Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

> I don't know how they do it. But they do it. I'm assuming geo-tracking of phones and noticing that you spent X hours together the previous night.

My boss and I were wondering the same thing after spending few days at client's site in another country, after which Facebook suggested the client as a friend to my boss, even though they never interacted electronically by means other than company e-mail. We thought about geo-tracking, but then concluded that it's more likely Facebook must have identified both of them were connecting to the same Wi-Fi network at the same time.



Somehow the Wi-Fi network theory sounds even creepier than geo-tracking.

Also in my example, I didn't usually connect to any Wi-Fi because I have LTE. But of course I don't turn off wi-fi when I leave the house so my phone still does network discovery.


Well, when you think about it they probably associate users who use the same IP address. From that perspective it's difficult to see how they could avoid using this data. It's like not noticing that two different people called you from the same number.


Does it now? By Wi-Fi tracking I'm thinking of SSID matching (which are public information), not pinging devices inside a single network.

Your data point about on-but-not-connected Wi-Fi supports both geotracking and SSID tracking. We need more data!


Facebook harvests a mountain of data about you when you use the new Messenger app. I swear the only purpose it has as a separate app is to be far worse than they think they can get away with the main Facebook app being.

It's probably from the messenger data that they correlated your proximity.


You know when websites ask for consent to share data with third parties? This is mostly how they do it. What you view on linkedin goes to fb and vice versa.

I had used a chat service only once to call some guys and the day after all participants were suggestions on linkedin.

To most of those free service a product is you and they monetize your graph in this way.

Then you have subsidiaries. You may gave not given facebook or messanger access to you contacts, but what anout whatsapp? Any app with access to your contacts may have sold them. And even if you have none the other party might, or even a third party.


> What you view on linkedin goes to fb and vice versa.

That's completely unfounded speculation, and I'd love to see any citation for it.

I'm inherently skeptical because for these companies their most important asset is their data. The idea that they would share that data willy-nilly with third parties makes no business sense.


it is indeed speculation, but at least anecdotal, not totally unfunded

http://www.interactually.com/linkedin-creepiest-social-netwo...

http://imonlinkedinnowwhat.com/2009/09/14/linkedin-people-yo...

there are quite a few instances of weird guess about the people you might know suggestions

of course the actual source is kept as one of their best secrets.


Neither of those links you shared provided any evidence whatsoever for LinkedIn and Facebook having a secret bidirectional data sharing relationship.

The far more intelligent explanation is that suggestions are, as LinkedIn repeatedly emphasizes, based on imported contacts. Note that even if you don't import contacts, the other person might have.

Try using Occam's razor before jumping to unprofitable conspiracy theories. (Again, why on earth do you think Facebook would voluntarily share data with Facebook?)


> Again, why on earth do you think Facebook would voluntarily share data with Facebook?

Because why not? Facebook isn't really competing with LinkedIn right now - the latter doesn't try to enter private life, and the former doesn't really touch work life much. Both could benefit from sharing data with each other - from better recommendations to creation of shadow profiles. I don't see how this is an unreasonable theory.


This is insanely creepy...


Another possibility is that your social graph might have more overlap than you realize, even if it's not very much, it could be enough to make them the 'nearest' to you in some way. At least some of the graph data can come from sources invisible to you (i.e. people looking at other people's Facebook connections).


Yes, that could be an option too, though in this case it would be surprising given that this was a new subcontractor of our actual client, in a place neither of us have ever been before. But I recognize the phenomena you mention; I sometimes see random people on Facebook with surprising amount of overlap with my social graph, even crossing the groups I consider mostly unrelated (say, a person is a friend with few of my buddies from university AND some other unrelated group of locals AND also knows that random person I met somewhere that lives on the other side of the country).


That does seem hard to explain. It's creepy sometimes to wonder just how much they can figure out from how little and where all the information siphons are actually located.


Honestly, I find it both empowering (oh the data I can find if I know where to look) and reassuring. Why reassuring? Because it reminds us that lies can and will be discovered. Some people like to think they'll get away with one but it's only a function of how much someone else cares. But reality is self-consistent and in the Great Web of Causality falsehood is just damage that needs to be routed over.

Conversely, lying is a malicious act. It'sdirectly sabotaging the model of the world and reasoning capabilities of other human beings, with recursive damage being dealt as that person propagates or reasons from false information. Unless placed in life or death situation (std::jews_hiding_in_your_basement), you really shouldn't do it. It boggles my mind how lightly people treat spreading falsehoods.


I think things will eventually come around once reputation in life is unavoidably attached to everything they say, though, but it might be messy getting there.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: