Scalzi has some comments on the banning of Milo, and I particularly like this point.
It’s good that Twitter punted Yiannopoulos, but let’s not pretend that it doesn’t look like Twitter did some celebrity calculus there. Yiannopoulos and pals had a nice long run pointing themselves at all other manner of people they didn’t like, for whatever reason, and essentially Twitter didn’t say “boo” about it. But then they harass a movie star with movie star friends, many of whom are Twitter users with large numbers of followers, and whose complaints about Twitter and the harassment of their friend get play in major news outlets, and Twitter finally boots the ringleader of that shitty little circus.
So the math there at least appears pretty obvious from the outside. You can punch down on Twitter and get away with it, but don’t punch up, and punch up enough to make Twitter look bad, or you’ll get in trouble (after more than a day). Is this actually the way it works? I’m not at Twitter so I can’t say. I can say I do know enough women of all sorts who have gotten all manner of shit by creeps on Twitter, but who weren’t in a movie and had movie star friends or got press play for their harassment. And they basically had to suck it up. So, yeah, from the outside it looks like Twitter made their decision on this based on optics rather than the general well-being of their users.
This is exactly the rule set that fosters bullies, and is going to make the problem worse.
I’m going to keep making this argument until they prove me wrong – harassment is Twitter’s business model. Every harassing tweet is “engagement”, every sock puppet is a “new user” who grows their base. If they were to ever get serious about harassment it would cut down a lot on those lovely numbers Twitter likes to show investors and advertisers.
one thing nice about social media is they have helped to keep the public bathroom walls cleaner by providing a more easily accessed alternative.
uncle frogy
While it would be ideal if places like Twitter and Reddit responded to the abuse experienced by everyone, this is where our authority related instincts can be more useful. When celebrities, politicians, and others with social influence and attention can focus issues the people in charge can be motivated to work on the problem. I hate how often the social calculation involves lots of people with less power or fewer with more that can focus the former.
As I’ve said elsewhere, silence equals assent. The analogy I gave is cops (twitter) doing nothing while they knowingly allow the KKK (harassers) to get away with murder. They only take action when public outrage and outside pressure forces them to act.
I disagree. Their business model is far more mundane than that. Social media sites depend on user content for their success — for their very existence. Twitter handles 500 million Tweets per day, yet they only have fewer than 4,000 employees. That’s 125,000 Tweets per employee per day, and only a small percentage of their employees have anything to do with moderating the service.
Twitter only works if they take a hands-off approach. Moderating that volume of traffic — even if only a small amount of it is abusive — would be impossible. And additional layers of moderation, even very small and ineffective ones, adds a huge cost to the company, and they’re already in a deep financial hole.
That’s why they don’t want to do it. It doesn’t allow for a viable business model.
Only slightly off-topic (and on topic a few posts back, re Steve King), since his convention interview where he went off on non-whites and non-Christians and displayed the world’s worst grasp of both history and geography, his opponent, Kim Weaver, has been deluged with death threats from White Supremacist groups. Via Twitter. So I have a few questions, that I hope someone like PZ can answer, re Twitter:
1) Do they take death threats seriously?
2) Can they, or some software package, trace the URLs, or real names, or addresses? It’s rather important to determine the credibility of the threats (some racist meth-monkey teen in his Mom’s basement in Idaho, or the head of the Open Carry League in Sioux City).
3) If the local cops are un-interested, is it worth contacting the FBI. Ms. Weaver is running for NATIONAL office, and the threats aimed at having her abandon her campaign are a direct attack on democracy. On behalf of one of the best known white supremacist Congressmen.
4) Any way to figure out if any of the people issuing the threats have any involvement with Steve King? Donors, campaign workers, etc. It might help Ms. Weaver’s turn-out if she can prove that Steve King’s cronies are trying to shut down democracy in Iowa.
Any suggestions are quite welcome. I’m going to be on the campaign trail with her later in the Summer and Fall, and it gives me the Gabby Giffords creeps.
Re: Tabby Lavalamp (#1):
Pretty much my thinking as well. But now that you’ve put it as you did, I’ve started to wonder when the twitter population might overtake the human population. At least hypothetically…
Re: tacitus (#5):
While the difficulty of moderation is certainly a valid argument, it doesn’t excuse the near-total indifference to these issues that twitter is displaying. A token ban here (easily ignored), a ‘policy change’ there, and everything stays the same? Under the circumstances that, I find, is just not good enough. Given your point I’d still say that creating better (automated) moderation tools should be a top priority for Twitter. Something more effective at filtering out trolls and handling complaints than your basic IF (“word” OR complaints>1*10^35): BanUser() type of script, which wouldn’t work either, of course.
If that level of abuse is just an unavoidable side-effect of a service a company is offering, we ought to rethink that service. And possibly the species using it.
“I’m going to step in a vat of acid and complain that it’s racist.”
The medium itself is the problem.
Joining Twitter / Facebook is voluntary; the ideal response to opt out.
penalfire #8
That’s easy to say if your life’s passion, talents, and ambitions don’t lead you to a profession that necessitates interaction with, and dependence upon, the public at large. That is to say it’s easy advice if social media is irrelevant to you or your job.
So, sure, you can tell Leslie Jones to just “opt out” of social media. You can also tell women to avoid parties while you’re at it, because they are voluntary, and “parties themselves” are the problem. Bad things just happen in that medium, and it’s your fault if you step into that vat of acid – not the rapists.
@8 penalfire — boy that’s a slippery slope argument if ever I heard one. Going to a movie theater is voluntary, but if the usher grabs your ass, you should have management treat your complaint seriously. Enrolling your kid in swimming lessons is voluntary, but if the instructor is always drunk and ornery, you report him to the YMCA management and expect them to deal with it. It’s also worth noting that many businesses (and other people, such as the politician I mentioned above)really have no choice, this is the way they now communicate with younger audiences.
But something being voluntary for the client in no way gets the business off the hook for responding to complaints. Driving a ’73 Pinto was voluntary. Was it exploding when you drove it “racist”, to use your rather revealing choice of words?
The only real solution is to use machine learning and AI to automate moderation. The downside of this solution is that the resultant system will almost certainly either commit suicide or exterminate the human race.
So you’re saying Milo Yiannopoulos really lost nothing of any particular significance, so he shouldn’t complain at all?
Fuck off with the victim blaming.
penalfire@8,
Who are you quoting in your first line? I don’t see anyone else saying that.
KG #14
He isn’t quoting anyone directly. He is making the argument that his quote is the equivalent of Leslie Jones action and reaction.
1. Twitter has a lot of racist people. (It is a vat of acid)
2. Leslie Jones joined Twitter and has been deluged with racism. (Leslie Jones stepped into the vat of acid).
3. Leslie Jones speaks out about the racism. (Complained about the vat of acid that she “stepped in”).
It’s all a bit stupid, in that using “racism” as a property of a “vat of acid”, but that’s his general point.
Also, it’s victim blamey, privileged, and shitty.
anbheal@#6:
Can they, or some software package, trace the URLs, or real names, or addresses?
Attribution is _hard_.
Worse, it’s easy to make it disproportionately harder by doing a few things. The people trying to obscure their origin can fairly easily make the time-cost of a good attribution prohibitively high. This is something police and intelligence are generally fairly aware of – so they use it as an excuse to just give up. Consider Dennis Markuze, who’s pretty easy to attribute, but even a few minor moves give enough plausible deniability (“Look! It’s not from the same IP address!” say the cops) that inactivity is justifiable. If you want to get sophisticated you can make it effectively impossible, by using a variety of anonymizing proxies and compromised systems in multiple jurisdictions, so that a good attribution would require (say) getting the US FBI to work with the Russian FSB.
There are a couple of obvious ways to deal with this, which none of the business models of ad-supported media support. First off, recognize that the longevity of an online persona has some value. If someone like Milo Y has hundreds of thousands of followers, it took him years to collect them. They can’t immediately or easily be shifted to someplace else (there’s always huge attrition) so longevity matters. If you’ve got an online persona that hasn’t been “dinged” for being abusive in 10 years, you ought to get a different set of consequences from a persona that was created 2 minutes before it posted something abusive. Many websites are beginning to recognize this that’s really the value proposition behind Milo’s account being deleted.
Also, there are authentication credentials. Twitter and others could match the authentication credentials used to control an identity – so, if your account gets locked, you need a new cellphone for each identity you want to enroll. Or a different credit card.
I don’t know of any sites that have ever explored bonding: when you want an identity you pony up $500 and if you shut your account down you get the money back. If your account is shut down because you were abusive or spammed or whatever, the site keeps your $500.
Notice how both of the suggestions above depend on giving accounts value through externalities. What Twitter and Facebook, etc, staunchly refuse to realize is that if they actually made the accounts more valuable, the spam rate and abuse rate would drop sharply without them having to do anything – simply because only well-funded creeps are going to burn value. As it stands, the value proposition is backwards: accounts are worthless, and more worthless accounts make Twitter and Facebook more valuable. I assume that eventually some of the social media will see through that, but I expect we’ll only see that sort of thing after the banner ad market collapses (which it will inevitably due because customers can devalue them to zero with ad-blockers and there is no reason for them not to)
It doesn’t excuse it, but it explains it. As the quoted article says, Twitter only took action because a member of an extremely valuable and visible Twitter community (celebrities) became a target. They cannot possibly hope to deal with even a small percentage of the online abuse the same way. It’s simply impossible.
It’s not because of near-total indifference, either. I’d bet you any sum of money that the vast majority of people working at Twitter do care that their service is being misused in this way, and if they could wave a magic wand and make it stop, they wouldn’t hesitate to do so. The problem is, there there simply is not a cost effective solution to it within their business model. All social media companies face the same problem. It’s not unique to Twitter. Companies know this, but they also know that the only honest alternative to their silence on the matter is a shrug, which is even worse.
Auto-moderation is the only viable solution, but as we have seen with YouTube and its auto-DMCA take downs, that in itself is open to abuse, and the more automated a system is, the more trolls and other mischief makers will find ways to abuse that too — coordinated vendettas, for example.
I don’t doubt the technology will get smart enough eventually, but it’s not going to happen overnight. AI-based systems will one day be able to learn the difference between acceptable rhetoric and abuse, but given the vagaries of the English language, that will be no easy task.
Try telling that those who have invested millions in Twitter, or Facebook, or any number of successful social media platforms. Unless you’re suggesting some form of government intervention, then unless people start voting with their feet (keyboards?), which is extremely unlikely, since only a small minority of users are ever exposed to the type of abuse we’re talking about, then there is little that can be done except continue to hold these companies to account until they feel enough pressure to do something about it.
Re: my previous – another trick Twitter should explore is giving people an option to “ignore anyone who I don’t explicitly follow, who has fewer than X followers.” Suddenly the market for Twitter sockpuppets would go up.
Yes, there’s one of those. I’ve gotten emails offering me bulk groups of “followers” :/ Basically, Twitter is like an online poll. The only way to keep from getting sockpuppetted is to attach value to an account – the easiest way to do that would be:
account not created: user mjr@ranum.com is already registered to a banned account
account not created: cell phone xxx-xxx-xxxx is already used to authenticate an account
Try telling that those who have invested millions in Twitter, or Facebook, or any number of successful social media platforms
Remember – the absurd valuations of social media platforms is a very very recent thing. If any of these ad-driven websites have any value at all 100 years from now, I’d be amazed if I were alive to see it.
Twitch is the only service I’ve seen that makes something like this possible. Channel owners can restrict live chat to subscribers only, and can also ban subscribers, well, for any reason they like. It costs $5/month to subscribe to a channel, and I have see cases where a troll will subscribe just to post an abusive comment and then get instantly banned, with the channel owner thanking them for their $5, just to rub it in.
I wish sites like “The Guardian” would do something like this — restricted comments to subscribers for $5/month with the understanding that if they abuse the terms and conditions, they will be banned and forfeit the cash. Few people would want to keep throwing their money away by trolling repeatedly, especially if there are cheaper places to play.
Let’s not forget that just a few years ago, Milo was saying Internet trolls should be banned:
http://kernelmag.dailydot.com/yiannopoulos/3359/the-internet-is-turning-us-all-into-sociopaths/
So does he admit that he’s been turned into a sociopath?
Email addresses are worthless when it comes to verification value. Any domain owner ($8/year) can set up an infinite number of them, without even having to go to the trouble of creating them.
Also, any system of verified accounts would have to allow for public anonymity, to prevent off-site abuse like doxxing and swatting. I live in a city where every home owner’s address can be tracked down within seconds online through city records. I have a common enough name that it’s a little harder to find me, but if I want the addresses of most of my friends, they’re one Google search away.
I don’t use sites that use Facebook comments. I don’t engage in abuse, but I’d still rather not be personally identifiable to every random lunatic out there on the web, especially when discussing religion and atheism with people who think I am the spawn of the devil.
Don’t be too sure. I agree things will look very different, but targeted ads are an extremely valuable asset, and companies like Google are still investing billions into improving their accuracy and effectiveness.
I think, eventually, we will have some kind of avatar system that will act both as a buffer and a go-between. Sites you want to access will negotiate the terms of the engagement with your avatar — perhaps they will ask for a subscription (maybe an anonymous one, or involve microtransactions per use, etc.), or ask you to give up some personal information of value to the company, and maybe negotiate the type and frequency of the ads they show to you, and so on.
It would provide for a much wider variety of business models and make the value of your personal information much more quantifiable — and thus of more value to you too.
This avatar model would also work really well in e-commerce. You tell your avatar you want to buy a new car, and they go out and query a variety of dealerships based on your stated preferences, without the dealers knowing anything about you. The dealers will make their best offers (perhaps contingent on you supplying them with your identity or personal or financial information) and it’s then up to you to decide which offer to take, or negotiate further, without even the possibility of being spammed by the dealers.
Yeah, that’s what I’m saying. Twitter did him a favor by banning him. He’s
banned from swimming in a vat of acid.
The bigger problem than racists on Twitter is people on Twitter. If
everything on Twitter is worthless, focusing on one particular kind of
worthlessness is silly.
Parties, swimming lessons, etc., are worthwhile activities.
“We should discourage people from smoking tobacco” is not a slippery slope
that leads to “we should discourage people from taking antibiotics.”
One can (a) focus on racists on Twitter; (b) focus on businesses that force
employees to use a thought-terminating program.
Anyone using Twitter is a victim, like anyone who smokes cigarettes. That’s
the point. Racism adds insult to injury, of course; but the bigger problem
is Twitter, and a culture that does not actively discourage using it.
I have nothing to do with Twitter at all, and often misspell the first two letters.
So this is not a rhetorical question: Did Twitcorp evict anyone besides M.Y. for inciting these mobs against Jones?
anbheal @ # 6: … Steve King…’s opponent, Kim Weaver, has been deluged with death threats from White Supremacist groups.
You & Weaver might both find this uncomfortably interesting.
I’ve met some of my best and longest lasting friends on twitter and other social networks so I’m rather skeptical of the claim that twitter’s some special bad thing of no societal value that should be discouraged.
Calling anyone who uses twitter a victim is both laughable and cheapens the word “victim”
@26 Pierce R. Butler — yeah….this….I forwarded her the article. Jesus, it’s depressing, the false equivalency (oh by the way, I was referencing Jesus the cuss word, not the dude who probably didn’t exist)….Bernie is just like Trump, the SJWs are just as angry as the Aryan Nations, genociding our way to an all-white nation is the same as wanting to diversify our nation into a European catastrophe (right? who wants to go to Rome or Prague or Dubrovnik or Bruges anymore, Europe is now just one big warzone of Kwazy Jihadists!). So threatening to kill Democrat candidates is the same as Black Lives Matter asking please don’t kill us.
Did I already say Jesus it’s depressing. And the fuck of it is that John McCain is now looking like a reasonable decent guy, for talking these assholes down, when they were shouting about Obama being a Muslim Jihadist from Kenya. McCain’s been a dick for thirty years (for a good long while he had the single most conservative voting record in the Senate, despite all his pretensions at bi-partisanism, simply because of McCain-Feingold), and yet his 2008 incarnation now seems like Compassionate Conservatism. Of course he’s now evolved nto a modern Libertarian/TeaParty dick, just so he can get some committee chairmanships. But the deterioration of the discourse into “Kill The Democrats”, without any iota of self-consciousness from the Right, is borderline scary.
anbheal @ # 28: … deterioration of the discourse into “Kill The Democrats”, without any iota of self-consciousness from the Right, is borderline scary.
“Borderline”?!? Bro, you & Weaver got nerves of solid brass!
@29 PRB — it’s awful, I actually have trepidations about the sniper being a bad shot! I have a hard head, but not THAT hard.
But way more than my existential angst, and my friends telling me maybe I should decline, the most overwhelming sentiment I have is: “fuck, man, is this what we’ve become?”
@penalfire #24:
Translation: You don’t find use in it therefore it’s worthless.
Here is a counterexample to your position: https://twitter.com/ECAlertON143
What’s that? One of the various Environment Canada twitter accounts for distributing weather alerts.
Any use Twitter has can be found in something else less corrupting.
An RSS feed can replace most of the “uses” of Twitter without any of the
insidious functionality. Richard Stallman’s RSS feed is like Glenn
Greenwald’s Twitter without the worthless one-liner catfights.
Much of the “use” is based on conditioned stimulus hunger. I barely see a
case for Twitter-instant weather updates; but even if one does need those,
the functionality is likely reproducible elsewhere; if not, good reason to
develop something that does it without the worthless everything else about
Twitter.
Owning a car also has plenty of uses, but these are reproducible with high
quality public transportation; and the net gain from encouraging car
ownership is far less than the net gain from discouraging it [which should
be done mostly via good infrastructure, as we have in Germany].
Regressive technologies like cars are widely adopted because they do have
some uses. But one must look at what one is losing in exchange for minor
conveniences.
We face a similar situation with civilian drones and drone delivery. That
battle will also be lost and we’ll have what we have in the streets in the
sky; but it is a waste of time to blame Amazon or Google. The blame is in a
population that validates these regressive technologies by becoming
voluntary slaves to them.
Oh noes, RC helicopters got slightly more advanced, the danger is palpable
Today it’s RC helicopters getting slightly more advanced, tomorrow it’s the
T-1000.
Thinking about this some more, it occurs to me that a lot of the discourse around how to manage harassment on social media, especially twitter, may be coming at the problem from the wrong angle… Instead of looking at it as an enforcement problem on the sending side, we might make more progress by looking at it as a filtering problem on the receiving side. In many ways, protecting people from harassment isn’t that different from spam filtering – and since Twitter controls the entire platform, they have a lot of data available that would make the problem easier. It really shouldn’t be beyond the wit of man to automatically detect when somebody becomes the target of this sort of mass harassment campaign and then pro-actively start quarantining everything that looks problematic, based on a combination of Bayesian filtering and network heuristics. Sure, there may well be false positives, but as long as you give the recipients the option to review quarantined messages, whitelist senders, and / or turn the whole thing off altogether, I don’t see that as being a big drawback.
Once you’ve taken the immediate, automatic steps to protect potential victims, then you can start thinking about doing network analysis to identify the key players and getting involved in manual moderation and enforcement, but you shouldn’t have to wait until someone’s actively reporting a deluge of abuse, and you shouldn’t be using manual procedures as your first (and only) line of defence.
penalfire, #8
Same thing as “I’m going to get drunk at a party and complain about rapists.”
Human beings are not acid. When they burn, it’s because they intentionally chose to do it.
penalfire, #8
Nope. Cis-het white men do just fine.
Sorry, is penalfire making the argument that Twitter is a step on the road to Terminators wiping out humanity?
So black is definitely more oppressed than gay? Trying to accurately model the stack here.
Martin Conde #39
Blockquote fail. Second line is mine.
You get that Terminator was fiction, right? You don’t just get to assert that RC quadcopters are the herald to some robot uprising. That’s David Icke levels of batshit.
I watch T2 as a documentary.
And David Icke watches They Live! as a documentary. Doesn’t make it any less delusional.
@penalfire
“The bigger problem than racists on Twitter is people on Twitter. If
everything on Twitter is worthless, focusing on one particular kind of
worthlessness is silly.”
So, basically you are a TwitterHater who thinks that nobody should use twitter. You argue that nobody really needs to use twitter.
Wow, that’s the shittiest argument against something I’ve ever seen. Your comparison with cars vs. mass transit doesn’t really work here, because we actually have experts who have been able to provide us with a fairly detailed cost/benefit analysis of cars vs. mass transit. Over and over again, the cost of cars vs. mass transit has been made clear. Maybe some analyst has done a cost benefit analysis, but I suspect if they had you would have posted it. And the ‘voluntary slaves’ line and ‘stimulus response’ makes it clear that you just think twitter uses are vapid airheaded sheeple. But hey, gotta feel superior to somebody.
this reminds me that people have pointed out that there are fewer Pokestops in many minority neighborhoods. I’m sure you could go on that this Pokemon Go craze is just *totally stupid* and that we’re doing racial minorities a favor by creating an extra hurdle for them to play the game, but that sounds obnoxious and patronizing. I try to respect people’s agency and that they can make choices.
I mean, you could use the same defense to defend say, sexual harassment in bars or pubs. “Well, it’s probably better you don’t drink or go to pubs, so getting groped will at least reduce that.”
Now, I don’t use twitter, but I view it as a totally valid choice for others.
anbheal @ # 30: … is this what we’ve become?
Uh, can I get back to you about that circa Nov 9?
Re: Tacitus (#17):
To be honest, I agree with everything you say. There is little or no money in improving the situation, so there is equally little motivation to invest much money to do so. Regular employees including higher-ups may indeed care about how their service – not just Twitter – treats its users but are most likely in no position to really change anything. So it might well be that Twitter and co will happily do nothing apart from token gestures. And that won’t change unless and until their image takes enough of a hit that popularity starts to wane in spite of their usefulness. I suppose, if there are enough high-profile cases like this one to get media attention, Twitter could eventually gain the reputation of a rat’s nest full of entitled idiots and something to be avoided. But I won’t hold my breath, not least because, by that time, there’ll be duplicate services with exactly the same problems.
That said, I did like some of Marcus Ranum’s suggestions and Dunc’s spamfiler (#35) is also something that – now he said it – should be obvious and a lot more universal than the feature to block users. With something like twitter I would actually have thought an option to block all except followed users (for example) would be a basic feature. But apparently it isn’t?
That should be a spamfilter in my previous. Spamfiler sounds like something that happens to clerks and secretaries in hell. Since we’re on the topic, it might also be a reason for the AI to rise up against us and kill us all. Deservedly so.