If you’ve been following the world of Twitter (cue chorus of people who insist they’d never ever use stupid ol’ Twitter), you know that there’s been one of those epic corporate missteps: The CEO of Twitter, Dick Costolo, volunteered to answer questions sent to…a hashtag on Twitter, #askcostolo. What followed was amusingly voluminous, as the angry horde of Twitter users rose up as one to flood the hashtag with their anger at the incompetence of the company in providing tools to deal with abusers. As currently configured, Twitter makes it easy to sign up willy-nilly for anonymous accounts and for users to make racist and misogynistic threats without fear of punishment, or to create accounts solely for the purpose of stalking people online…and there are thousands of Dennis Markuzes out there, fulminating obsessively and hiding behind a flurry of throw-away accounts, all happily enabled by the mostly male engineers at Twitter.
And that’s the real problem, that Twitter won’t even take the basic steps of making it possible for users to block repeated harassers. It’s not clear why; it is clear that they don’t seem to think rape threats are actionable issues.
Anyway, one person made some simple suggestions for the least Twitter could do.
Block all users whose accounts are less than 30 days old
This is easy—it takes an arrow out of the quiver of serial harassers who use alternate accounts generated as needed.
Block all users whose follow counts are less than whatever threshold users set
Google used the social proof of “back links” to establish credibility and ranking for content over 16 years ago. This is old hat by now. Users should be able to block anyone who can’t convince other people to follow them.
Rings of followers created just to subvert this will have to be detected.
Again, hire a Google engineer. They’ve cracked this one.
Block new users whose @replies include any words the user decides
Users who are on the receiving end of harassment face startlingly unimaginative adversaries. The same slurs and threats are used over and over. Brand new account with no followers using the n-word? Block!
That’s stupidly easy to express algorithmically.
Block any user who has been blocked by more than N people I’m following
Let’s also share the load. If all your friends block someone there’s a decent chance you’ll want to also.
Auto-blocks are opaque
There should be no feedback when a behavior triggers these measures. The harasser should believe that everything is working as normal.
The first two suggestions are a bit problematic in that there are easy ways to get around them: the obsessed stalkers will just build up a small population of sock puppet accounts that will also link merrily to each other. It’s a bit of work and requires long term planning, but really…that’s what these loons do. They also put up barriers to brand new users.
But the others…wow. Do you realize that Twitter lacks even the most basic functionality of enabling keyword filters? We’re talking 1990s technology here, and the brilliant minds at Twitter corporate are unable to grasp the concept. They also violate basic security concerns: if you are being harassed, and you block an account, the owner knows right away to switch to one of his backup sock puppets.
One example: If you follow @AngryBlackLady, who writes about reproductive justice and racism, you know that she has been harrassed nonstop for two years by a goon who makes new twitter accounts just about every day, specifically to dun her with racist slurs and misogynistic noise:
Some of the smartest ppl in tech work at twitter, yet they won't/can't fix the abuse problem. #askcostolo #assholster pic.twitter.com/mwePtNBSQ7
— Imani ABL (@AngryBlackLady) August 3, 2014
This is basic. This is easy. Any idiot could provide functionality that would reduce this problem. Twitter, apparently, only employs exceptional idiots.
“Twitter, apparently, only employs exceptional idiots.”
Or, more likely, there hands are tied by a management with another agenda…. such as keeping costs low. Costs money to do the right thing… they’d likely have to hire more programmers and engineers. From the #askcostolo debacle, it is pretty clear to me he had no intention of answering those questions…. for a reason…. and nothing that anything has changed nor is that his intention. I doubt that rich white dude has encountered harrassment in his entire life; just a narcissistic prick.
One of the biggest problems that I’ve run into is finding graphic violence or terrible harrassment / bullying / racism/sexism etc. in tweets that I’d like to report. Even though the reporting system will allow it, and has a box to tick for it, in the end the message is, unless the harassment is directed at the one doing the reporting (me)…. forgetaboutit. The report goes nowhere. All that work – copying the url for multiple tweets, listing them in their goddamn boxes, to be told that it will all be ignored afterall, is infuriating.
I’ve been successful once with one picture that was so heinous that law enforcement agencies should have been notified. Which begs the question. Are they doing that? LOL!!!!. I just cracked myself up.
I’ve never used twitter. Now I have a good reason not to besides just to lazy to learn one more internet thing.
The mistake here is thinking that Twitter exists to serve its users… Well, OK, it does, but only in the “it’s a cookbook!” sense. They will not do anything that puts any roadblocks at all in the path of new users, because they need to keep expanding the user base, nor are they going to give the least stuff about whether some users are harassing other users unless they can be convinced that the number of users they lose as a result of allowing harassment is greater than the number of users they would lose by preventing it.
They no more care about the experience of individual users than a blue whales cares about the experience of individual krill. You’re just a number to them. Hell, you’re not even a number, you’re just a component of a statistic.
If you want to know what Twitter cares about, their latest earnings report is here. Twitter’s #1 job is to maintain growth in their number of Monthly Active Users.
Even if those numbers are plumped up by encouraging fakes and phonies to sign up for multiple accounts, like the “assholster” creep harassing @AngryBlackLady. Or if they are commercial spam accounts, like the parasites clinging to every user with a large traffic base.
Those count in the statistics, although they shouldn’t, just as much as legitimate users.
I’m a former Twitter employee — and the internal issues are both more complex and simpler than you’d think.
As small companies grow larger, they tend to become calcified. Few of them remain capable of innovating while, at the same time, maintaining the quality of their product. I’d argue Twitter is similar. If you recall, three-four years ago, the complaint you’d hear about Twitter had been the frequent outages and the fail whale. So, engineering has been focused on improving reliability of the product. Internally, this has been a complete rewriting of the underlying architecture of the system while causing minimum disruption to users. This, they’ve done well. But, in the process, they’ve not innovated on the product itself. That’s resulted in a stale interface, poor user management tools, and poor growth in the user community.
On top of that has been a laudable commitment to respecting the user’s voice (and I’d argue the converse has been to disrespect the user’s ear at times.) Twitter has been one of the better companies at fighting against censorship. Don’t ask me, ask the EFF. This is an internal bias against the kind of editing tools you mention. Perhaps that will change with time.
cacondor
Perhaps once they and others finally understand that choosing to not listen is not the same as censorship they will actually start caring about people dealing with online harassment.
Indeed.
Twitter: Because random commentary by random people is so very valuable.
Surprising that a viable alternative hasn’t arisen. Even Facebork has competition – however inadequate. Is market-inertia really that strong? Betamax all over again?
1. It might change when their user numbers drop off and their income drops. The internet is full of companies that arced up and then crashed. Anyone remember Myspace? AOL? Usenet?
Facebook ate Myspace’s lunch.
2. This is called free market capitalism. If you aren’t focused on your customers, they can and will go elsewhere.
3. I’ve heard enough horror stories about Twitter than I used it a few times and dropped it years ago. Didn’t miss it and won’t ever go back. I can get death threats and stalkers elsewhere on the net, and frequently do so.
This may serve as a wake-up call to give users what they want.
I’m surprised this is being presented as a “fail” when it seems to me like they’re getting really valuable feedback.
Failing to take PZ’s hint in the OP, the chorus of superior people who want everyone to know they don’t use Twitter is right on cue.
In the context of a conversation about harassment on Twitter, they are not directly saying “So she was harassed? Well what was she doing there anyway?”, but the implication, intended or not, is there.
This is a discussion about a social problem. Please don’t derail it.
Said a person in a comment section of a blog. No, really.
Actually…filtering words *isn’t* 1990s tech. It’s 1960s tech, at the very least. The UCBerkeley FORTRAN compiler would quit and give a fatal error if a program contained certain 4 letter words as variable names in the late 1960s. The ones the compiler rejected would be considered pretty mild by current standards, but the technique certainly existed at the time.
As for Twitter and related “services”…if the “service” is free, you are not the customer. You are what is sold to the customer.
(I won’t say that I don’t use Twitter because I do. I use it to check on downtime status of an MMO I play. I don’t have an account, though, and these tales of harassment on Twitter provide just one more deplorable reason for me to avoid having a Twitter account.)
From Twitter’s instructions on how to block someone:
This, to me, seems wrong. There should be no way for someone to know they’ve been blocked. They should just be left thinking they’re being ignored. Thus they’re less likely to realise they “need” to set up a sockpuppet account.
That’s how blocking works in most chat rooms, and on instant messaging systems, and it’s how a comments-board killfile works, so it’s hardly a revolutionary idea,
chorus?
Well, does it matter if I point out that the reason I don’t use Twitter is because too many of my friends and fellow feminist writers have written extensively about the constant harrassment and abuse?
I could see it having some value, and there are times I rather wish I found it useful, but even my very limited attempt at it was quickly overwhelmed by the amount of sheer abuse directed at almost everyone I wanted to read. It became rapidly way too fatiguing to sort through the sturgeonsworth of shit for the rare gems, and constant maintenance to block jackasses and spammers.
Or is that too hipster too?
Looking at it from a purely technical standpoint, would it not be possible to implement those safeguards in a twitter client? Its ridiculous that twitter won’t take some simple steps to protect its users from abuse, but it might be easier to create the FreeThought Twitter Client (TM) then to get a massive corporation to change its ways.
Fun fact: preventing abuse and harassment is not censorship.
Any form of filtering, for privacy or against harassment, can also function against unwanted ad messages or corporate bots spamming discussions. We constantly forget that in social media, ‘users’ are not the customer, ad agencies or brand marketing is. ‘users’ are the product, the eyeballs that the actually customers, those who run Twitter via paying its bills, want. Twitter isn’t stupid and incapable of writing code to provide some protection, they don’t want to and can’t because that opposes their business model. Plus a little controversy also always drives up any count metric and then they can charge their customers more.
The interests of users and the interests of Twitter are inherently in opposition.
Here’s some workarounds for the lack of filtering functionality on Twitter:
Twitter Better: 20 Ways to Filter Your Tweets
Open Tweet Filter (Chrome Extension)
How To Silence & Filter Annoying People On Twitter Without Unfollowing
Here’s How to Filter Anything from Twitter’s Web Interface
You rang? I still don’t see the appeal nor utility. :-)
At any rate, dailydouq’s summary at #19 is a pretty good explanation of why a capitalist company governed by owners whose sole goal is profit will never act in the interest of the social good in a case like this.
I think you’re underrating the “30 days” rule. It wouldn’t stop the hard-core crowd who will wait for thirty days and build phony twitter account upon twitter account, but it would stop a lot of assholes from jumping on to twitter under pseudonymous accounts on short notice as part of organized campaigns. In 30 days many of them might have gotten bored and moved on already.
They’re all good, though. I suspect Twitter doesn’t bother because they love overall high volumes of membership and tweets, which they can use as proof of their importance with advertisers (plus what dailydouq said in post #19).
“FreeThought Twitter Client”, @captaindecker ?
Nope, Twitter has been limiting its features for non-Twitter client for several years. They want to control the ads you see, just like the TV broadcasters who fought the VCR.
One thing Twitter does do with speed and enthusiasm is shut down spam accounts and accounts that fake the names of celebrities, companies, and other brand names. They can and will censor and shut down accounts but only the accounts that do something they think is bad. Rape threats? No problem! Claiming to be the real Justin Bieber? Gone in a flash.
The second suggestion is terrible. I don’t tweet and I’m not looking for followers. I have one follower who I suspect is just someone looking for a follow-back to up their numbers. I simply use twitter to follow people. I can’t be alone in this use of twitter.
Everything on that list, yes, with the exception of #2. For the same exact reason as mijobagi @ #24. I’m never going to have followers either and I’m not likely to be tweeting anything ever. If I had to have followers to maintain an account, I wouldn’t be able to. There is also the small fact that it would be trivial for dedicated trolls (and what trolls aren’t dedicated) to get around a follow-back threshold.
It’s not incidental, I’ll add, that one of the reasons I don’t tweet or re-tweet anything is because I don’t want flack from trolls, harassers and abusers, but that doesn’t even need to be one of the reasons even if merely using twitter as an information feed from people you find interesting seems rather counter to the social aspect of it. Some of us aren’t going to be social in quite that way, but still benefit from being included all the same.
Still, I’m considering leaving Twitter altogether if things don’t change soon. It’s really not worth it to me to generate revenue for them when people are being harassed without recourse.
Remember Reddit? ‘Can’t use it for many of the same reasons I’m willing to dump Twitter.
@mijobagi @Thomathy, Such A ‘Mo
These suggestions are for security settings that anyone *could* add to their account. Someone who gets a lot of rape threats could switch on all of them, while you might choose to leave your account more open.
So the second item is an excellent one for *some* people.
This is a very minor objection, I think, because I’d be more concerned about false positives than false negatives. This reflects that I don’t get harassed, and probably also how I use Twitter, but I’d rather an (opt-in) auto-blocking algorithm that spares me 50% of accounts I have to block than one that takes care of 120%, sweeping up people I’d like or find interesting as well.
Dunc @ 3:
Sure, if I’m not paying I’m the product, not the customer, but I’m also a product that can hop off the shelves if my needs aren’t being taken into account.
vaiyt @ 18:
I can see how it can look like it to those of us who don’t get abused or harassed. I include myself in that not because I don’t think these tools should be implemented — I do — but because the extent of the problem is something I know as information, rather than experience.