Manchester & Social Media

[stollcpa]

I've seen reports that there were a couple Tweets that might have warned authorities. What responsibility do you think Twitter, Instagram, FB and other social media have to alert law enforcement? I think they have responsibility to alert law enforcement on any posting that even hint of a person propensity to commit any crime.

 

[TheOriginalHappyGoat]

I've seen reports that there were a couple Tweets that might have warned authorities. What responsibility do you think Twitter, Instagram, FB and other social media have to alert law enforcement? I think they have responsibility to alert law enforcement on any posting that even hint of a person propensity to commit any crime.

How?

 

[stollcpa]

How?

Don't they have filters to determine when to take down posts? When they're taking down posts related to hate speech, potential crimes, etc., why can't they be compelled to forward them to law enforcement?

 

[TheOriginalHappyGoat]

Don't they have filters to determine when to take down posts? When they're taking down posts related to hate speech, potential crimes, etc., why can't they be compelled to forward them to law enforcement?

They have people who examine posts which are reported by users. I don't think they search for violators on their own. Still, how would a filter for crime even work? What are the search terms?

 

[stollcpa]

They have people who examine posts which are reported by users. I don't think they search for violators on their own. Still, how would a filter for crime even work? What are the search terms?

No computer expertise but there must be search terms they presently use for the posts they view and take down.

 

[TheOriginalHappyGoat]

No computer expertise but there must be search terms they presently use for the posts they view and take down.

First of all, again, NO, I don't think they do that. They only examine posts people report. Second of all, even if they had a filter installed, what kinds of search terms could you use to look for signs that someone might commit a crime? You don't need computer expertise to answer that question.

 

[RBB89]

No computer expertise but there must be search terms they presently use for the posts they view and take down.

The content they take down is reported.

 

[Hank Reardon]

They have people who examine posts which are reported by users. I don't think they search for violators on their own. Still, how would a filter for crime even work? What are the search terms?

Eh, there is something in place. A couple of months ago at my place of employment a half dozen squad cars showed up with deputies sporting AR-15's. Turns out a guy who works (er...worked) here was joking around on Twitter about going all postal on his co-workers. He made six or seven posts in a row, and 20 minutes later in rolled the heavily armed LEO's. Allegedly no one specifically called it in. Some sort of monitoring picked it up.

 

[TheOriginalHappyGoat]

Eh, there is something in place. A couple of months ago at my place of employment a half dozen squad cars showed up with deputies sporting AR-15's. Turns out a guy who works (er...worked) here was joking around on Twitter about going all postal on his co-workers. He made six or seven posts in a row, and 20 minutes later in rolled the heavily armed LEO's. Allegedly no one specifically called it in. Some sort of monitoring picked it up.

I didn't hear about anything like that. Did it not hit the news?

Anyway, I bet someone called it in. Maybe not even someone in the building. Just someone somewhere else whose feed it popped up in.

 

[Hank Reardon]

I didn't hear about anything like that. Did it not hit the news?

Anyway, I bet someone called it in. Maybe not even someone in the building. Just someone somewhere else whose feed it popped up in.

It wasn't in the news. The guy was just a clown messing around. They cuffed him, searched his van, etc. It became clear that he was just a jackass. They didn't arrest him and after about 90 minutes of talking with company reps etc, they uncuffed him.. He was fired, or course. To illustrate what a moron the guy is a few hours later he's posting photos from Home Depot bemoaning the paltry faucet selection. He'd worked here 10 years.

You could be right about someone calling it in. I obviously don't know. Just repeating what we were told.

 

[El Drado]

It wasn't in the news. The guy was just a clown messing around. They cuffed him, searched his van, etc. It became clear that he was just a jackass. They didn't arrest him and after about 90 minutes of talking with company reps etc, they uncuffed him.. He was fired, or course. To illustrate what a moron the guy is a few hours later he's posting photos from Home Depot bemoaning the paltry faucet selection. He'd worked here 10 years.

You could be right about someone calling it in. I obviously don't know. Just repeating what we were told.

Someone told me about that...wasn't it at WaterFurnace? That is a special kind of dumb

 

[Hank Reardon]

Someone told me about that...wasn't it at REDACTED. That is a special kind of dumb

I can not confirm this.

 

[stollcpa]

First of all, again, NO, I don't think they do that. They only examine posts people report. Second of all, even if they had a filter installed, what kinds of search terms could you use to look for signs that someone might commit a crime? You don't need computer expertise to answer that question.

I have no idea if this can be done but if these social media companies can monitor my account and bombard me with ads why can't they identify possible terrorist? They pull down offensive posts. They can't be looking at millions of posts a day. They must have some kind of filters in place to narrow the ones they actually look at? I just ask the question could they help. I guess they can't be compelled to help even if they could help.

 

[Sope Creek]

I've seen reports that there were a couple Tweets that might have warned authorities. What responsibility do you think Twitter, Instagram, FB and other social media have to alert law enforcement? I think they have responsibility to alert law enforcement on any posting that even hint of a person propensity to commit any crime.

Does a neighbor have a duty to inform law enforcement if they see something that might "hint" at a person's "propensity" to commit "any" crime?

What does "hint" mean? What does "propensity" mean for purposes of this exercise? And wouldn't this standard require a social media platform to have full knowledge of all criminal laws in every jurisdiction in which it operates?

The standard you've set here, stollpa, is such a hair trigger that neither the social media platform nor law enforcement would have the manpower to do anything with the reports required under it. It's simply unworkable.

And that's on top of the private sector big brother implications of your thought. I just don't see it as being either practical or desirable in the overall scheme of things. Too much nanny state thinking . . . .

 

[TheOriginalHappyGoat]

I have no idea if this can be done but if these social media companies can monitor my account and bombard me with ads why can't they identify possible terrorist? They pull down offensive posts. They can't be looking at millions of posts a day. They must have some kind of filters in place to narrow the ones they actually look at? I just ask the question could they help. I guess they can't be compelled to help even if they could help.

Stoll, your question has been answered multiple times. These sites have a reporting function. Users report posts that violate the rules. That's how they know which ones to examine. There is no magic space-age math that will allow a computer to identify which tweets are regular trolls and which are potential terrorists. Person of Interest is a fictional show.

 

[anon_mlxxvlbug9dpa]

Stoll, your question has been answered multiple times. These sites have a reporting function. Users report posts that violate the rules. That's how they know which ones to examine. There is no magic space-age math that will allow a computer to identify which tweets are regular trolls and which are potential terrorists. Person of Interest is a fictional show.

Stoll, whether he knows it or not, is advocating for artificial intelligence to feed law enforcement investigations into a funnel. It's as impracticable an idea as that of crazed when it comes to agreeing to an executable set of keywords or behaviors that should automatically trigger some sort of mythically resourced pre-crime unit.

Human will is what needs to change. Technology and vigilante watchers can only do so much. Until we (the global we) bend the curve away from religion as the only source of news and guideposts, we're going to have a level of extremism.

 

[stollcpa]

Okay. Stupid question.

 

[toastedbread]

There is no magic space-age math that will allow a computer to identify which tweets are regular trolls and which are potential terrorists.

I sincerely doubt that. There is some fairly heavy stuff going on in the field of machine learning and natural language processing. I'm not sure to what extent it's been applied to this case, but it's certainly not beyond possibility.

 

[TheOriginalHappyGoat]

I sincerely doubt that. There is some fairly heavy stuff going on in the field of machine learning and natural language processing. I'm not sure to what extent it's been applied to this case, but it's certainly not beyond possibility.

Obviously people are working on this kind of AI. But we are nowhere near having computers find terrorists for us online.

 

[toastedbread]

Obviously people are working on this kind of AI. But we are nowhere near having computers find terrorists for us online.

There are probably already some rudimentary models that aren't yet public. I don't think it's that far off. It's not going to tell you with 100% certainty who is/isn't a terrorist. It will give you an idea about a persons sentiment, and whether or not they are trolling. Probably you can also determine who is more likely to commit an act based upon the severity of their sentiment.

 

[TheOriginalHappyGoat]

There are probably already some rudimentary models that aren't yet public. I don't think it's that far off. It's not going to tell you with 100% certainty who is/isn't a terrorist. It will give you an idea about a persons sentiment, and whether or not they are trolling. Probably you can also determine who is more likely to commit an act based upon the severity of their sentiment.

Oh I'm sure there is something like that already in action. As one of many tools in the digital arsenal. But remember what OP was about. This thread was about whether or not the social media services should have a duty to report potential criminal behavior. That's just not workable. A robot in some NSA basement scanning Twitter relentlessly? Absolutely. Some kind of magic formula that immediately warns some techie in Frisco that it's time to call the cops? No way.

Computers know a lot more about us than people would be comfortable with if they really thought about it, but it's all brute force. Google knows which ads to serve you because it has years worth of your search history to tell it what you're interested in. The NSA robot above can compile a list of accounts that agents should keep an eye on, based on previous posting history, list of followers, etc. But what OP is hoping for here, sort of an early warning system based on a bare minimum of information, isn't going to work. In the past hour, 23 Twitter users have used the phrase "I will kill you" in a public post. Now multiply that by the number of phrases that you might need to put on your search list to be useful. You'll be swamped in very short order.

Now, if you combine the watch list from our NSA robot with a search list of worrisome phrases, that might result in a manageable number of serious red flags to look at, but it will still be almost entirely false positives, and the actual positives are only going to cover a minority of the people broadcasting future crimes online. It's still not going to catch the high school senior who snaps and shoots up the school, or the middle-aged guy who only uses social media to talk about sports, but found out last week his wife is cheating on him, and has decided to do something about it. The kid's "Today, everyone pays for what they did to me," and the man's "Time for the bitch to suffer" are likely not getting caught up in those filters, because the computer never had any reason to have those users on the watch list in the first place.

 

[stollcpa]

Oh I'm sure there is something like that already in action. As one of many tools in the digital arsenal. But remember what OP was about. This thread was about whether or not the social media services should have a duty to report potential criminal behavior. That's just not workable. A robot in some NSA basement scanning Twitter relentlessly? Absolutely. Some kind of magic formula that immediately warns some techie in Frisco that it's time to call the cops? No way.

Computers know a lot more about us than people would be comfortable with if they really thought about it, but it's all brute force. Google knows which ads to serve you because it has years worth of your search history to tell it what you're interested in. The NSA robot above can compile a list of accounts that agents should keep an eye on, based on previous posting history, list of followers, etc. But what OP is hoping for here, sort of an early warning system based on a bare minimum of information, isn't going to work. In the past hour, 23 Twitter users have used the phrase "I will kill you" in a public post. Now multiply that by the number of phrases that you might need to put on your search list to be useful. You'll be swamped in very short order.

Now, if you combine the watch list from our NSA robot with a search list of worrisome phrases, that might result in a manageable number of serious red flags to look at, but it will still be almost entirely false positives, and the actual positives are only going to cover a minority of the people broadcasting future crimes online. It's still not going to catch the high school senior who snaps and shoots up the school, or the middle-aged guy who only uses social media to talk about sports, but found out last week his wife is cheating on him, and has decided to do something about it. The kid's "Today, everyone pays for what they did to me," and the man's "Time for the bitch to suffer" are likely not getting caught up in those filters, because the computer never had any reason to have those users on the watch list in the first place.

You're correct in what I was looking for. I see the folly of it now.

 

[Sope Creek]

Okay. Stupid question.

Don't beat yourself up over it . . . that's what we're here for.

In all seriousness, I understand the instinct to want to "do something" about stopping terror attacks if at all possible. My sense is that the best response we can have - aside from investigating, rounding up and prosecuting all those involved, plus taking care of those affected as best as possible - is to live the way we live and let the bastages know that whatever they dish out we can take, which makes their little suicide ventures even more useless and meaningless than they already are.

 

[stollcpa]

Don't beat yourself up over it . . . that's what we're here for.

In all seriousness, I understand the instinct to want to "do something" about stopping terror attacks if at all possible. My sense is that the best response we can have - aside from investigating, rounding up and prosecuting all those involved, plus taking care of those affected as best as possible - is to live the way we live and let the bastages know that whatever they dish out we can take, which makes their little suicide ventures even more useless and meaningless than they already are.

Agreed. We certainly have to be strong and live our lives and not show fear to these bastards. I am sure we're using technology as best we can to identify these people. I often wonder though like in this case where authorities say they were aware of the guy exactly what they're doing with that knowledge.

 

[TheOriginalHappyGoat]

You're correct in what I was looking for. I see the folly of it now.

It's not folly. We're just not really there yet. But the more evidence we leave behind online, and the smarter AI gets, the closer we'll get.

The real question should be this: do you even want to be there? Do you really want to live in a world where a computer interprets your online posts to determine your intent and then reports you to the government?

 

[anon_mlxxvlbug9dpa]

It's not folly. We're just not really there yet. But the more evidence we leave behind online, and the smarter AI gets, the closer we'll get.

The real question should be this: do you even want to be there? Do you really want to live in a world where a computer interprets your online posts to determine your intent and then reports you to the government?

Can anybody tell me exactly what Alan is trying to say here?
http://www.foxnews.com/opinion/2017/05/23/dershowitz-terrorism-persists-because-it-works.html

 

[i'vegotwinners]

I've seen reports that there were a couple Tweets that might have warned authorities. What responsibility do you think Twitter, Instagram, FB and other social media have to alert law enforcement? I think they have responsibility to alert law enforcement on any posting that even hint of a person propensity to commit any crime.

perhaps if someone could show one of the tweets to which you refer, we could have a credible discussion on how said tweet should have been handled.

without such an example, it's IMPOSSIBLE to really debate how such a tweet should be handled.

 

[Sope Creek]

Can anybody tell me exactly what Alan is trying to say here?
http://www.foxnews.com/opinion/2017/05/23/dershowitz-terrorism-persists-because-it-works.html

Dershowitz seems to be taking sides, and blaming Palestinians' financial backing of, and payment for, terrorist acts as the predominant root cause of terrorism. He's saying that punishing terrorists and demanding the entire world - including the Arab world - not to support/pay for terrorist acts is the primary solution.

That said, the former Archbishop of Canterbury said this in 1982, after positing that the oppressor/victim relationship is errantly thought of in terms of a "master-slave" relationship in which one partner in such a relationship is always and solely defined by the other:

"God is not 'with' the victim in order to make us [oppressors] victims; so the preaching of the resurrection affirms. Yet we go on seeking some firm ground for a 'justice' that will invert the existing order to create new victims out of old oppressors - or, worse still, new victims out of 'neutrals'. One of the greatest historical tragedies of our epoch has been the fate of the Palestinian Arabs. Europe's attempt to atone for a nightmare of incalculable violence against the Jewish people has produced a new race of victims in the Palestinians, and so set up a further chain of terrorist counter-violence as a result of the violence (so much compounded and embittered in recent years by the reckless extension of Israeli settlements into the occupied territories, and the leverage possessed by extreme conservative political groups in the Israeli political system) offered to the indigenous Arab populations. It is distressingly sharp illustration of the circularity of oppression; and we should make no mistake about its origins in European Christian anti-semitism." (Williams, Rowan: Resurrection, Interpreting the Easter Gospel, pp 11-12.)

I have no idea what Williams' commentary would be today regarding Dershowitz' op-ed piece specifically, or regarding the Israeli-Palestinian mess of today more generally. But his point about the circularity of violence/oppression/victimhood remains spot on, IMHO. Dershowitz gets at interrupting that circularity by placing the onus solely on the Palestinian/Arab elements in these conflicts. IMO that is an avoidance of responsibility and accountability on the rest of the world . . . too.

 

[UncleMark]

Can anybody tell me exactly what Alan is trying to say here?
http://www.foxnews.com/opinion/2017/05/23/dershowitz-terrorism-persists-because-it-works.html

He's blaming the Palestinians for the terrorism we see today, because "it works". Although the state of Israel, born of terrorism, is blameless.

 

[i'vegotwinners]

Dershowitz seems to be taking sides, and blaming Palestinians' financial backing of, and payment for, terrorist acts as the predominant root cause of terrorism. He's saying that punishing terrorists and demanding the entire world - including the Arab world - not to support/pay for terrorist acts is the primary solution.

tell Alan, good luck with that.

 

[stollcpa]

It's not folly. We're just not really there yet. But the more evidence we leave behind online, and the smarter AI gets, the closer we'll get.

The real question should be this: do you even want to be there? Do you really want to live in a world where a computer interprets your online posts to determine your intent and then reports you to the government?

Well that was a cold splash of ice water to my face. We probably don't want to go there.

 

[stollcpa]

perhaps if someone could show one of the tweets to which you refer, we could have a credible discussion on how said tweet should have been handled.

without such an example, it's IMPOSSIBLE to really debate how such a tweet should be handled.

Unfortunately a reporter made the statement with no photos of tweets.

 

[TheOriginalHappyGoat]

Well that was a cold splash of ice water to my face. We probably don't want to go there.

I agree. We all know what the road to hell is paved with.

 

[anon_mlxxvlbug9dpa]

Dershowitz seems to be taking sides, and blaming Palestinians' financial backing of, and payment for, terrorist acts as the predominant root cause of terrorism. He's saying that punishing terrorists and demanding the entire world - including the Arab world - not to support/pay for terrorist acts is the primary solution.

That said, the former Archbishop of Canterbury said this in 1982, after positing that the oppressor/victim relationship is errantly thought of in terms of a "master-slave" relationship in which one partner in such a relationship is always and solely defined by the other:

"God is not 'with' the victim in order to make us [oppressors] victims; so the preaching of the resurrection affirms. Yet we go on seeking some firm ground for a 'justice' that will invert the existing order to create new victims out of old oppressors - or, worse still, new victims out of 'neutrals'. One of the greatest historical tragedies of our epoch has been the fate of the Palestinian Arabs. Europe's attempt to atone for a nightmare of incalculable violence against the Jewish people has produced a new race of victims in the Palestinians, and so set up a further chain of terrorist counter-violence as a result of the violence (so much compounded and embittered in recent years by the reckless extension of Israeli settlements into the occupied territories, and the leverage possessed by extreme conservative political groups in the Israeli political system) offered to the indigenous Arab populations. It is distressingly sharp illustration of the circularity of oppression; and we should make no mistake about its origins in European Christian anti-semitism." (Williams, Rowan: Resurrection, Interpreting the Easter Gospel, pp 11-12.)

I have no idea what Williams' commentary would be today regarding Dershowitz' op-ed piece specifically, or regarding the Israeli-Palestinian mess of today more generally. But his point about the circularity of violence/oppression/victimhood remains spot on, IMHO. Dershowitz gets at interrupting that circularity by placing the onus solely on the Palestinian/Arab elements in these conflicts. IMO that is an avoidance of responsibility and accountability on the rest of the world . . . too.

Right. My read of his manifesto was Israel good ; Palestine bad.

Terrorism far predates Palestine. It's crazy how hot and cold Dershowitz can be. Perhaps he's just a populist.

 

[IyamCornholio]

Don't they have filters to determine when to take down posts? When they're taking down posts related to hate speech, potential crimes, etc., why can't they be compelled to forward them to law enforcement?

The intelligence agencies could have similar filters looking for phases that would be signs of terrorist activity. What they do with the information is another question. Identifying the real threats may take human involvement.