Instagram Has a Massive Harassment Problem
theatlantic.com
theatlantic.com
People wildly underestimate how dedicated and adept some people are at finding PII given even minute details [1]. The only real way to avoid sharing PII is to stay off the internet entirely.
Even then, pandora's box has already been opened. Your PII is available even if you haven't explicitly chosen to share it [2].
Once it's out (and spoiler alert, it's out), "stepping away from the computer" is no longer an option. The article describes in detail how it transitions from online to real-life harassment: Swatting, stopping by your personal residence, calling your work, calling your friends and family, etc.
Yes, it happens to people of moderate means and very minor celebrity. Yes, you can suffer physical harm or death as a result [3]. Yes, we should also try and solve the swatting problem from the police side as well.
1: https://www.dailydot.com/unclick/4chan-shia-labeouf-secret-l...
2: Equifax
3: https://www.cnn.com/2018/01/30/us/kansas-swatting-death-affi...
The moral of the story is that no matter how careful one protect their identity, they always leak more than they thought they are leaking.
I feel bad for the new generation of Twitch/Youtube people who are exposing themselves to the Internet. There are some freaky good stalkers out there.
It starts even before they expose themselves in the form of parents who post everything about their kids on Facebook.
That doesn't quite do the situation justice. The stalking gets crowdsourced; often to people who are well practiced at it.
> a flag emblazoned with the words "He Will Not Divide Us" would be flown for the duration of the presidency. The camera was pointed up at the flag, set against a backdrop of nothing but sky.
> Within 38 hours of resuming transmission, the flag was located by a collaboration of 4chan users, who used airplane contrails, celestial navigation, and other techniques to determine that it was located in Greeneville, Tennessee.
https://en.wikipedia.org/wiki/LaBeouf,_R%C3%B6nkk%C3%B6_%26_...
Or, for that matter, extradited.
That doesn't always work either. A lot of PII shared completely offline finds its way to the internet, and third party hacks make it accessible.
I don't necessarily have a perfect solution all lined up, but just correctly identifying the problem is step one. I've mentally noodled with a number of models. But the thing they have in common is that none of them start with a new user being able to fling whatever they want to whomever they want for free.
[1]: I don't actually consider it a failure of the system if you encounter an asshole every so often. That's life. But when you start a video stream and you end up with so many assholes typing asshole comments so quickly that you literally can't even read them before they scroll away, even if you want to, that's a structural problem, not just a social problem.
I suppose I should point out explicitly that my goal is not to create a 100% comfortable world in which you never encounter an asshole, but to create something a bit more like the real world where by the time you're an adult and you've found a place for yourself in the world, by and large you probably don't have a problem dealing with lots of people-you-consider-assholes personally. (You may professionally, but that's a different sort of problem.) Even if you maybe have a problem with one or two people in particular, you almost certainly do not set a foot outside your bedroom door and immediately get several thousand people screaming at you at the top of your lungs. But you're going to encounter a few, sure.
More on the Reddit model, I outlined something here: https://news.ycombinator.com/item?id=16570017 The observation that it is computationally expensive is true, but I'm not convinced it's intractably so. There the idea is that if some anklebiter starts replying to your every comment, by default you at most see something greyed out or collapsed if you did not have any previous connection to them, and they can't bother you. I think to jumpstart the system you'd always have to have a couple of random comments show up, and some of them may be assholes, but it wouldn't be the case I was describing where you can get this unrelenting torrent of abuse.
(I think such a system would have more manipulation resistance, too. $MEDDLER could go to some political forum that works that way, and maybe they could convince a few people or a subcommunity to upvote them, but they wouldn't be able to completely take over an entire topic by sheer mass the way they can on Reddit. You can't get a world where $MEDDLERs don't exist, but you don't have to have a world in which only $MEDDLERs and $MEDDLER-approved accounts exist on your forum.)
This change of communications with assholes rate brings asymmetry that is a cause of the problem: you still have the same response capacity and emotional response decay time as in the offline world, but the incoming rate of assholes is so huge, so you can't cope with it anymore.
So perhaps any solution should be seen from this perspective: decreasing the upper boundary rate to some reasonably acceptable value.
But it also may lead to the opposite conclusion – we now in an online world, with different speeds and rates of everything. Maybe solutions "like in the real world" simply won't work anymore here. Perhaps we can use some new tools to solve these sorts of problems, that we never had in the offline world?
For example, there was a brilliant idea to set a price tag for your attention in the online world. Imagine an email system where the sender should pay some price for you to receive it and read. This price can be dynamic and vary according to various parameters – from sender's identity to my current mood. If someone sends me a public speaking request or is trying to hire me, they'll pay like 5$ per mail, but for my friends or coworkers asking for help it could be 0.03$. And if my inbox is empty (ha-ha) and I'm in the mood to explore who's writing me, I can temporarily lower the price for all. That's of course, just a rough sketch of the idea, but it might work in the future if cryptocurrencies become ubiquitous for example.
Making this a market seems like an interesting idea.
Maybe the cost is paid in non-monetary ways even (can't message much when anonymous) and maybe the cost is along a continuum depending on how much anonymity is requested (email, but no real name, or location and age but no personally identifying info).
"A peanut gallery was, in the days of vaudeville, a nickname for the cheapest and ostensibly rowdiest seats in the theater, the occupants of which were often known to heckle the performers." (Wikipedia CC-SA, via Google)
That's a new one on me.
If a cost is evenly moderately costly to create, and abusive comments quickly result in bans, then the ratio of abusers quickly goes down to a very low level.
The problem is that making it costly to sign up goes against the business model and growth model of sites like Twitter, Reddit, or Instagram.
The problem is that it makes it hard to get enough of a critical mass to convince people that it's worth the ten bucks.
That being said, something awful did it well over a decade ago [1] and has a relatively low amount of assholes/harassment, but it's not zero.
On the one hand you have google/facebook pushing people to reveal their real names on accounts... on the other you have sites similar to this one where you don't have to reveal much, but rely on karma/reputation a bit more.
In the end, it's a matter of dealing with the best and worst of mankind. As sad as it is, bullying didn't start with the internet, and it won't end with it. It sucks to be on the receiving end, and although I, personally, haven't always been the best at dealing with it, people need to learn that skill.
The world, despite best optimistic intentions, is not a kind, fair place to be.
To get linked, you initiate a the "add me to your group" request. This request is just a flag which asserts "X is interested in linking with Y". It cannot be repeatedly asserted, and no text or images may be attached to it. (If someone wants to harass you with these contact requests, they will have to keep creating new profiles, and fill those profiles themselves with hateful content targeting you.)
The system could use Bayesian filtering on profile content to determine which kinds of accounts you don't like linking with.
If you take away or limit that, what are you left with? will your users be content? and more importantly, will advertisers keep spending the same or more money on your platform?
How do you propose to have any conversation, even like this one, if you don't see anyone responding to you, and they in turn don't see you?
The "metaverse" . Where you can be what you want people to see w/o others knowing what you are. #highfidelity
That is a nice idea but I'm not sure it will work because consider Stack Overflow: most users there are little more than a number and a geometric shape, as far as what others see, and yet they still felt they needed to formalise a code of conduct based on survey results that certain demographics felt unwelcome.
Another way to look at it is, good comments are capable of inspiring more good comments, but low-quality comments always seem to lead to more low-quality comments. So comment quality in general is also subject to entropy in most forums and social networks that I've seen.
Some combination of moderation and lacking appeal to general audiences allows specialized forums like HN to reach a steady state, with probably more assholes than desired and fewer good comments than desired, but at least it's not continually hurtling towards collapse.
> Like Twitter, Instagram enables the easy setup of endless anonymous accounts: All you need is an email address and you can start posting within minutes. Abusers leverage this functionality to create armies of fake accounts to attack people. But while Twitter now allows users to protect themselves—by muting replies from people who don’t follow you, who you don’t follow, aren’t verified, haven’t confirmed an email address, and more—Instagram has only implemented some of these controls.
Instagram had a nicer overall "personality" to it initially because it was centered around the posting of attractive, innocuous photos. But in the end, it's still a platform that allows anyone to disseminate information of any kind (not just comments, but screenshots and images of text memes), so there's no reason to think that as it becomes massively mainstream, it won't have the exact same problems as Facebook and Twitter.
edit: fixed grammar
Is the "easy sign up" such an important selling point, that anything more complicated drives away users?
I'm wondering how could be possible to avoid hoarding burner accounts just for the sake of trolling, while at the same time not requiring 1Billion people to send in scans of their ID to prove their identity.
Not to mention that even requiring ID for proving identity, one can still troll and harass as long as it's not caught on the fact
I use Instagram, however, the only "harassment" I receive is spam and occasional mentions from p_rn bots. I also keep my profile and I'm not trying to push, sell, or make myself a celebrity.
What I think that may be happening: There are a lot of people who are trying to leverage their message or themselves to a level of celebrity and they don't realize that this is one of the consequences of it (or that they refuse to accept it).
In saying this, I'm definitely not saying I do... I take a lot of things far more personally than intended, and can only try to not let my instincts drive me when I do. It isn't easy to be on the receiving side of things... that said, I don't think outright censorship is the answer. Couldn't this kid report/block other users on the system giving him grief? Also, for those blocked, probably shouldn't indicate that they've been blocked.
I've always found it funny that on Twitter, when someone blocks you, you can no longer see their public profile and tweets. It'd be far better to allow you to see it, but only the blocked see their own retweets/replies. I don't think the people making decisions always think through their decisions.
The problem with something like this drawing attention and potentially reacting too quickly is the chance to make things worse is far more likely than actually approving anything.
You seem to be pretty out of the loop when it comes to Instagram's primary user group. Many teens/early 20s these days don't even have a computer, a smartphone is all they use. They're definitely accustomed to phone UIs.
> I use Instagram, however, the only "harassment" I receive is spam and occasional mentions from p_rn bots.
Don't generalize from a single datapoint, especially when that datapoint is not a representative sample at all.
> What I think that may be happening: There are a lot of people who are trying to leverage their message or themselves to a level of celebrity and they don't realize that this is one of the consequences of it (or that they refuse to accept it).
This is just straightforward victim blaming without even trying to understand what is actually going on.
They may be accustomed to the UI but from using the IG App you don't have the full capability to be aggressively persistent on the phone as you would on the desktop. Yes, mean things do get said, but I would hardly believe that it's as easy as they claim.
2.
This was my experience. I'm a private individual whom doesn't use it to build an audience, sell anything, or any of the like.
3.
Celebritihood has it's upsides and downsides. Am, I saying it's a good thing that it happens no.
tl;dr the HN demographic is not the one that gets harassed on almost any platform, but especially IG.
Excuse me but what? I thought it was common knowledge that Instagram is one of the worst places on the net when it comes to bullying.
There are all kinds of things that can cause that to happen, famously redesigns and changes to moderator actions has done it in the past. But in general having a continuously shitty experience as people harass you is a pretty compelling reason to not use a social site. I cannot help but wonder if enough people will leave to trigger a mass exit or not.
http://wilwheaton.net/2018/08/the-world-is-a-terrible-place-...
https://medium.com/@AmberEnderton/wil-wheaton-has-a-listenin...
Wil Wheaton isn't a platform, he's a private user, and it's not entirely unironic that he "got" what he was promoting for others (being de-platformed)...
Their T&Cs prohibit having more than one account, so they would never implement this feature no matter how much sense it makes.
Maybe we need systems where you can either be anonymous and censored, or identified and uncensored. You get to pick.
They do however harm the most vulnerable in society, either excluding them from platforms or making it easy for their harassers to find and target them.
Anonymity is at worst neutral for harassment. Removing it doesn't do what people hope.
FTFY
Internet crazies swatting you because you got a portafort in fortnite, or some equally trivial offense like that, most definitely is a problem.
And that's not even getting into the more sinister stuff that the fbi has to deal with.
Woah, woah, woah.
I said internet crazies showing up at your place of employment is a problem. That doesn't make it assault.
For instance, the outrage brigade might show up to protest in front of the company where you work for a few weeks. This is their right as Americans, and I realize and acknowledge that. At the same time, I believe it is harassment. I didn't say we should do anything to stop it. I only said that it most definitely is a problem. And if it happened to you, I'd bet that right around the time your boss called you into his or her office to give you your final paycheck, you would start to think it is a problem too.
I think you went about 5 bridges WAY too far with calling a lot of that harassment "assault". A lot of the harassment is not assault. That's why it's so hard to stop. The assault stuff is the stuff that the fbi would be called in on generally speaking.
What's with nytimes, theatlantic and the rest of the media and their endless spam and whining about social media?
If you don't like it, stop using it? Leave HN alone.
HN has a massive nytimes, theatlantic, etc spam problem too. We don't have a way to filter useless clickbait media.
If the "journalist" at theatlantic don't like instagram, go start your own version of "nice" instagram and make billions.
You realise these things appear because HN users submit them, and other HN users upvote them?
Your comment is as inane and oblivious as me telling you to stop whining about The Atlantic, because it’s free and if you don’t like what you see you should just log off.
What's with liftbigweights and their endless spam and whining about interesting, on-topic submissions?
If you don't like it, stop using it? Leave HN alone.
liftbigweights has a massive HN spam problem too. We don't have a way to filter trolls.
If the "user" at liftbigweights don't like HN, go start your own version of "nice" HN and make billions.
People often (or at least used to) share their current location on Twitter without having any idea they were doing it. Venmo payments default to public, again something that people often don't know.
And often it's not you who leaks the information but someone you know. Good luck controlling the social media presence of everyone in your family and circle of friends.
"Facebook demanded driving licence, then posted data from it": https://www.reddit.com/r/LegalAdviceUK/comments/9n8ulu/faceb...
> A couple of days ago Facebook suspended my account as apparently someone had reported me as having a fake name on there. It advised I needed to upload a copy of my ID to get the account unlocked. I sent them a copy of my driving licence to prove my name was real. Today Facebook has unlocked my account but has added both my middle names onto my facebook account so it displays my full name to everyone. If I try to change it back on my profile page it says that I need to wait at least 60 days before updating my name again.
> I do not want my middle names on Facebook, I consider that a security risk as it is one of the questions my bank uses for security. I did not give Facebook permission to copy that data from my licence and certainly not permission to then display it to others.
> Do I have any recourse or way of getting them to change the name back? I can't see any way to contact a human as all support tickets have just been closed.
And even then, pandora's box has already been opened. Your PII is available even if you haven't explicitly chosen to share it [2].
1: https://www.dailydot.com/unclick/4chan-shia-labeouf-secret-l...
2: Equifax.
Don't normalize harassment.
This is just the sort of comment that some might view as an invitation, or as you challenging them. And on a place like HN, you probably have quite a few people and groups around who would be well qualified to "make a point" so to speak.
It's one thing to be aware of the potential risks, but it is another for a 13 year old on Instagram not even realizing there are consequences.
Presumably, the people who are most unhappy about Instagram harassment are people who use the social network to make connections and friendships that they can’t make otherwise — there is no line between their social network relationships and “real life” relationships, especially as the former so often becomes the latter. That Instagram is so massively popular, and even an engine for economic activity, is a sign that it’s not something that people can simply log off from. Imagine telling a developer to just “log off” and get a “real” job if Microsoft were to kill every Github repo that didn’t agree to abide by new draconian licensing and terms of service.