3 ms·

I wonder if rather than focusing on completely opting out, if the author had tried a deliberate confusion strategy in addition to minimizing her digital footpri

I wonder if rather than focusing on completely opting out, if the author had tried a deliberate confusion strategy in addition to minimizing her digital footprint would she have been "more" successful? (I recognize the difficulty in measuring "successful" outcomes with this metric). One of the "dirty" secrets of Big data is that for the Analytics to work is that the data has to be "prepped". How about deliberately using different names, genders, etc etc in social media, stores (first abbreviation, maiden name etc). Once the data trail gets muddy enough, you cannot track anything. Hmmm Might even be an "app" idea in this :)

bediger4000 12y ago
Would it be possible to maximize your footprint, instead of minimizing it? I think it's almost impossible to opt out, as the story says, so I try to get other names associated with my address, get other people to use my grocery store loyalty card, and I let my kids do school homework searches under my Google log-in. I also fill in as many surveys as I can with semi-fake answers, and on Netflix or Hulu (Can't recall which one does this) I alternate checking "yes" and "no" for "Is this ad relevant to you?" If you create enough "data smog" will the Big Data people be able to penetrate it?
- nostrademons 12y ago
  If you create enough data smog the Big Data systems will just give wrong answers. Whether that's a good or bad thing depends on how the data's used. It seems to be a common misconception that the Googles and Facebooks of the world have someone nefariously looking at your data to profile your every move. No. The average employee at Facebook does not care about you, and even if they have access to your personal data (I know that this is significantly locked down at Google, and probably is at Facebook these days), they don't care who you are or what you do. Instead, they're developing algorithms that mine lots of data from lots of consumers to produce products for lots of consumers. If you put in wrong information, it's basically GIGO. And since the basic models are trained on millions of data points, if just you do it, you only hurt yourself - you're a drop in the bucket compared to everyone's data, but your own personalized interface will be based on the profile you give it, not what you actually want. You can see this in action with shared Netflix accounts. Share a Netflix account with someone and basically you get garbage for recommendations, because the system tries to make sense of two different "streams" of data that were arbitrarily merged together. (There are ways to separate them out, but they add a lot of complexity to the system.) Things may be different if you're talking about Palantir or the NSA. They specifically do look at individuals. But then, if they get a bunch of random data their assumption is probably going to be that the person is deliberately trying to obscure their tracks ("tradecraft"), which means that the suspect warrants even further scrutiny.
  - dman 12y ago
    What happens during wartime when civil liberties are curtailed? I suspect each company will have to follow their law of the land which might involve regrettable use of data. In other words I dont have a lot of trouble believing that the Googles and the Facebooks of the world will be good stewards of data in normal circumstances (Its in their self interest after all). The other issue worth thinking about is that we are in the midst of the golden age of building networks so the economic models are all viable. What happens 10-20 years from now when the roving eye of economic development has moved onto other sectors leaving the economics of ad supported saas products non viable. I fully trust Google and Facebook, but if they were to ever be liquidated / sold in pieces I dont know if my data will end up in safe hands.
    - nostrademons 12y ago
      This is a legit concern - there's a long history of organizations doing unscrupulous things with their data & IP when their actual business goes south. (SCO is perhaps the best example.) However, these are usually more self-limiting nuisances than long-term problems. If the business is going south, then it means the organization's actual power is declining, and this limits the damage they can do with their data. They're quite capable of making a lot of noise and annoying a lot of people, but they'd then lack the power to deal with the inevitable blowback. What would happen if Google goes under in 30 years and sells all the data it's accumulated on all Americans? Let's say they sell it to a worst-case buyer, someone who explicitly uses it for blackmail & extortion. If they do this on a small scale, then by definition it's on a small scale, they can't make any money off it, and they get shut down legally when they pick on the wrong journalist or wealthy individual. (They're in a catch-22 when picking targets: if they pick poor, powerless people, there's no money they can extract from them, while if they pick fat wealthy targets they'll probably get fought hard legally.) If they do it on a large-scale, the public outcry will get laws passed faster than you can say "re-election", and they'll get shut down by Congress/lawsuits. Either one is a no-win for the buyer. A totalitarian government is more worrying, but the government already has all your personal data anyway.
- saraid216 12y ago
  If they're looking for you, specifically, then yes. It is possible to have the smoke trail have an erroneous local maximum, but it's unlikely that you can pull this off. If they're just looking for patterns of which you are just a statistic, then it'll merely be hit-or-miss; you're unlikely to generate enough "smog" to make it so they never actually tag you correctly.
gress 12y ago
Even if that strategy were more effective, people who employ it still face a lifetime of additional stress to maintain it.
inerte 12y ago
But that means someone somewhere will know she (or a variation of her name) is pregnant. I have arrived on the US only 8 months ago and since 3 months ago during every week I am getting two mailings from Financial One, a company I've never directly (or knowingly indirectly) made business with. One is addressed to "First M Last" and the other to "First Middle Last". Financial One doesn't care, it simply send the ad to the two of me.
pessimizer 12y ago
My mother has done this her entire life, and passed the habit down to me. She alternates between her married name (from a 35 year old divorce) and her maiden name and both, hyphenated and unhyphenated, intentionally typos when she can get away with it, uses a lot of off-by-1-or-2 for numbers in registration profiles, old addresses, and alternately combines and separates her first and middle names (because of her particular name, it works.) She's also a data person professionally, and once worked at Acxiom. At the time, when she checked into what they had on her, she found that Acxiom thought she was no fewer than seven people, some of them male:)
- x0x0 12y ago
  I find it weird that a person so concerned w/ her privacy would work for Acxiom; did she ever mention how she squared that?
  - pessimizer 12y ago
    There's no contradiction between those two things that I'm aware of. All jobs and all purchases under capitalism play their part in the general oppression of the working class:) There is the day to day disgust with how directly what you're doing enforces the status quo - but alternatively, there aren't a lot of good (or morally good) computer jobs in Arkansas; ask a Wal-Mart employee.
- vitd 12y ago
  And how does that affect her ability to do things like get a password reset? My wife used to do a little of that, though not to the extent of your mother. (Mostly just putting in that she was 120 years old on forms with ages, and stuff.) At one point, she lost the password to an account, and to reset it, she had to enter the fake data she had supplied to get it back. She was never able to remember which years she had chosen for which account. Luckily, it wasn't for anything too important. Also, are there any legal implications to doing something like that? It's probably fine for signing up to a web site, but does she do that on things like auto loans or mortgages? Has it ever been a problem? I like the idea, it just seems like it's inviting trouble.
DEinspanjer 12y ago
Read a fun book that had that as a central plot device a while back: https://en.wikipedia.org/wiki/Feed_%28Anderson_novel%29 https://en.wikipedia.org/wiki/Feed_%28Anderson_novel%29 Unfortunately, it didn't turn out so well for the person who tried to follow that plan.
ihsw 12y ago
Indeed, this work for humans too. You cannot remove information once it enters people's heads, but you sure as hell can spread misinformation to drown it out. The problem is that there's is a very large body of information associated with you that is not under your direct control. In some cases spreading such misinformation is quite illegal (caller ID, email forwarding services), unfeasible for most people (GPS location, HTTP referrer), or technologically impossible (email sender information). Fortunately this is becoming more accepted with the advent of Bitcoin and Tor, however even those services are generally regarded by society as criminal at worst and sketchy at best.

bediger4000 12y ago

Would it be possible to maximize your footprint, instead of minimizing it? I think it's almost impossible to opt out, as the story says, so I try to get other names associated with my address, get other people to use my grocery store loyalty card, and I let my kids do school homework searches under my Google log-in. I also fill in as many surveys as I can with semi-fake answers, and on Netflix or Hulu (Can't recall which one does this) I alternate checking "yes" and "no" for "Is this ad relevant to you?" If you create enough "data smog" will the Big Data people be able to penetrate it?

nostrademons 12y ago
If you create enough data smog the Big Data systems will just give wrong answers. Whether that's a good or bad thing depends on how the data's used. It seems to be a common misconception that the Googles and Facebooks of the world have someone nefariously looking at your data to profile your every move. No. The average employee at Facebook does not care about you, and even if they have access to your personal data (I know that this is significantly locked down at Google, and probably is at Facebook these days), they don't care who you are or what you do. Instead, they're developing algorithms that mine lots of data from lots of consumers to produce products for lots of consumers. If you put in wrong information, it's basically GIGO. And since the basic models are trained on millions of data points, if just you do it, you only hurt yourself - you're a drop in the bucket compared to everyone's data, but your own personalized interface will be based on the profile you give it, not what you actually want. You can see this in action with shared Netflix accounts. Share a Netflix account with someone and basically you get garbage for recommendations, because the system tries to make sense of two different "streams" of data that were arbitrarily merged together. (There are ways to separate them out, but they add a lot of complexity to the system.) Things may be different if you're talking about Palantir or the NSA. They specifically do look at individuals. But then, if they get a bunch of random data their assumption is probably going to be that the person is deliberately trying to obscure their tracks ("tradecraft"), which means that the suspect warrants even further scrutiny.
- dman 12y ago
  What happens during wartime when civil liberties are curtailed? I suspect each company will have to follow their law of the land which might involve regrettable use of data. In other words I dont have a lot of trouble believing that the Googles and the Facebooks of the world will be good stewards of data in normal circumstances (Its in their self interest after all). The other issue worth thinking about is that we are in the midst of the golden age of building networks so the economic models are all viable. What happens 10-20 years from now when the roving eye of economic development has moved onto other sectors leaving the economics of ad supported saas products non viable. I fully trust Google and Facebook, but if they were to ever be liquidated / sold in pieces I dont know if my data will end up in safe hands.
  - nostrademons 12y ago
    This is a legit concern - there's a long history of organizations doing unscrupulous things with their data & IP when their actual business goes south. (SCO is perhaps the best example.) However, these are usually more self-limiting nuisances than long-term problems. If the business is going south, then it means the organization's actual power is declining, and this limits the damage they can do with their data. They're quite capable of making a lot of noise and annoying a lot of people, but they'd then lack the power to deal with the inevitable blowback. What would happen if Google goes under in 30 years and sells all the data it's accumulated on all Americans? Let's say they sell it to a worst-case buyer, someone who explicitly uses it for blackmail & extortion. If they do this on a small scale, then by definition it's on a small scale, they can't make any money off it, and they get shut down legally when they pick on the wrong journalist or wealthy individual. (They're in a catch-22 when picking targets: if they pick poor, powerless people, there's no money they can extract from them, while if they pick fat wealthy targets they'll probably get fought hard legally.) If they do it on a large-scale, the public outcry will get laws passed faster than you can say "re-election", and they'll get shut down by Congress/lawsuits. Either one is a no-win for the buyer. A totalitarian government is more worrying, but the government already has all your personal data anyway.
saraid216 12y ago
If they're looking for you, specifically, then yes. It is possible to have the smoke trail have an erroneous local maximum, but it's unlikely that you can pull this off. If they're just looking for patterns of which you are just a statistic, then it'll merely be hit-or-miss; you're unlikely to generate enough "smog" to make it so they never actually tag you correctly.

gress 12y ago

Even if that strategy were more effective, people who employ it still face a lifetime of additional stress to maintain it.

inerte 12y ago

But that means someone somewhere will know she (or a variation of her name) is pregnant. I have arrived on the US only 8 months ago and since 3 months ago during every week I am getting two mailings from Financial One, a company I've never directly (or knowingly indirectly) made business with. One is addressed to "First M Last" and the other to "First Middle Last". Financial One doesn't care, it simply send the ad to the two of me.

pessimizer 12y ago

My mother has done this her entire life, and passed the habit down to me. She alternates between her married name (from a 35 year old divorce) and her maiden name and both, hyphenated and unhyphenated, intentionally typos when she can get away with it, uses a lot of off-by-1-or-2 for numbers in registration profiles, old addresses, and alternately combines and separates her first and middle names (because of her particular name, it works.) She's also a data person professionally, and once worked at Acxiom. At the time, when she checked into what they had on her, she found that Acxiom thought she was no fewer than seven people, some of them male:)

x0x0 12y ago
I find it weird that a person so concerned w/ her privacy would work for Acxiom; did she ever mention how she squared that?
- pessimizer 12y ago
  There's no contradiction between those two things that I'm aware of. All jobs and all purchases under capitalism play their part in the general oppression of the working class:) There is the day to day disgust with how directly what you're doing enforces the status quo - but alternatively, there aren't a lot of good (or morally good) computer jobs in Arkansas; ask a Wal-Mart employee.
vitd 12y ago
And how does that affect her ability to do things like get a password reset? My wife used to do a little of that, though not to the extent of your mother. (Mostly just putting in that she was 120 years old on forms with ages, and stuff.) At one point, she lost the password to an account, and to reset it, she had to enter the fake data she had supplied to get it back. She was never able to remember which years she had chosen for which account. Luckily, it wasn't for anything too important. Also, are there any legal implications to doing something like that? It's probably fine for signing up to a web site, but does she do that on things like auto loans or mortgages? Has it ever been a problem? I like the idea, it just seems like it's inviting trouble.

DEinspanjer 12y ago

Read a fun book that had that as a central plot device a while back: https://en.wikipedia.org/wiki/Feed_%28Anderson_novel%29 https://en.wikipedia.org/wiki/Feed_%28Anderson_novel%29 Unfortunately, it didn't turn out so well for the person who tried to follow that plan.

ihsw 12y ago

Indeed, this work for humans too. You cannot remove information once it enters people's heads, but you sure as hell can spread misinformation to drown it out. The problem is that there's is a very large body of information associated with you that is not under your direct control. In some cases spreading such misinformation is quite illegal (caller ID, email forwarding services), unfeasible for most people (GPS location, HTTP referrer), or technologically impossible (email sender information). Fortunately this is becoming more accepted with the advent of Bitcoin and Tor, however even those services are generally regarded by society as criminal at worst and sketchy at best.