your life on average, is word-of-mouth really the most efficient way we can come up with to tell you about the things you do actually want to spend your money on?
Word-of-mouth vs. paid advertisements is a false dichotomy.
Also, a friction isn't a bad thing. You don't have to "research every thing that exists, or even to hear about every possible product in passing." It's fine to pick a good enough thing from a smaller set.
> In theory, this is a great business -- customers want to spend money, companies want to sell things, and the information/discoverability asymmetry means that companies are inclined to get word of their products out there with customers _also_ wanting to hear about those products (if they're sufficiently personalized). If "advertising" were good enough, I'd pay money for it.
Advertising not a great business in theory, because it's corrupted by a fundamental conflict of interest. Without draconian regulation, it's never going to be aligned to your interests as a consumer.
A better business would be some kind of product review magazine, where they research products and write articles about them.
Personally, I favor draconian regulation. Nationalize the ad agencies. Companies submit a request to the government ad agency for an add, they write a neutral ad with a couple of photos descriptive photos of the product, its name, and a brief outline of features, and that's what gets run.
> Word-of-mouth vs. paid advertisements is a false dichotomy.
I think a lot of people confuse paid advertisements by influencers as word-of-mouth. For whatever reasoning, the concept of hired spokesperson seems to have been lost with social media influencers.
I have a mouse problem. have had it for years (like most people). If you make a better mousetrap today I want it - but I'm not researching this every year.
Where the AI makes a difference here isn't regular personalized advertisement (which already isn't all that great, based on the percentage of ads I get for products I would never consider at all, or are downright offensive to me), but in understanding your existing consumers, and attempting to do habituation effects.
So imagine you have a bunch of money, watch sports while drinking and are bad at math, and therefore are considered to be a great target for sports betting companies. Making sure you get used to betting most of the time you watch a game is very valuable for the company, so just realizing what teams you like, when they play, and what kind of bets might look good to you, but are really pretty iffy is very valuable to them. Just like they would love to know when you are bored, or depressed, and maybe betting on the game that is going on right now would be appealing: A level of access to you that, say, a casino, or a bar that you haven't visited in a while just doesn't have. And habituation models are simple, you don't need a very expensive system to know when offering you a discount to entice you to don't break a gambling streak will pay off
Now that is using AI in ways that are quite antisocial by most standards: the current advertisement that tries to sell me hair growth when I have all my hair isn't all that scary.
Yep attacking gamblers is definitely one way to use AI.
Also there are plenty of other possible ways if you have the information. Think of people going thru breakups. People with eating disorders or other forms of body dismorphia, you could throw rather horrific ads at them.
if you draw a venn diagram of all the stuff i get advertised on and all the stuff I actually buy, the two circles are in completely different locations with virtually no overlap whatsoever. the only time i get ads that are even remotely related to my purchases, are only ads that come after I've made the purchase and am done. personally, i don't see how they make any profit off me whatsoever.
Maybe you're confusing who is meant to be making the profit. The people lying to you about receiving relevant, personalized ads are telling the same lie to those buying ads. The ad company tells both sides the lie and their profits are soaring.
Those buying ads have of ways to track what works. they needed that 200 years ago already and were developing it. (Not all of course, but the big ones)
i think the whole "personalised advertising" thing is way oversold and more for the benefit of a sales pitch for the advertisers but reality is far from it. google makes their money on volume, not accuracy. and so all the "information" they collect, doesn't seem to translate into more targetted advertisement.
"It annoys me that big-tech marketing has made most people believe that "personalised advertising" means they get ads which are more "useful" to them."
"relevant" is another term seen in addition to "useful"
But "relevant" is relative
For example, "relevant" to what?
It's only if Big Tech has collected data about the ad target and, e.g., made some guess about their intent, that the ads could be "relevant"
Whether the ads are truly "relevant" is a question for the reader. The term "relevant" might just be marketing fluff
Either way, Big Tech will keep the data vacuum humming
When I bought a vacuum on Amazon, I got cross-sells for other vacuums for several weeks. I don’t know how many avid vacuum collectors Amazon studied to conclude that those were the best ads to show me (not a vacuum collector).
Some of that is those advertisers want toeget everyone. You are not a women but odds are you live with one (at some point in life) who asks you to go buy something for her.
It annoys me that there aren't laws to prevent this. Or that anti-monopoly law wasn't effectively used to separate the largest advertising company in the world from a consumer software browser product which is clearly being used to facilitate and amplify these outcomes.
I'm thoroughly annoyed that adblockers aren't installed by default and require an opt out to disable. This will not at all touch first party advertising, but, it will put a huge dent into dynamic third party advertising. Which seems to be the source of the problem you describe.
Our government is genuinely failing to represent the majority on this issue.
To be fair, if I choose to buy something, it's almost by definition because I consider the thing useful. It's pretty rare that I purchase something I'd learned about from an ad, but I have done so a few times and benefited from doing so. How is anyone else to determine whether I've been adversely manipulated, i.e. whether the cost of the thing outweighs its benefit to me?
Buying something from an ad isn't really fundamentally different from being influenced by an HN post. For example, thanks to https://news.ycombinator.com/item?id=46294574, I read up on and have decided to experiment with TLA+ on my next project. It's really no concern of mine what Martin Kleppman's commercial interests may ultimately have been in publishing that blog post; I received value from the information all the same.
Personally, I'm not particularly bothered by ads per se. I'd be more bothered by information being withheld from me during searches, e.g. if Brand A could pay Amazon to delist Brands B and C from organic search results, since that would directly guide me toward less optimal purchases. But as far as simply going about my day and seeing a billboard or promoted social media post every now and then, I don't see the big deal. It's generally easy to ignore, it costs me almost nothing, it occasionally helps me, and ultimately it funds a lot of things I like and take for granted (e.g. Chromium, Firefox, and Android).
I'm not saying that people who routinely waste money on irrational purchases don't exist. I just don't find that to be a compelling argument against the existence of a particular market which overwhelmingly benefits almost all of us.
I do have quite strong concerns regarding aggressive data collection, however, and I certainly wouldn't opt in to greater erosion of my privacy — but I see that as a separate issue. To the extent that ad-driven revenue models provide an incentive for companies to facilitate greater privacy invasion, I agree that it's a significant concern which warrants much stronger pushback from the public than it receives. I just think it's important to highlight that mass data harvesting per se is the major issue, more so than any perceived manipulativeness of the fact that brands pay money for exposure.
Then of course there's the issue raised in this post, which is yet again another matter entirely. I'm all for using AI to optimize pricing and efficiency, but "dynamic pricing" as described in the article sounds like a euphemism for price discrimination, and should be more strictly regulated IMO regardless of whether or not AI is involved.
Is it possible that most people are like you and advertisers don’t make money off you but there’s a small amount of people where advertising works and advertisers make big money off those people?
I'm sure there are some "whales" who buy a lot of things they don't need, as suggested earlier, but I also believe people are entitled to the freedom to make their own decisions even if others disagree with those decisions.
That being said, I wouldn't agree that people like me don't provide value to advertisers. Aside from the fact that I do occasionally buy things from ads:
* I also get curious and research products or product categories that I learn about from ads, which translates into increased word-of-mouth, even if I never personally have a need for the thing
* If I see a lot of high-quality and/or high-profile ads for a certain brand over time, I'll be more likely to remember it and my perception of its legitimacy will improve; if I end up ever needing that kind of thing, I'll include the brand in my research of potential options
* Even if I've already purchased the thing, higher perceived legitimacy of its vendor on my part due to advertising presence could still help increase my confidence in its likely longevity, level of support, and mainstreamness, which might potentially translate into mentioning it in certain contexts where I wouldn't have otherwise
If I had to guess, that's probably where the majority of value in advertising lies. Not pure click-through rate or direct conversion, but contribution to broader mindshare. The distinction might be meaningless from a starting point of low or zero mindshare, since the marginal value of any additional mindshare at that point will be higher and the necessary timeline to produce a financial return may be tighter, but a Fortune 500 company probably doesn't buy a TV ad spot with the expectation that a handful of shopaholics will suddenly feel compelled to buy new cars or whatever on the spot.
AI is just another tool to implement a trend that has been going on for quite a while, probably starting with massive data collection through the internet. I think it’s only a matter of time until we will be seeing individualized pricing everywhere, including retail like grocery stores or gas stations
Maybe I am getting paranoid. But to me a lot of innovation in the last years feels openly hostile and primarily designed to extract maximum money while providing only little actual benefit. AI will just accelerate this trend.
>Maybe I am getting paranoid. But to me a lot of innovation in the last years feels openly hostile and primarily designed to extract maximum money while providing only little actual benefit. AI will just accelerate this trend.
This is the natural evolution of capitalism in a system where regulation is weak. Political donations from companies need to be outlawed and the proper regulatory environment restored. Capitalism is like a powerful engine that will happily rev to redline and tear itself apart if the engine lacks a governor. Regulation is that governor.
Where are you finding the API to get these prices?
Paying an individual to physically go to the store and check the prices on every good is presumably pretty expensive even if it's just quickly scanning each isle with a special phone app
The real tragedy is that we don't already have regulations already on the books in response to the endless data-hoarding that's been inflicted on us since at least the first tracking cookie.
This has been true for decades.
I recall my university classes in mid 2000s talking about examples of machine learning models for grocery store purchase patterns.
You may be thinking about this article about how Target knew that a woman was pregnant before her family knew: https://news.ycombinator.com/item?id=3598558
I wish we had an update on what the situation looks like today.
Go over to friend's place and watch the ads they get, you'll get a good idea of what kind of health concerns or illnesses they may have.
So far, in situations where it wouldn't be rude to ask, I've been able to determine with pretty good accuracy that at least someone in the household has the advertised health concerns.
You can also get an idea of their financial situation, given what buckets advertisers put them in and what they're advertised, as well.
Similarly, advertisers know when you're at friend's location, or elsewhere, and may show ads tailored to your profile.
I wouldnt jump to conclusions too fast. I'm a very thin person, but I get a lot of ads for GLP-1. My doctors have always said I need to gain weight, so I can assure you I'm not searching for weight loss solutions. Nor am I diabetic.
GLP-1 drugs are such money makers that it's worth advertising them so that you 1) know they exist 2) might recommend them 3) might take them yourself at some point.
A month's supply of Wegovy/etc costs nearly $2k.
Same thing with the monoclonal antibody drugs, you can be looking at tens of thousands of dollars per month for a single patient.
You also see this with luxury brands, that advertise to segments that cannot afford their products, in order to build image of luxury and prestige.
That said, I asked my friends/family if they had <not embarrassing or too intrusive> condition, and more often than not, someone in the household had it.
To be fair, for GLP-1 drugs I see enough of them in untargeted ads (eg. billboards) that I just assume they're carpetbombing everyone rather than doing precise targeting.
A few months ago, during a boardgaming session, I used a friend's computer to play some music. I made fun of the AI porn ads he had playing whenever I visited a YouTube page.
I only later realized that he's not using an adblocker, so this might actually be tailored to his profile.
> I only later realized that he's not using an adblocker, so this might actually be tailored to his profile.
I often forget how different my online experience is because I use Brave
You could have some fun by looking up symptoms of all kinds of diseases, so the profile of you will be filled with errors.
If you've been to a doctor's office, clinic or hospital lately, you've likely signed away the right for them to share your information with their partners during your initial intake.
My point was you cannot stop people from collecting data about you. But you can fill it with nonsense.
It's more fun to do this wild stuff when using someone else's wifi so it's also associated with their network for later savoring.
How do advertisers know when you're at your friend's location? Through your phone?
Connect to their WiFi, or your WiFi is off and still scanning for (E/B)SSIDs and thus can triangulate your location (this is what Google location services does), same thing with Bluetooth, apps/OS will grab your GPS location, and your phone carrier is selling your live and historic location data to anyone who wants it, including your government.
Two users sharing the same residential IP address
It’s a meme that will never die, but there’s no proof it ever happened:
> This story doesn’t even show that Target tried to figure out whether the girl was pregnant. It just shows that she received a flyer that contained some maternity items and her weird dad freaked out and wanted to talk to the manager. There’s no way to know whether the flyer arrived as a result of some complex targeting algorithm that correctly deduced that the girl was pregnant because she bought a bunch of lotion, or whether they just happened to be having a sale on diapers that week and sent a flyer about it to all their customers.
https://medium.com/@colin.fraser/target-didnt-figure-out-a-t...
> I wish we had an update on what the situation looks like today.
My wife and I spent 3 years in fertility treatments, which involves a lot of online activity similar to that of someone newly pregnant (buying pregnancy tests, researching symptoms, etc).
We were constantly bombarded with pregnancy related advertising, it really ramped up after the first year. Tons of "congratulations" cards, coupon books, "new mom" magazines, up to and including unsolicited shipments of formula and branded blankets.
So to answer your question, it's still happening, and it's disgusting. I strongly suspect Carrot Fertility sold our information because the peak of it all happened a couple months after I gained access to them through my employer.
(We did eventually succeed, our baby is nearing 6 months)
They have even better psychometric profiles on everyone now than they did previously. This is why Zuckerberg can confidently tell people during an interview that he knows they want at least 15 friends¹ & he is going to deliver those friends to them w/ his data centers.
¹https://www.linkedin.com/pulse/why-mark-zuckerberg-thinks-yo...
I would guess that purely observational psychometrics completely fail to predict how people will respond when challenged or stressed. I think they're trading on fools gold.
Observational psychometrics over a long enough timeframe (e.g. social media profile lifetimes) probably include periods of challenge or stress, which may help the predictive behaviour.
>You may be thinking about this article about how Target knew that a woman was pregnant before her family knew: https://news.ycombinator.com/item?id=3598558
"before her family knew" is a pretty low bar to clear, especially if the daughter was actively trying to hide the pregnancy (eg. by wearing baggy clothing). Moreover if we're taking the example of this specific story, where the women presumably knew she was pregnant (as opposed to the more sensational story of "Target figured out a women was pregnant before she even knew!!1!" that also makes the rounds), it's not hard to imagine how Target might be in a better position to infer her pregnancy without being galaxy brained or creepy. Take the examples given in the article:
>Take a fictional Target shopper named Jenny Ward, who is 23, lives in Atlanta and in March bought cocoa-butter lotion, a purse large enough to double as a diaper bag, zinc and magnesium supplements and a bright blue rug.
We don't know, but since the girl in question was only 14 it is believable she didn't know yet, but in that case she would be to a doctor soon after.
>The manager apologized and then called a few days later to apologize again.
>On the phone, though, the father was somewhat abashed. “I had a talk with my daughter,” he said. “It turns out there’s been some activities in my house I haven’t been completely aware of. She’s due in August. I owe you an apology.”
That quote, especially the "there’s been some activities in my house I haven’t been completely aware of" part makes me think she knew and was trying to hide it. There's also the expectation that if she really didn't know, that person writing the article would put that detail in, given how extra sensational it would make the story.
That is possible - even likely. However we still don't know which was my point.
I think that is still creepy, but that example just seems horrible?
Oh no, a woman bought lotion, a purse, and a rug. Must be pregnant!
Galaxy-brained, no; creepy, yes.
It's creepy for companies to keep track of what you buy? How do you think your Amazon order history works?
To track it? Maybe, maybe not, depends on what the conpany is, how the purchases are made, etc.
To analyze it to infer personal information? Starting to be creepy, even in the cases where tracking it isn’t.
And then use the inferred information for marketing explicitly and overtly around the inference? Definitely getting creepier.
I continuously get Spanish advertisements despite being unable to speak Spanish. (The only foreign language media I visit on YouTube is Anime/Japanese and Germans stuff).
There are surely large numbers of dumb cases of being mispredicted out there in.
The miss rate is still wildly high based on the ads in my instagram feeds.
But the really-good hits are probably tough to notice. They won't stand out as "boy this is a stupid ad" and even if I just scroll past a well-targeted ad, it's probably doing its job of making that company/product a bit closer to the top of mind...
Some of the miss is intentional. they don't want you to know what they know so they give some random things you don't want to throw you off
Yeah AI is taking a lot of damage when the actual problem is capitalism
> Yeah AI is taking a lot of damage when the actual problem is capitalism
And I think it's fair to to throw flak in AI's direction, if what it does is make capitalism less tolerable by removing some of the "inefficiencies."
While apologists for capitalism have done a good job of pushing me towards wanting to burn it all down, I doubt that's in the cards any time soon and limits on AI technology are far more likely.
Seems ironic. Perhaps we should celebrate AI as the accelerator of capitalism’s ultimate unworkable demise
We aren’t getting regulations on AI. The military industrial complex includes the tech industry now. It’s an existential race to beat China.
The sad reality is that for all of our potential futures, we aren’t getting the Star Trek post scarcity utopia. Our onboard ship computers aren’t generating Earl Grey, hot, they are generating trillionaires on one hand and poverty on the other.
>The military industrial complex includes the tech industry now. It’s an existential race to beat China.
Too bad nobody really even knows what this means but a lot of people will use it as a slogan to convince people to give this money!
It annoys me that big-tech marketing has made most people believe that "personalised advertising" means they get ads which are more "useful" to them. I regularly see people opt in to personalised advertising because of this.
Personalised advertising is about collecting every detail about your life and using it to extract as much money as possible from you. AI advancements might be making this even more effective but it's been this way for a long time.
This is one of those places where it's worth disentangling the status quo from an optimal alternative.
Currently, every factlet you leak to one of these systems poisons them toward their profit (and almost unanimously against your best interests). Advertising draws your attention away from the products that would make your life better (cheaper, heathier, tastier, whatever) and toward profitable alternatives.
It doesn't have to be that way though. You physically don't have time to research every thing that exists, or even to hear about every possible product in passing. Supposing some of those would improve your life on average, is word-of-mouth really the most efficient way we can come up with to tell you about the things you do actually want to spend your money on? In theory, this is a great business -- customers want to spend money, companies want to sell things, and the information/discoverability asymmetry means that companies are inclined to get word of their products out there with customers _also_ wanting to hear about those products (if they're sufficiently personalized). If "advertising" were good enough, I'd pay money for it.
That only falls apart because of a lack of trust and ethical behavior. Instead of being treated like the information market it is, it's thrust onto individuals to try to prey on their weaknesses.
> It doesn't have to be that way though. You physically don't have time to research every thing that exists, or even to hear about every possible product in passing. Supposing some of those would improve your life on average, is word-of-mouth really the most efficient way we can come up with to tell you about the things you do actually want to spend your money on?
Word-of-mouth vs. paid advertisements is a false dichotomy.
Also, a friction isn't a bad thing. You don't have to "research every thing that exists, or even to hear about every possible product in passing." It's fine to pick a good enough thing from a smaller set.
> In theory, this is a great business -- customers want to spend money, companies want to sell things, and the information/discoverability asymmetry means that companies are inclined to get word of their products out there with customers _also_ wanting to hear about those products (if they're sufficiently personalized). If "advertising" were good enough, I'd pay money for it.
Advertising not a great business in theory, because it's corrupted by a fundamental conflict of interest. Without draconian regulation, it's never going to be aligned to your interests as a consumer.
A better business would be some kind of product review magazine, where they research products and write articles about them.
Personally, I favor draconian regulation. Nationalize the ad agencies. Companies submit a request to the government ad agency for an add, they write a neutral ad with a couple of photos descriptive photos of the product, its name, and a brief outline of features, and that's what gets run.
> A better business would be some kind of product review magazine, where they research products and write articles about them.
Australia has https://www.choice.com.au/ - a subscription non-profit product review website & magazine.
> Word-of-mouth vs. paid advertisements is a false dichotomy.
I think a lot of people confuse paid advertisements by influencers as word-of-mouth. For whatever reasoning, the concept of hired spokesperson seems to have been lost with social media influencers.
I have a mouse problem. have had it for years (like most people). If you make a better mousetrap today I want it - but I'm not researching this every year.
Opt-in product catalogues are fine for that. Plenty.
Where the AI makes a difference here isn't regular personalized advertisement (which already isn't all that great, based on the percentage of ads I get for products I would never consider at all, or are downright offensive to me), but in understanding your existing consumers, and attempting to do habituation effects.
So imagine you have a bunch of money, watch sports while drinking and are bad at math, and therefore are considered to be a great target for sports betting companies. Making sure you get used to betting most of the time you watch a game is very valuable for the company, so just realizing what teams you like, when they play, and what kind of bets might look good to you, but are really pretty iffy is very valuable to them. Just like they would love to know when you are bored, or depressed, and maybe betting on the game that is going on right now would be appealing: A level of access to you that, say, a casino, or a bar that you haven't visited in a while just doesn't have. And habituation models are simple, you don't need a very expensive system to know when offering you a discount to entice you to don't break a gambling streak will pay off
Now that is using AI in ways that are quite antisocial by most standards: the current advertisement that tries to sell me hair growth when I have all my hair isn't all that scary.
Yep attacking gamblers is definitely one way to use AI.
Also there are plenty of other possible ways if you have the information. Think of people going thru breakups. People with eating disorders or other forms of body dismorphia, you could throw rather horrific ads at them.
if you draw a venn diagram of all the stuff i get advertised on and all the stuff I actually buy, the two circles are in completely different locations with virtually no overlap whatsoever. the only time i get ads that are even remotely related to my purchases, are only ads that come after I've made the purchase and am done. personally, i don't see how they make any profit off me whatsoever.
As long as there's some company willing to bid the minimum amount, they'll happily serve up those ads.
I have YouTube ad personalization off, and sometimes get frequently repeating ads. I suspect when that happens, they are the only bidders.
Maybe you're confusing who is meant to be making the profit. The people lying to you about receiving relevant, personalized ads are telling the same lie to those buying ads. The ad company tells both sides the lie and their profits are soaring.
Those buying ads have of ways to track what works. they needed that 200 years ago already and were developing it. (Not all of course, but the big ones)
i think the whole "personalised advertising" thing is way oversold and more for the benefit of a sales pitch for the advertisers but reality is far from it. google makes their money on volume, not accuracy. and so all the "information" they collect, doesn't seem to translate into more targetted advertisement.
"It annoys me that big-tech marketing has made most people believe that "personalised advertising" means they get ads which are more "useful" to them."
"relevant" is another term seen in addition to "useful"
But "relevant" is relative
For example, "relevant" to what?
It's only if Big Tech has collected data about the ad target and, e.g., made some guess about their intent, that the ads could be "relevant"
Whether the ads are truly "relevant" is a question for the reader. The term "relevant" might just be marketing fluff
Either way, Big Tech will keep the data vacuum humming
Well I can list some things which are completely irrelevant (happens even in online ads despite the advancements).
I got an offer for life insurance for US veterans - I’m not a US veteran so this has nothing to do with me.
I got an ad for women’s hygiene products, but I’m not a woman. So that’s completely wasted on me.
I just bought a mattress, and I don’t need a 2nd mattress, so all of those are irrelevant.
When I bought a vacuum on Amazon, I got cross-sells for other vacuums for several weeks. I don’t know how many avid vacuum collectors Amazon studied to conclude that those were the best ads to show me (not a vacuum collector).
Some of that is those advertisers want toeget everyone. You are not a women but odds are you live with one (at some point in life) who asks you to go buy something for her.
It annoys me that there aren't laws to prevent this. Or that anti-monopoly law wasn't effectively used to separate the largest advertising company in the world from a consumer software browser product which is clearly being used to facilitate and amplify these outcomes.
I'm thoroughly annoyed that adblockers aren't installed by default and require an opt out to disable. This will not at all touch first party advertising, but, it will put a huge dent into dynamic third party advertising. Which seems to be the source of the problem you describe.
Our government is genuinely failing to represent the majority on this issue.
To be fair, if I choose to buy something, it's almost by definition because I consider the thing useful. It's pretty rare that I purchase something I'd learned about from an ad, but I have done so a few times and benefited from doing so. How is anyone else to determine whether I've been adversely manipulated, i.e. whether the cost of the thing outweighs its benefit to me?
Buying something from an ad isn't really fundamentally different from being influenced by an HN post. For example, thanks to https://news.ycombinator.com/item?id=46294574, I read up on and have decided to experiment with TLA+ on my next project. It's really no concern of mine what Martin Kleppman's commercial interests may ultimately have been in publishing that blog post; I received value from the information all the same.
Personally, I'm not particularly bothered by ads per se. I'd be more bothered by information being withheld from me during searches, e.g. if Brand A could pay Amazon to delist Brands B and C from organic search results, since that would directly guide me toward less optimal purchases. But as far as simply going about my day and seeing a billboard or promoted social media post every now and then, I don't see the big deal. It's generally easy to ignore, it costs me almost nothing, it occasionally helps me, and ultimately it funds a lot of things I like and take for granted (e.g. Chromium, Firefox, and Android).
I'm not saying that people who routinely waste money on irrational purchases don't exist. I just don't find that to be a compelling argument against the existence of a particular market which overwhelmingly benefits almost all of us.
I do have quite strong concerns regarding aggressive data collection, however, and I certainly wouldn't opt in to greater erosion of my privacy — but I see that as a separate issue. To the extent that ad-driven revenue models provide an incentive for companies to facilitate greater privacy invasion, I agree that it's a significant concern which warrants much stronger pushback from the public than it receives. I just think it's important to highlight that mass data harvesting per se is the major issue, more so than any perceived manipulativeness of the fact that brands pay money for exposure.
Then of course there's the issue raised in this post, which is yet again another matter entirely. I'm all for using AI to optimize pricing and efficiency, but "dynamic pricing" as described in the article sounds like a euphemism for price discrimination, and should be more strictly regulated IMO regardless of whether or not AI is involved.
Is it possible that most people are like you and advertisers don’t make money off you but there’s a small amount of people where advertising works and advertisers make big money off those people?
I'm sure there are some "whales" who buy a lot of things they don't need, as suggested earlier, but I also believe people are entitled to the freedom to make their own decisions even if others disagree with those decisions.
That being said, I wouldn't agree that people like me don't provide value to advertisers. Aside from the fact that I do occasionally buy things from ads:
* I also get curious and research products or product categories that I learn about from ads, which translates into increased word-of-mouth, even if I never personally have a need for the thing
* If I see a lot of high-quality and/or high-profile ads for a certain brand over time, I'll be more likely to remember it and my perception of its legitimacy will improve; if I end up ever needing that kind of thing, I'll include the brand in my research of potential options
* Even if I've already purchased the thing, higher perceived legitimacy of its vendor on my part due to advertising presence could still help increase my confidence in its likely longevity, level of support, and mainstreamness, which might potentially translate into mentioning it in certain contexts where I wouldn't have otherwise
If I had to guess, that's probably where the majority of value in advertising lies. Not pure click-through rate or direct conversion, but contribution to broader mindshare. The distinction might be meaningless from a starting point of low or zero mindshare, since the marginal value of any additional mindshare at that point will be higher and the necessary timeline to produce a financial return may be tighter, but a Fortune 500 company probably doesn't buy a TV ad spot with the expectation that a handful of shopaholics will suddenly feel compelled to buy new cars or whatever on the spot.
AI is just another tool to implement a trend that has been going on for quite a while, probably starting with massive data collection through the internet. I think it’s only a matter of time until we will be seeing individualized pricing everywhere, including retail like grocery stores or gas stations
Maybe I am getting paranoid. But to me a lot of innovation in the last years feels openly hostile and primarily designed to extract maximum money while providing only little actual benefit. AI will just accelerate this trend.
More perfect and consumer reports did a joint investigation and have uncovered this already happening.
https://youtube.com/watch?v=osxr7xSxsGo&pp=0gcJCR4Bo7VqN5tD
>Maybe I am getting paranoid. But to me a lot of innovation in the last years feels openly hostile and primarily designed to extract maximum money while providing only little actual benefit. AI will just accelerate this trend.
This is the natural evolution of capitalism in a system where regulation is weak. Political donations from companies need to be outlawed and the proper regulatory environment restored. Capitalism is like a powerful engine that will happily rev to redline and tear itself apart if the engine lacks a governor. Regulation is that governor.
why doesnt someone make a coinmarketcap for grocery prices across all the stores and update it in real time?
Where are you finding the API to get these prices?
Paying an individual to physically go to the store and check the prices on every good is presumably pretty expensive even if it's just quickly scanning each isle with a special phone app
what if we could incentivize the grocery store employees themselves, i am sure nobody minds earning an extra hour of work
ahh yes. it's not just this, it's that.
The real tragedy is that we don't already have regulations already on the books in response to the endless data-hoarding that's been inflicted on us since at least the first tracking cookie.