Data and Information

My previous post was written extemporaneously as a response to comments. This post had been written first and I had expected it to be the last on the subject of the response to mass shootings. I intend to turn to other subjects.

Whenever and wherever in the world a mass shooting takes place everyone hears about it, people demand action, and governments respond. That response could be more expensive – in dollars and consequences – than effective.

Will the government intelligence agencies be given a broader mandate and more capability to increase their surveillance of the public in response to a demand arising from mass shootings, terrorism, and criminal activity?

Maybe.

To extend technology-based surveillance capability to its limits, you would have to capture every phone conversation and every e-mail, and put CCTV with facial recognition on every street – all of that has been tried. To extend human-based intelligence capability to its limits you would have to deploy flocks of agents or make everyone a snitch – historically, that too has been tried. That and the once popular use of torture are, however, now considered both problematic and anachronistic.

On the other hand, the private sector has successfully made maximum use of technology and minimum use of interpersonal interactions to gather data to increase sales. Emphasis on successful. It has, it is true, been aided by the mind-boggling willingness of the consumer to give up their personal data, their preferences, and their habits. Governments may not have that same advantage but might still be persuaded that technology-based approaches are best.

Technology is more intrusive in a less-intrusive way. Snitching is less intrusive in a more-intrusive way.

In Canada, our legislators will continue to encourage whistle-blowing but they could also infer – even without the persuasive impact of glossy brochures and a free lunch – that using technology can be the way to go. After all, technology is so passive that the public won’t even know it is there. They could extend the agency mandates within existing legislation without parliamentary debate. And then all that is left is to write the cheque. Success should be so simple.

Data are fact. Information is opinion. No one can argue fact – although some still try – but anyone can have an opinion. Putting aside the errors that arise from not collecting all of the data, or collecting the wrong data, the primary source of intelligence errors is in the analytical processes that convert data to information.

This is where the private sector has an advantage. Their mission is to get you to buy and the consequence of failure is far less than for government if the mission is to prevent death and destruction. The marketing data are more discrete and the private sector can afford to refine their analysis through trial and error using self-improving systems. Time is on their side. If they make an error, then they don’t get the sale today. For them, technology works.

The data underlying the government mission contain more variables and require more extensive use of human analysts to connect the dots. What happens if an error is made in a government system? The act is not prevented because the actor was not foreknown. The innocent who are targeted in error are left to recover as best they can. The demand and response increase, seeking and providing more and more intrusive means. And consequently, to spend more money.

Both approaches are prone to error but in one case being wrong doesn’t matter.

Attempts by government to protect us – at our demand – from mass shootings, terrorism, and criminal activity through broader and deeper surveillance has unintended and unwanted consequences.

Let’s be careful what we ask for.

Stuart Syme