When does privacy matter?

[dreamshark]

As we slither down the slippery slope to a post-privacy world, it gets trickier and trickier to figure out when to be outraged, when to be slightly unsettled, and when to just shrug and say "and so it goes."  Almost everything that Facebook does outrages me, because their explicitly stated business plan outrages me (and they are pretty consistent about implementing features that move them towards that goal). I am more kindly disposed towards Google and Amazon because I have found both of them so useful, but they're making me increasingly nervous too. Here are some questions I am currently pondering.

• Would I care if comments I made on Facebook about a product were used in ads for that product targeted at my "friends?"   [Yes]


• Would it be even worse if Facebook used my picture in these ads? [Yes. Fortunately I'm not on Facebook]


• Does it make a difference that for the purposes of this discussion, "product" can include political candidates? [Not sure. In some ways I'd rather have my paltry personal influence used to support "my" candidates than to sell consumer products.]


• Do I care if my online shopping patterns are tracked and used to present me with targeted adverts? [No, at least so far. Right now I'm just finding it funny, as I am now seeing advertisements EVERYWHERE for a product I have already purchased]


• Do I want my online search results "informed" by the opinions of "friends" on social media? [Hell, no! Bing is horrifying.]


• Are facial recognition algorithms creepy?  [Yes, especially combined with all those Facebook initiatives to "monetize" their user base.  On the other hand, it's really cool when I'm using it to sort my digital pictures in the privacy of my own computer. Split decision.]


• Is Twitter inherently evil? If not, is it good for anything that I care about?   [decision still out]

Comments

[dd-b.livejournal.com]

Yeah, it's a big topic, and a terribly important one -- and I don't think we have anything close to social consensus on it, so discussion is important. (We're clearly facing changes; at the very very least, we must codify things that are currently only traditional, but in fact I think actual changes are inevitable and probably desirable.)

• I don't think I'd mind if comments I made on a product were displayed to others, so long as those others already had visibility to the comment.
  
• I think it's okay if they use my standard icon, or the one I used for the comments they're quoting, in the place they quote them.
  
• I don't immediately think political candidates are that different.
  
• I don't like the tracking of online shopping patterns, because I do think that should be private. Also because they do such a piss-poor job of targeting ads based on it; if it made the ads more interesting, or I thought it actually helped them some, I might consider reconsidering.
  
• Not sure about social input to searches. In theory, my friends often know about things I care about. I don't know that it's likely they'd actually do it so that it was a benefit though (certainly your comment aboug Bing suggests it doesn't work well there).
  
• I wish I had useful facial recognition algorithms. Facebook can't even accurately identify what's a face, let alone which two are similar. There are potentially VERY creepy possibilities with widespread public surveillance and facial recognition, and also for inferring friend circles beyond what one admits to.
  
• Twitter must be very cool, because Neil Gaiman and Kyle Cassidy and Patrick Nielsen Hayden and Steven Brust all rather like it. But I've never gotten much of anywhere with it myself.
  

Looking at this, I think my pattern is that I don't mind so much their doing things with what I publish (I post comments about a product publicly, I don't mind their using that too much), but I don't want them using stuff I don't publish (online buying patterns; that's between me and the store, I don't want it spread).

I think I'm MUCH more private online than a lot of people (especially some modern teens); I tend not to post even in friends-locked venues anything I'd really hate becoming public. (I do tend to treat email as private, though.) And I work in a field full of rather open-minded people, and not under strong political pressure from the outside. If I were a day-care worker or a school teacher for example, there are things I let at least hints of into my online posts that I'd want to completely black out.

As I understand the concept of privacy, one part of it is that people get to choose a fair amount about how information of theirs is used. I'm making middling-liberal choices, but I don't think that should be forced onto everybody else.

[dreamshark.livejournal.com]

So far, all of these technologies are in their infancy. So how we feel about them should include the assumption that they are going to get much more effective (which may be better or worse depending on your point of view). So far, the targeted ad thing seems to be working about as well as Netflix's generally inept attempts to predict what movies I'll like, so it's more a source of amusement than potentially helpful. However, so far it hasn't bothered me so I have made no attempt to shut it off. Someday it might actually improve my online experience.

My comments about Bing are based on the concept, not on how well it actually works (I don't KNOW how well it works, since I don't use it). When I search for information online I want to get predictable, fact-based results. I do not want the information I'm presented with skewed to match the expectations of my presumed online cohort. Right now it's under development, and who wants to use beta-test search algorithms? But when it DOES work, it may be even worse, leading to self-fulfilling prophecies. For instance, if a lot of your friends think that vaccinations cause autism, your online searches to see if they might be right would be more likely to turn up the same nutball sites said friends have been looking at.

Actually, as the Facebook advertising algorithms get more sophisticated they will probably have similar effects, creating little islands of friendship groups that are being presented with similar sets of dubious informational links.

Since I don't use Facebook, I don't know how well their facial recognition algorithm works. Have you tried Google's tool? I downloaded a pc-based version of Picasa, which turns out to be a very nice little photo management program. Just for fun, I turned on facial recognition to see what it did. It ran through every picture in my images folder, creating head shots of every face it found. This was handy in itself, as there are always lots of nice head shots lurking unnoticed inside of larger pictures. Of the thousands of images it identified as faces, I think there were TWO that actually were not.

I started tagging faces, and it very quickly started learning. Some faces it identified confidently (and always correctly). More often it would present me with 2 or 3 possibilities to choose from, most of the time including the correct choice. As you make corrections you can see it learning, as the tentative identifications under each individual change immediately to reflect the new information. It's kind of fascinating, actually. Or would be if it weren't so creepy in its implications.

I'm also pretty private online, certainly compared to "kids today." But it turns out that other people's choices have effects on me. Even if I never posted an online picture of myself, my picture is still all over the place: on other people's Facebook pages, public webcams, security pictures and probably backgrounds of strangers' pictures. Once facial recognition gets really good, all those random shots can be tagged automatically, and eventually bots could be turned loose into the Internet to sum up all these random sitings and track the movements of pretty much anybody.

And if I use social media, more and more of the things that I see on that social medium will be controlled by the choices that have been made by my online "friends." Even if I never click on a Like button, my friends' Likes will be used to determine the ads I see, and God knows what else. For instance, my account might be given a different set of default features and settings based on how my friends' accounts are set up.

[barondave.livejournal.com]

My introvert and extrovert sides clash at the edges. On one hand, I'm a fairly public persona and most of my major views were broadcast decades ago. On the other hand, there's stuff I don't necessarily want aired, not so much because it's private (though some is) but because dealing with people knowing about it is time consuming.

My general response to a post-privacy world: Add noise to the system. I'll make a post extolling the virtue of some politician I don't like, or claim I'm from a different city (which is why FB is confused) or talk about some product or other. Not so much noise that anyone who actually knows me will be fooled, but enough to make the patterns less effective and therefore less economical to generate.

Having FB use your comments to target people for a certain product is no different, in my view, from people paying money to buy a t-shirt with a product name on it. The scale is not quite the same, but you're still paying money for other people to advertise on you.

I'm a little creeped out about facial recognition, but not, as yet, a lot.

[dreamshark.livejournal.com]

"My general response to a post-privacy world: Add noise to the system. I'll make a post extolling the virtue of some politician I don't like..."

We'll have to see how hilarious you think that is when your picture starts popping up on your friends' screens endorsing the next Republican presidential candidate.

[barondave.livejournal.com]

Stuff like that happens all the time in Real Life (tm), such as Paul Ryan's washing of clean dishes in a place that didn't want him to be there in the first place. It backfires on the politician, and is pretty funny. And the place Ryan made famous got lots of donations... from the other side.