Should social networks tell you if you’ve been following a fake account?

https://shkspr.mobi/blog/2020/09/should-twitter-tell-you-if-youve-been-following-a-fake-account/

by edent • 6 years ago

89 106 6 years ago

dreen5• 6 years ago

Unless you require a proof of ID, theres not gonna be a reliable way to identify fake accounts. You'll get the same sort of algo-driven mass mistakes we now see in Youtube.

rootsudo• 6 years ago

Only because the platforms need to rid fake accounts to keep engagement high.

Makes you wonder, would the platforms ever be truly transparent with their product?

Example: Soylent Green.

mcphage2• 6 years ago

They may not get identify all fake accounts, but there are still tons of accounts which can be clearly identified as fake.

dreen5• 6 years ago

My point is that proof of ID or ownership should be implemented, as that would get rid of the problem.

Edit: I need to point out that this should be optional. Anonymity should still be the default.

notahacker• 6 years ago

This assumes they care. Even Facebook, the social network that used to kick you off for not using your real name, now informs you that low effort sex-bots (nothing but a pouting stock photo, a single post inviting you to join their 'adult dating' website at a dubious-looking link and a lot of random friend requests) don't breach their guidelines in any way if you report them, which presumably involves an actual human in the loop.

Other platforms are of course more favourably disposed towards multiple accounts, automation assisted high volume posts, shifting identities and parody accounts

edent2• 6 years ago

Not all accounts are from people, though. Some accounts claim to be newspapers or technical organisations.

dreen5• 6 years ago

This can still be proven with some sort of document or proof-of-ownership

olah_1• 6 years ago

Have you heard of the BrightID project? I think it could be a pretty cool way to determine real identity without leaking info to the network at large.

dreen5• 6 years ago

I haven't and admit I don't have the expertise required to understand it unfortunately. I can't really see how it could work to actually prove someone is not a bot in large scale social networks. Like, what's stopping three real people to get verified and then use that to verify any amount of bots?

quelsolaar2• 6 years ago

More importantly it should retroactively tell me about what I have read that is shown to be fraudulent. "On September 4th you clicked on link "Hillary Clinton runs child porn ring", this article has been fact checked as misleading and the account sharing this has been identified as a fake account operated from Russia".

If they do look for fake accounts and misinformation, its not enough to remove the misinformation they have to inform people they have been misinformed.

edent2• 6 years ago

I agree. But there is a "creepy" factor that a social network knows what you click on. I suspect lots of people are unaware of that.

Would you want a librarian chastising you for the books you check out? Would that make people more or less likely to use the library?

Wowfunhappy3• 6 years ago

Would you want a librarian chastising you for the books you check out?

If what I checked out was fake? Yeah, personally I would like to be told.

thwarted• 6 years ago

The librarian analogy is a little stretched, but I do expect librarians to properly put fiction books in the fiction section and nonfiction books in the nonfiction section.

TeMPOraL• 6 years ago

Chastising? No. But if there was a service available in which the librarian would tell me when a work I've borrowed has been criticized (or even thoroughly debunked) in some other publication, I'd definitely opt-in to that.

toyg• 6 years ago

It’s just a matter of careful wording. Instead of “you clicked on <bad source>” just say “the source linked was bad”.

mschuster912• 6 years ago

Would you want a librarian chastising you for the books you check out?

With cars that turn out to be defective, in Germany the authorities can issue mandatory recalls and notices to owners of cars of the recalled model.

If I borrowed a book at a library or read a scientific study that has been later proven to be wrong or outright misleading, yes I would like to be notified of that.

alkonaut• 6 years ago

Would you want a librarian chastising you for the books you check out

No I wouldn't want anyone chastising me, nor any contact with a human (that can be awkward). But I'd be happy to get an email saying "Please be aware that the book X from the science section that you borrowed earlier has been found to contain inaccurate information and has been moved to our controversial science/science fiction section".

Retractoins occur all the time in newspapera but they build on the model that people actually read the newspaper. It doesn't work when people read one article from a social media link. Fact checking and retractions should be updated to work in the age of facebook 1) Respectable media would publish retractions to facebook, saying "Please notify anyone who followed the link to reuters.com/article/123 that the article has been corrected". 2) Fact checks would do the same thing (because this is most important with less respectable media).

For this to happen obviously facebook would want to agree to annoy a fraction of their users who don't want to find their favorite links unavailable, littered with warnigs, or followed by messages that they are eating an unhealthy news diet. Hopefully they realize that this is a goal worth pursuing.

chrononaut• 6 years ago

Maybe I'm misunderstanding the analogy here but I would expect the library to be fully aware of what books I have checked out. If I don't return a book, they would nag me about it. Why would that be strange or unusual for someone?

acdha3• 6 years ago

Is this a librarian chastising you or, on your next visit, telling you that the reference book you’d checked out had been shown to be a work of fiction and recataloged under conspiracy theories? Libraries have plenty of books with inaccuracies and they haven’t contributed to the downfall of popular discourse.

tzs• 6 years ago

To make the library analogy match how social media works, the book should be one you checked out because the library recommended it to you earlier based on books you had checked out earlier.

throwaways885• 6 years ago

There's been too many instances of people with edgy opinions banned for being fake, so this could make things worse.

spacebear• 6 years ago

The response you imagine somebody having to that alert: “oh wow, I’ve been mislead. I should reevaluate some of my thinking”

The response somebody actually has: “Twitter is part of the pedophile ring and also they’re watching me.”

ZephyrBlu2• 6 years ago

It's depressing how stupid we (Humans) are...

Cthulhu_• 6 years ago

Exactly; they don't trust fact checking, because to them the word "fact" has been co-opted by The Man. There's no reasoning with some people.

Actually there's no reasoning with most people; most people can't be convinced that e.g. the earth is flat even when presented with loads of evidence. Speaking for myself, I don't feel like I've changed my opinion on a lot of things for a long time.

TheSmiddy• 6 years ago

If this results in them leaving twitter then is that not a positive?

We can't save the people who have already gone down the rabbit hole but we can prevent more jumping in after them.

quelsolaar2• 6 years ago

I think it really matters. A lot of people are being misled in a much more subtitle way, and aren't full blown conspiracy theorists. I think it would make people more careful, if they where told that what they have read in the past have been proven false. I think there are more people like me, who thinks we have reasonably good idea of what to trust, but probably have read some fake news we haven't picked up on, and would like to know what was fake to help us avoid it in the future.

Barrin922• 6 years ago

sure the diehard conspiracy people can't be hlped even with that information, but there's a non-zero amount of people who will understand that they've read false information.

I don't really see the downside to correcting false information. If someone's too deep down the rabbit hole fine, but there are people who are not.

deanclatworthy• 6 years ago

This is a great idea. I wouldn't go as far as requiring fact-checking of the fake accounts, but just an indication that what you read may have been inaccurate and point out all that you have read.

sigsergv3• 6 years ago

"Hillary Clinton runs child porn ring" — is not a fake data, it's a Russel's teapot, you can't fight it with reason. And marking as a fake is a bad move because it inflates notion of “fake”.

101008• 6 years ago

Completely honest question, how is that not fake data? Or you think there is a dinstction about fake news and fake data? Would you mind sharing this dinstinction? Thank you!

acdha3• 6 years ago

This is true in abstract but most of these claims have details which can be tested. You can’t disprove a perfect conspiracy but you can show that the named people don’t exist or weren’t in the place at that time, the “proof” image has incorrect details (e.g. remember the birthers alleging Obama had a birth certificate which mixed the name and seal of different countries and time periods?).

If it has enough detail to be tested, you can more justifiably mark it as fake. The vague unfalsifiable stuff is much less satisfying for people to spread.

IncRnd3• 6 years ago

They will probably never expose their decisions to the wide world in that manner. That would incur liability for incorrect decisions by publicizing them.

intended2• 6 years ago

Do it.

Don't expect anything significant.

Cigarettes have images selected to be revolting or upsetting to keep people away.

That is not the reason people stop smoking cigs, or the reason cig consumption is down.

Social reasons, huge taxes and a concerted effort to make this unviable for a majority of people, in combination has made the dent it has.

And yet, people smoke today.

mistermann2• 6 years ago

this article has been fact checked as misleading and the account sharing this has been identified as a fake account operated from Russia

Something I've always wondered about stories of accounts being portrayed as "from" a certain country, my impression has always been that things like location and ip address data can be spoofed to make it appear that you are in a location other than where you actually are.

Does anyone happen to have technical depth in this area? Is it known with high certainty that it is impossible to spoof this information and then make posts from a social media account (Yes/No/Maybe)? (And, how is it known?)

smitty1e2• 6 years ago

Some of the great Twitter accounts (e.g. Titania McGrath) are utter frauds.

The ones that irritate are the Stolen Valor type, where people claim military service. The game tends to fall apart quickly, though.

The reason is that these experiences are binary. Think sysadmins: you've either done it, and can sustain a conversation about Linux or you haven't.

Fakery has a short half-life.

sshine3• 6 years ago

Fakery has a short half-life.

But an infinite regress.

ZephyrBlu2• 6 years ago

Some of the great Twitter accounts (e.g. Titania McGrath) are utter frauds

Hmm? I thought Titania McGrath was a parody account. How is it (And other accounts) a fraud?

smitty1e2• 6 years ago

I was using the two terms synonymously.

mdpye• 6 years ago

I think this would probably be a good thing, but that it won't happen because it will reduce engagement and trust for the platform by drawing to attention how full of garbage it is - not in aggregate, lots of people know that, but in your immediate sphere, which I suspect most people like to believe is cleaner than the average.

So: good for the world, bad for the platform. Not gonna be implemented by the platform owners...

minot• 6 years ago

I am not even sure if it is good for the world. When presented with evidence to the contrary, many people might simply dig in to their existing beliefs.

andy_ppp3• 6 years ago

I would go further - I want to be able to nominate people who I follow and Twitter should tell me if they have marked Tweets I have interacted with as untrue.

Why can't we peer review this stuff and let the "market" decide what is true. We know that this can work quite well and on average it should lead to better results than the social networks trying to do this themselves.

You can even gamify voting things as untrue - each person could have a page rank of things they have said are false and if they consistently downvote "the other side" for things that are true their voting ability will be proportionally reduced.

sJ646U9k6c6gME9• 6 years ago

Why can't we peer review this stuff and let the "market" decide what is true.

In a sense we already do. Right-leaning and left-leaning groups already believe mostly non-overlapping sets of “alternative facts”. Each side builds its own narrative not so much because of hate for the other side, but because in aggregate the total amount of misinformation has eroded the potential for credibility of any person or statement that doesn’t come from one’s own “in” group.

In today’s Internet you build your own picture of what’s true. Everything else is noise.

andy_ppp3• 6 years ago

When I see “Hydroxychloroquine is not able to cure COVID-19” is it better that warning is flagged by the community or by Twitter. I prefer a community solution; saying it’s not possible is IMO a lack of inventiveness around how the tagging could work. This also scales much better.

jpttsn• 6 years ago

Poe’s law means I don’t know if I’m engaging in good or bad faith here but...

Your suggestion sounds like letting the ”crowd” decide; basically voting on the truth democratically. Religion is (kind of) an implementation, where people can “vote” for the denomination whose truth they prefer. They vote with their feet or with their guns, depending on surrounding historical context.

Letting the “market” decide would probably require some additional mechanisms, like financial reinforcement. Advertising is (kind of) an implementation: the more money you spend, the stronger your signal will be. Alas, it’s an incomplete implementation and pernicious ground truth often has an undue influence.

andy_ppp3• 6 years ago

You could try peer reviewed fact checking, Wikipedia seems to have managed okay, so it’s possible even if the terminology and models above are not.

sshine3• 6 years ago

All your facebook friends are fake friends.

Delete it and go outside.

dvh• 6 years ago

Why not simply delete the fake account?

martin_a3• 6 years ago

What would be the benefit for Social Networks?

Their metrics are active monthly users and ad revenue. The more buzz is going on, the better for them.

They will sometimes have to delete the most obvious fake accounts, yeah, but besides that they couldn't be more happy when fans of the People’s Front of Judea have virtual battles with the fans of Judeas Front of People (i know the joke, but I don't know it in English, just German, hope this translation is correct).

All of that generates just more buzz, drives people to the platform, generates clicks, pageviews, ad revenue.

Social Networks will always only do the bare minimum to fulfill (the rather low) requirements of the law and let the rest flow...

sammorrowdrums• 6 years ago

Blessed are the cheesemakers!

You were almost spot on. I believe it's the Judean People's Front. The People's Front of Judea is correct.

Spare a coin for an old ex-leper?

martin_a3• 6 years ago

I would be interested in buying either a pack of small stones or ocelot ears. Do you have any of those?

vonmoltke• 6 years ago

Serious question, to you and the many other people who make points like this on these stories: what makes you think these pissing matches and other toxic engagements increase ad revenue? Why do you think people in the midst of these will engage with the ads? What advertisers do you think want to have their ads and brands associated with these engagements?

martin_a3• 6 years ago

what makes you think these pissing matches and other toxic engagements increase ad revenue?

Because the number of monthly active users/the number of pageviews gets higher. That's a talking point when doing ad sales. It's about potential reach of advertisements, which also defines the price of your campaigns.

Why do you think people in the midst of these will engage with the ads?

They probably won't. You'll see lots of reach but no conversions. That's the point where ad-agencies or the social networks itself jump in and "help you" "optimize" your ads for conversions. Maybe some of it will even stick in the end, who knows. Anyway, some people along the chain made money.

What advertisers do you think want to have their ads and brands associated with these engagements?

Do advertisers really know? I think in most cases your targeting is kept fuzzy enough so you're not able to pinpoint it down to a post level or whatnot, at least you will reach your "target audience". (That also sounds great when your marketing presents the numbers: "Big hit in the target audience, but no conversions yet! We'll have to fine tune our ads even more...") Whether they are sharing cat pictures or are discussing genocide... You'll never really know in the end.

Of course there was some uproar when companies found their advertisments as pre-roll ads on the YouTube videos of ISIS, but that's a storm in a water glass. Nothing to really worry about after all.

dutch30002• 6 years ago

lame. the real problem is the algorithm generated feeds that apps like instagram, FB and twitter use. their main goal is to occupy the eyeballs for as long as possible. this manifests in people being sucked deeper down into their own echo chamber. yes, that echo chamber is filled with garbage, but the garbage lined, endless pit of media streaming and it’s impact on our society squarely originates at the core and maniacal focus of SM platforms to to increase app use time. period. everything else is window dressing.

kyrers• 6 years ago

I assume the fake account will be deleted, which means I will stop following it.

However, I do think they should notify every follower of said fake account, letting them know that what they've read might be false.

sjg0072• 6 years ago

Yes.

mikece2• 6 years ago

Caveat emptor. It says a lot about people on social networks that they can't see through a fake account. Either they don't have the intelligence to discriminate between fake and real or they are so easily led about that anyone -- even fake people -- can sway them one way or the other. The latter is what scares me the most, especially when they vote. The naive and easily swayed are the very people who elect and empower persuasive dictators and tyrants. Social media isn't "the problem" but it's accelerating an existing problem.

intended2• 6 years ago

https://spotthetroll.org/

Fake accounts, especially for swaying public opinion are not that easy to spot.

I failed at spotting several, and I used to spend more time than average dealing with user content.

It's hard to look at an account and simply say "this isn't a person.", perhaps more so if you have a broad view on the range of possible human behavior.

Maliciously formed content works hard to be indistinguishable from normal behavior, Its essentially a pseudo turing test. Can we mass produce social media accounts that pass for human to both humans and automated systems.

I'm really not sure how someone can identify this unless they are trained.

mikece2• 6 years ago

My larger point is that people who cannot distinguish between reality and not-reality are an extreme drag on a society as they require others to think for them. Such people shouldn't be on social media and shouldn't be voting... it's how we end up with the politicians in office that we have now. People who have a solid grasp on reality and their own convictions won’t be swayed by online activists whether they are flesh and blood or bits and bytes.

dingaling• 6 years ago

Fake accounts, especially for swaying public opinion are not that easy to spot.

If you don't know the account holder in person, or through word of mouth in a social context, don't follow them.

mark_l_watson• 6 years ago

My answer would be yes, or rather not a binary yes/no but the probability from their models that an account is fake.

sjg0072• 6 years ago

People don't understand probability. Clinton had a 70% chance of winning in 2016.

akerro• 6 years ago

So don't say in probability but in stars:

Chances this account is a bot: (I had utf-8 stars here)

* * */- -

awinter-py• 6 years ago

should ebay tell you if you purchased something from a fraud account and it's never going to arrive?

sneak• 6 years ago

Those are real accounts, run by real people: people who are spammers or propagandists. The idea that there are “fake” or “bot” accounts is just pro-censorship nonsense. There is no AGI, these are just alt accounts for human beings (that potentially have an agenda).

I, for one, don’t want Twitter, or anyone else, arbitrarily deciding what (or from whom) I’m allowed to read.

Twitter’s shadowbanning a lot of terms from search lately, especially those related to conspiracies or organized, armed groups (even groups without histories of violence).

It’s becoming really terrifying to not be able to get a clear or complete picture of the relative size/penetration of these memes, or to know what these groups are saying or being told (regardless of whether I buy into their ideologies or not).

Keeping tabs on what’s going on in one of the most volatile and potentially dangerous times ever is now being actively hindered by the social platforms, and it’s terrible.

We do not need more of that, and we desperately need less.

Better scoring, grouping, tagging, and classification tools would be GREAT. More censorship would not.

Should social networks tell you if you’ve been following a fake account?

Related Stories