A man, obviously “down on his luck” as they used to say, shuffles along a city street mumbling, occasionally shouting at no one in particular. People pass by, maybe averting their eyes, perhaps even crossing the street.
If he becomes more aggressive, possibly shouting, swearing, lunging at people, someone may call the police or, if in a public place like a shopping centre, alert security.
If he goes as far as physically assaulting someone, again people will likely call the police, alert security, sometimes a member of the public will intervene.
When people get louder, more strident, and belligerent, they become more noticeable; does that automatically make them a threat? Probably not.
There is a continuum: ignore, call someone to deal with them, or intervene.
How could that work in cyberspace?
Does the fact someone is mouthing off, saying something unpleasant, disagreeable, to some people even insulting or unacceptable, mean we need more controls or create new laws?
No, it means we need the means of knowing when that person becomes threatening, potentially violent and dangerous.
It is when their talk becomes a threat we need to listen. When behaviour becomes a clear and present danger, we need to intervene.
The trouble is, online, there is no one to call. If there were, how would people know the difference between meaningless chatter and genuine threats?
At the moment, many solutions are along the lines of more-police-on-the-beat, let’s-install-CCTV or some other form of monitoring.
Aside from the lack of evidence such approaches work, any response is bound to be too little, too late.
We need to figure out how to: alert, assess and act.
The trouble is online, it is hard to tell who or what you are dealing with.
Is it youthful braggadocio or genuine jihadist threat?
Is it a wimpy kid acting macho or a potential school shooter?
We need to find creative ways of finding and assessing content — content far too nuanced to rely on algorithms to find, far less assess.
Relying on the public is not a good idea, either, particularly when assessment easily converts to action.
In Kenya, for example, suspected young criminals, have been profiled on various Facebook groups by so-called “gangster hunters”.
“They profile them on Facebook, after one week or a month they shoot them, and put pictures of their dead bodies on Facebook,” Wilfred Olal from the Dandora Community Justice Centre says.
Photos, — close-ups of heads with gunshot wounds and lacerated bodies — are posted with a warning that other criminals face the same fate. At times Facebook blurs the photos, which users can choose to un-blur.
So much for crowdsourcing.