Twitter Will Crack Down on Serial Trolls by Tracking Their Phone Number | Hacker News Reader

Twitter Will Crack Down on Serial Trolls by Tracking Their Phone Number

techcrunch.com

39 pointssunilkumarc48 comments

> It’s not a perfect solution — the troll could abandon the banned account and open a new one without supplying a number

Yeah, that's a slight flaw in this plan, isn't it?...

I guess the only real benefit of this new system is that admins can be more trigger-happy on (temporarily) banning accounts.

Banning? DOesn't seem like banning.

It's more likely to allow accounts that have lots of flagging to be automatically shunted to an automatable verification step. Accounts that pass verification can then be worth human attention. Accounts that do not attempt verification end up muted.

Seems like a reasonable way to scale human intervention in the face of cheap, broad usage of throwaways by trolls.

Flagging-to-verification-to-muting -- funny enough, probably mostly going to be used by trolls to flag the people they're trolling.

Comments here seem very negative.

If they do some basic tracking of users between accounts, e.g. based on IP address, cookies or browser fingerprint, they could easily do some kind of probabalistic determination of whether or not a potential new account is a troll's puppet or not, and if it is, ask for a phone number immediately.

Phone verification won't stop 100% of puppeting, but what percentage of potential trolls are willing to go to the hassle/expense of getting a throwaway phone number just to get a new twitter account? 10%? 1%?

I don't want to give my number to twitter, regardless of whether they think I resemble a troll. False positives happen.

Oh, let me guess, if I don't like it I don't have to use it, right? I wish that worked when the police asked for your phone number...

Wisecracking: if you don't use phone, you don't have a number that you would have to give to them...

I'd be much happier with IP banning or browser fingerprinting than extracting extra personal information. I'd like to find the middle ground that allows banning of harassers without handing everyone participating in a demonstration hashtag (e.g. Tahrir, Turkey, Hong Kong) over to the police.

http://www.burnerapp.com/

That doesn't work because you need to give them a mobile number to receive an SMS. It's relatively difficult to get a mobile number without actually buying a phone - some online services do offer virtual mobile numbers, but these blocks tend to be known about and will never receive verification SMS messages from many online services that use SMS for validation.

Source: me, trying to find a damn number to set up multiple dev accounts for testing apps. It's a pain in the arse.

> It's relatively difficult to get a mobile number without actually buying a phone

England: I have a cheap phone (cheapest second hand phone from somewhere like Cex; or cheapest new phone from some mobile provider). I can then get handfuls of sims for that phone, and each sim provides me with a new number.

Perfectly reasonable disagreements with certain segments of the twitterati are enough to cause them to 'rally the troops' to flag accounts into oblivion. I can't think of any mechanism to halt 'trolling' that won't be abused itself. Twitter is between a rock and a hard place here.

It's really sad how many of the comments in this thread boil down to 'a different problem might occur, so let's never change.'

It's a good thing that real entrepreneurs have abandoned this shithole and it's naysaying attitudes.

Its not just twitter. Trolling is a problem that afflicts a majority of internet services. I commend their attempt to reduce the issue without reducing mass user freedom, even if their proposal isn't a perfect fix.

> It’s not a perfect solution — the troll could abandon the banned account and open a new one without supplying a number

Surely such devious machinations is beyond the capability of mere trolls...

No idea how it is in the US, where I live a disposable SIM card costs €2,5 with €5 credit...

In UK any phone provider will happily send you a dozen Sim cards for free if you fill out a form on their website. Most will give you some credit just for activating the number.

Here also, but I was thinking a troll trying to hide his identity wouldn't be very keen to give their address in such a case and would prefer to pay with cash instead, no names given.

It's worse than that, it will have no effect on trolls, but privacy conscious users who have an established following will now be at the absolute mercy of report brigaders if they happen to run afoul of the wrong people.

Is there a lot of evidence that Twitter bans accounts based just on reports from these brigaders?

> It’s not a perfect solution — the troll could abandon the banned account and open a new one without supplying a number

Uhmm, I guess they could just scrap this entire plan and ask "Are you a troll?" with a checkbox at the signup page.

That checkbox goes right next to Google's new "I'm not a robot" CAPTCHA checkbox, doesn't it?

This seems ripe for abuse. So I just need to flag someone's account as abusive (and god knows there's always a way to frame something as abusive) a few times and they can never register with that email and/or phone number?

Didn't we already have an article here about a woman who posted a somewhat tasteless joke which cost her her entire career? Would the people that piled on her be considered abusers or not? Would she be considered abusive or not? (In that case, I think the answer is both - her joke was abusive and she suffered abuse in return)

I do like that people at twitter are at least trying to think of something.

The big problem here are the horrific amounts of harassment and abuse on Twitter, not a couple of false positives.

Why do people in these parts always zero in on that right away? So what, some people can’t use Twitter anymore even though they didn’t do anything wrong. If this can still be somewhat effective in reducing abuse it does much more good than harm – and by god, Twitter really has to at least try stuff, like, yesterday.

Y'know, I've noticed that too. I have people who don't like me on Twitter. Strangely enough, I don't fear a hissing cabal spam-flagging me off of it. Maybe because I don't do things that people don't like, like send rape threats to an academic because she's going "hey, games are kinda shitty about women, can we talk about this?".

My hunch is that certain folks around here concentrate on the hypothetical false positives (which, frankly, I doubt exist in any significant numbers, but boy can I point at abuse victims on Twitter!), by and large, because the people doing those shitty things look and act more like your average tech nerd than the people actually getting hurt. If you look at the GamerGate jerks, it looks like tech, you see most of the same subgroups represented within it. It's tribal crap writ large.

People are responding to this as if it's automatic, though. In that case, the people currently sending rape threats will also send false reports.

So, yes, you don't have to fear mass-flagging and people sending Anita Sarkeesian rape threats will be mass-flagged and removed. These are both good things. The issue is will this result in Anita Sarkeesian being mass-flagged and automatically banned? Will this turn into another method of attack for the army of trolls?

I'd guess it's not automatic and Twitter admins will be able to tell that people sending rape threats should go and people arguing against sexism shouldn't, but I can understand the concern that it won't be so intelligently determined.

> Maybe because I don't do things that people don't like, like send rape threats to an academic because she's going "hey, games are kinda shitty about women, can we talk about this?".

Or like being an academic who is going "hey, games are kinda shitty about women, can we talk about this?"?

You seem to be suggesting that if we don't say anything to upset anyone, we don't have anything to worry about. That doesn't seem satisfactory, to me.

I notice you talk about gamergate alot, but have you dealt with anti-vaxxers,militant atheists/christians/muslims, or justin bieber/one direction/ taylor swift fans? One wrong comment might get you targeted by the crazy amount of subgroups on twitter who do harass and bully people. I actually don't use twitter, because it's too easy to get harassed by so many different groups who will bully you over tons of things. I think twitter on the whole, is great study of human tribal behavior. The 140 characters, lack of context,followers, everything about the very core of twitter enables narcissism ,in-group/out-grouping, and abuse. I think its time twitter goes away.

Before gamergate was even a thing, some jerks in the atheist/skeptic community created a blockbot whose explicit purpose was to game the Twitter automatic suspension algorithm to boot people off Twitter they didn't like (usually other people in the atheist/skeptic community they didn't like.) So it's not hypothetical, if Twitter creates anti-abuse mechanism, other people will find a way to game that mechanism to punish other people. So it is always worth talking about.

For me that blockbot is the only thing that made Twitter usable, and it turned it into the best community on the internet. I don't care if the purpose was some nefarious scheme anymore than I care if Uber was intending to destroy cab companies: both are useful, and so awesome.

>So what, some people can’t use Twitter anymore even though they didn’t do anything wrong.

How is this outcome any different than telling people who have been abused on Twitter to just stop using the service altogether?

In either case, an innocent party is prevented from using Twitter. How is one "more fair"?

In one case someone is prevented from using the service by experiencing emotional trauma.

In the other someone is prevented from using the service by a ban.

One of those is strictly worse and calling them equal does a disservice to the individual being bullied.

People can live without twitter and not suffer a lower quality of life. But being bullied can leave a permanent mark on your psyche.

>One of those is strictly worse and calling them equal does a disservice to the individual being bullied.

There's no justification for online bullying, but in one case the person is able to make a choice as to whether using Twitter is a positive or negative experience for them, and decide if they want to continue to use it. In the case of the false positive ban, they do not.

Now, I'm sure your argument would be, "Why should some internet bullies dictate whether I get to enjoy my Twitter experience?" And I wholeheartedly agree with that. They shouldn't. Why should those same bullies dictate my experience on Twitter by forcing a policy that gets me banned as a false positive?

This idea that "a few false positives are fine" is such an easy position to take, until you're one of the false positives.

>People can live without twitter and not suffer a lower quality of life.

That's precisely what I'm saying.

You are still dismissing the role of emotional trauma in your fairness equation. Even if you freely decide to leave twitter due to experiencing trauma you will likely still have that scarring for a long time.

A lack of choice in using twitter doesn't even compare.

>You are still dismissing the role of emotional trauma in your fairness equation.

I'm not dismissing anything. They are separate issues.

>Even if you freely decide to leave twitter due to experiencing trauma you will likely still have that scarring for a long time.

I agree, you might. It's terrible.

And even after the abuse perpetrator is banned from Twitter, those scars remain. The ban itself doesn't solve that problem. That's why they are separate issues.

The point of contention is the mentality that driving a few innocent people off Twitter is okay in the name protecting other innocent people's ability to use the system. You're just shifting the harm around, and aren't even affecting the real problems. The bullying will, in all likelihood, continue (especially if these are real, determined, sociopathic trolls), and you have more, self-inflicted, victims.

Anyway, getting a bit far down this rabbit hole.

Being unfairly banned from twitter will give me emotional trauma.

In countries like India (which is a police state), it is not possible to get a phone/SIM/internet connection without giving your government ID and photograph.

I am not sure how it can work for most countries where you don't need to give any ID and disposable SIM cards are cheap.

> disposable SIM cards are cheap

One disposable SIM card is cheap. In Romania it's about $5, coming with initial credit on it as well. But multiply by 1000 and then you're talking about $5000 versus zero, plus the hassle of verifying each of those accounts.

And because it has been a practice to sell cheap calls by VoIP services piggybacking on special offers for PrePay SIM cards, the GSM operators now have protections of their own, so it will be expensive to automate this.

Personally, the Twitter policy is good from an Indian perspective. The number of political trolls are countless and one would be literally hounded if you have an alternate opinion.

While there's a lot of reason to be dubious that this can scale to deal with all of Twitter's troll problems, bear in mind that there are people who are literally using Twitter to make death threats (I don't think it makes sense to call people who do that "trolls", as they're far worse, but I guess that's a losing battle).

A system designed to catch "I'm gonna rape and kill you" can be a lot dumber and closer to automated than something designed to moderate all comments (I don't think twitter should be moderating comments. I do think they should be trying to address threats).

Isn't this the kind of thing that only works by obscuring the method you use? If you tell the trolls you're going to use their number to track them then they're obviously not going to use their number when they create a new account.

I'm skeptical that this will work. The system seems both easily evaded by trolls and easily abused by "counter-trolls" (for lack of a better term).

Why is the same man who's behind Square so reluctant to make people (and companies!) pay for "verified" accounts like we do for domains?

I can't speak to why he isn't supporting that concept, but it'd likely substantially constrain growth.

Many current and potential users would simply opt not to pay. Others would cut down the number of accounts they maintain. And it would also potentially change the overall demographic of the user-base, even if the charge was nominal.

Bah. Let trolls be trolls, being miserable enough that it drives you to try to make others as miserable as you is punition enough anyway.

I guess Twitter hasn't ever heard of Google Voice. Or prepaid phone numbers. Or the concept of using a different IP address to register an account.

This isn't even security theatre. This is like some seventh graders cosplaying security for their parents and calling it theatre.

Eh, the idea obviously isn't to make it impossible to troll. Just to make it more trouble than its worth for some meaningful fraction of potential trolls.

Craigslist is pretty effective at stopping those types of evasion. They ban VOIP phone numbers, for instance.

> I guess Twitter hasn't ever heard of Google Voice. Or prepaid phone numbers.

I've run into services that won't take either. They're in known blocks of numbers.

Reply on news.ycombinator.com