“Big data” is going to solve a lot of problems we are told. The ability to amass and analyse huge corpuses of information will assist in spotting business trends, preventing diseases, combating crime and other amazing things. The collection of all this will allow us to become better predictors of a whole gamut of phenomena. There is even a phrase for this new ability: “predictive analytics”.
Remember “Minority Report”, the movie featuring Tom Cruise based on a Philip K Dick (NB one of my favourite SciFi authors) short story? In a nutshell, we will in a future world be able to predict and prevent future crime through the help of mutant “pre-cogs” who see crimes before they are committed allowing the police to “pre-arrest” people. Kinda scary, isn’t it?
Pardon me if I am skeptical about all this new-fangled “science”. I was pleased to see a piece in the Telegraph by an old friend, Jamie Bartlett from the UK thinktank Demos, on using data to predict terrorism and radicalisation (see it here). He warns that having the public report everything and letting authorities to have access to everything we have EVER said or written is disastrous – and I couldn’t agree more.
Stopping crime, or terrorism, is not merely a case of having all possible information. It is more about having the right information about the small number of people who actually do violent things. And that number is very small indeed.
Bringing it back to radicalisation and terrorism, it is a fact that most people who have radical thoughts, or even violent ones, never progress to act on those thoughts. And there is no model, no framework, no matter how much “Big Data” you want to throw at the problem, that will allow you to narrow the field.
Setting aside the very worrisome concerns about privacy and government power (should any one body have that much access to information? Then again, lots of people don’t seem to be too antsy about putting their whole lives on FaceBook or other social media), more data does not necessarily enhance predictability. It would actually make things worse by flooding the responsible agencies with mostly useless fluff and I am not sanguine on the ability of software to successfully separate the wheat from the chaff to any degree of accuracy. Intelligence analysts already “drink from a fire hose”: making the fire hose wider will not help.
Let’s be realistic about all this. The large majority of terrorist attacks will be stopped by good intelligence and law enforcement action: some will get through – accept it. We do not need CSIS and the RCMP to track all people who may at one time utter a sentence that may be consistent with violent radicalisation. What we need is smart people who understand the threat and know how to put it into context.