Unencumbered API access to the social graph is exactly what led to the Cambridge Analytica scandal, and promising to restrict API access was part of Facebook's settlement with the FTC.
Perhaps a "social graph" shouldn't belong to any company. The concept itself might be anti-privacy.
Perhaps it's time to move back to simpler things like OPML. (Which people seem more comfortable sharing freely - not sure if that's still a "social graph" tho.)
Perhaps specific relationship details shouldn't be housed in the social network - just the permission to follow.
Perhaps companies who house a "social graph" should be regulated - similar to what is done with medical data. It should be a toxic substance.
I just think we're a long way from doing this right and there are a lot of untried options - and I lack imagination as well.
Good riddance to them.
Which one is propaganda? The one you disagree with? The one that doesn't fit with your preconceived notions? The one that doesn't fit with your personality profile? Both positions? Neither position? Do we just shut down all thinking on the internet? Who watches the watchers?
> In the 20th century, the term propaganda was often associated with a manipulative approach, but historically, propaganda has been a neutral descriptive term.[1][2]
Letting people be mislead about vaccines harms people, letting people get brainwashed into mistrusting the legitimacy of their institutions harms people, letting people use their platform to incite genocide harms people.
They actually have metric to measure those things, it’s just that they prefer to dial down the heat, not kill it. It’s just too good for engagement.
Don’t “both sides” death and misery. It’s not a philosophy debate.
To be fair they are clearly fighting a delicate balance of censorship and community moderation. It's different from say a forum of yesteryear, because a forum can have a clear set of community values that should work for everyone.
Facebook and all social media is a single platform for many communities, and so given the new role of moderator they are in the impossible position of making a set of community values that apply to all communities.
I think the reason a lot of material isn't outright banned where you or I would see obvious misinformation or hate is because of this.
I urge anyone here to define metrics for harm reduction that are operationable.
If you get any far with that, then tell me how you feel if the tools that will achieve these operational metrics were inverted in their purpose.
I’m surprised they don’t go after it as a violation of their terms of service.
Also, not asking right wing “news” orgs to be fact checkers or having GOP operatives override content policy team decisions regarding misleading stories that were reported by users would be another great way to reduce “propaganda”.
It’s not a technically unsolvable problem, it’s just one Facebook doesn’t want to solve. Don’t give them a free pass.
They are a multi billion dollar company, they have the all the ressources they need if they cared.
If we're not okay with other companies having access then why is it okay for a single company to have that access? Do you know and trust every facebook employee including those that don't work there yet with your data? Do you trust that that data will be securely handled when the company is sold after declaring bankruptcy?
I think your objection is an accidental straw-man. I don't think it was intentionally made, it is a valid thought to ponder but it should exist independent of the question of sharing data.
Should Facebook have all the info your bank has for instance? Or auto mechanic, your realtor? This is a slippery slope of logic of shared data for all leads to.
Some of these "social graph" features will do things like link two people as friend suggestions if they both have your phone number in their contacts book, even if you haven't consented to Facebook building up some information about you.
This doesn't sound like willing access & knowledge of what information you provide to Facebook.
PS by fighting this point you’re merely pointing out that giving more businesses this data isn’t a good idea
It doesn't have to run on companies spying on their users either.
Sam Walton of Walmart would record everything about his competitors legally and illegally as well as clients yet you never see that in the news, why? Because it’s not sexy, it’s not about the internet.
This is why Equifax which was a much more horrible data breach isn’t thought about, yet everyone gets their panties in a bunch about Facebook
The ironic beauty in that would be awesome. Also, I'd kind of like to get it about myself for similar reasons :)
They had no real idea what else it would or is being used for.
They have to monetize and I approve.
Hell I’m willing to bet if Facebook puts into their agreement that every photo you upload to their platform is partially owned by them then people will still upload photos.
It’s not the data that we find valuable, it’s the connections they make or will make, and to this company. Next time you take a photo and find it’s shared with not just Facebook but a whole slew of other companies you tell me if that’s the right thing to do.
The person above said “do you trust all Facebook employees” well you can make that blanket argument for every company in existence and giving more companies our data isn’t a good solution
I mean Facebook might perform facial recognition on your photos and infer connections to other people you didn’t explicitly tell them about.
I mean that facebook might track your location and app launches and identify where you go on vacation and when you shop.
These are the kinds of thing nobody gave consent for.
The consent thing can both be good and bad, just like the invention of the internet and electricity before that. I’m trying to make that point that giving away your data is gonna be a commodity once everyone realizes they’re not a special snowflake
The difference is that Disney grants license to its content after (potentially drawn out) negotiations and payments, while the typical Facebook users "Agrees and accepts to grant Facebook a worldwide license"
BTW it doesn't confer a license (mostly) to anybody else but normal users do not sue.
Facebook builds shadow profiles on non-users and then fills them out with, amongst other things, data it gets from their actual users that they didn't no they were sharing.
This is particularly sneaky when Facebook has used dark patterns to trick app users into granting them permissions they don't strictly need for service.
Check out LiveRamp for example; they literally scan in physical documents like car mechanic paperwork (seriously) and dropship them into S3 buckets to huge data brokers every day.
Take that and merge it with facebook’s social graph, how many ms you look at a photo on Instagram and how long you take to reply to someone on WeChat, and you have an awful lot to work with.
What difference does that make? Facebook almost certainly knows my banking information, without me giving it directly to them
Privacy is in the process of becoming antiquated. It is sad, but I think it is inevitable. Surveillance technology advances with technology in general, and counter-surveillance becomes more and more cumbersome. Eventually, we will reach a point where anyone can afford tiny stealth drones to watch what is happening in their neighbors' houses. The counter-measures necessary - hermetically sealed living space, faraday cage enclosure - are so burdensome that only people who currently live in bunkers will implement them for their living spaces.
If we accept that premise, let's think about what we are losing. In my understanding, there are two main reasons why privacy is so important:
1) Allowing individuals to avoid becoming targets of persecution for having some attribute. Targeting by oppressive governments for holding idea X, targeting by an evil megacorp for threatening to disrupt industry Y, etc.
2) Allowing individuals to avoid social stigmatization and shame for having some attribute. Getting caught eating one's boogers, viewing porn of people dressed up as pieces of furniture, etc.
For point 1, I think the real issue is imbalance of information. Schemes to abuse access to someones previously private information mostly wouldn't stand up to public scrutiny themselves. Unless, of course, you have some ultra-powerful organization like a well run totalitarian regime, in which case individuals by definition have no privacy rights anyways.
For point 2, I think shrinking privacy will seriously reduce the social response to this kind of information. "So what if I watch porn of people dressed up as furniture; you masturbate while dipping your toes in peanut butter! That's way weirder!" Sure, there are plenty of painfully normal people out there, but I think most shameful weirdness would become commonplace when the full extent of weirdness is out in the open.
All that being said, I do think anonymity is important, and can probably be preserved, as it is distinct from privately acting in physical space, or online while tied to some form of identity. Also, I don't want to argue against efforts to preserve and advance privacy for individuals for as long as possible. I just think that per point 1, we might want to devote effort to reducing the privacy of organizations to try to maintain parity of information access as our individual privacy slips away.
> The counter-measures necessary - hermetically sealed living space, faraday cage enclosure - are so burdensome that only people who currently live in bunkers will implement them for their living spaces.
There are a lot of situations we forbid activity that is difficult to prevent (and sometimes difficult to detect).
Whether trying to use the law to deter here is worthwhile, meaningless, or harmful isn't necessarily clear - but it deserves consideration in addition to outright prevention.
It definitely somehow limits their use, but anybody really interested is able to obtain them illegally, or produce them themselves.
I'm afraid the same will be true about micro-drones or whatever else the post suggests.
It's different from drugs in that one of the participants is unwilling, though I don't know whether that's a big enough difference - particularly given that that participant is also unwitting.
I'd say it's not ok, but people keep on giving Facebook all that information (and it's very tempting to do so). There are technical solutions which anonymize/privatize social network information, and then no entity has such access.
You can only briefly see which friends have also already signed up.
So while I value privacy, I'm not sympathetic to using privacy as an argument to restrict your own access to your Facebook data, for you to decide yourself on how that's shared. If you misuse your view of your friends' data, that's on you and up to your friends to punish/discourage you for.
Your sharing your FB friends data with third parties, could be seen by the friends as impolite or even hostile. It's a matter of trust between friends, not technical ability.
I'm not defending VK — it does some increasingly user-hostile stuff too after having been acquired by Mail.Ru Group — but you can still integrate it into whatever the hell you want with ease.
How would you know?
The worst thing to ever happen with that open data was that some debt collection agencies used it to threaten people and their friends, and I believe that's why those service tokens were introduced. And that story was rather widely publicized: https://www.the-village.ru/business/finance/217569-banki-pis... for example.
https://towardsdatascience.com/steeling-faces-for-investigat...
https://nakedsecurity.sophos.com/2016/05/02/facial-recogniti...
Facebook itself is a third party.
I think a case could be made that knowledge that <A friend B> belongs to A and B equally and that neither has standing to preclude the other from sharing that fact. Of course they could both mutually agree to such an arrangement ahead of time, but outside of such an agreement, I have a hard time seeing the case for "B prevents A from telling C that A and B are friends".
Another aspect of this though, is "how much can A tell C about B without permission from B?" We know that IRL, people do get angry when friends reveal overly personal details about their lives to others, so this is definitely something to consider.
I'm not sure how you got that from what I said above. Obviously there are such things. But I do think it's an open question to what extent we can compel our friends to keep things private. Of course if someone is really a "friend" you would expect them to keep confidences if requested... that's part of the definition of "friend" in my mind.
Agreed. And I am not arguing that it is, or should be. But the mere fact that two people are friends seems to me, at first blush, to be something that normally either person would be free to share.
If the API is open and there's forced interoperability, the client device can call the service's API directly with the user's own credentials. No need for a third-party middleman, so no need to give a third-party access to the data.
It can work the same way IM applications like Pidgin work: one app, multiple accounts.
Why? If they shared it with me, that should mean they trust me to decide to whom I delegate the access no? I could as well just screenshot it and post it to twitter, what does make the social graph so special?
The point is to move toward standards around social data and then force the big tech companies to allow each user to export their own data, quickly, in these interoperable formats.
There is risk, but saying that users can’t be trusted with an API to their own social graph seems to be throwing the baby out with the bath water.
[0] https://www.healthit.gov/topic/health-it-initiatives/blue-bu...
The counter-argument to this is having the government provide the graph and an API to allow networks like Facebook connect to it. This should rightly terrify just about everybody.
I like the concept Diaspora had back in the day, it just wasn't workable really as your mum/aunt/non-tech friend wouldn't be able to understand how something like that works and would ultimately lose/delete/corrupt their tokens.
Multiparty secure computation is one possible way. There are a few, emerging implementations.
> To further strengthen the control over his or her own data, where the processing of personal data is carried out by automated means, the data subject should also be allowed to receive personal data concerning him or her which he or she has provided to a controller in a structured, commonly used, machine-readable and interoperable format, and to transmit it to another controller.
The problem is that if I export my social graph and share it with another company, it includes information about who my friends are -- at the very least, it shows that they are my friends. That violates their privacy. Maybe they did not want their relationship with me to be known outside the social network where they established it.
Example, you are chatting at a party with someone and they ask you if you know Bob since he works at your company. You say yes and you say you are good friends with him. You violated Bob's privacy here?
If the question were framed more like:
"Do you consent to sharing your entire address book, so that we can better market to you and your friends and offer targeted ads based on that data to our real customers?"
Would most people actually answer yes?
Presumably you would be granting this system access to this information.
>Would most people actually answer yes?
some people would, do they not have that option?