Minority Report is real : we can now foresee crime

Dec 6, 2018
Vote on Hacker News

British police walk the beat

1956: That was the year the book “Minority Report” was written by Philip K. Dick.
2019: That’s when it will become a reality.
With the National Data Analytics Solution, welcome to the age of preventing crime before it happens. Maybe.

Preventing crime

“Minority Report” is a novel written by Sci-Fi author Philip K. Dick. In his dystopian world, crimes can be foreseen and prevented. This leads to people being detained before anything happens (ie. “detained for no crime”).
West Midland Police (UK) is currently working on a project to predict serious violent crime. It relies on artificial intelligence, statistics, and big data. Enter the National Data Analytics Solution (NDAS).

How it works

Using records of prior arrests, Police database, mental health records, the NDAS will analyze as many as 1400 indicators. Of these 1400 indicators, 30 seem particularly powerful in predicting future offenses.
This only applies to people having committed a previous felony. NDAS is not designed for law-abiding citizens… yet.
The “known associates”, and the records of the social group the individual is in, are taken into account.
The idea is to identify the people most likely to turn to more violent crime.
Once identified, they won’t be arrested but will benefit from counseling with social workers or mental health professionals.
Machine learning will help make the analysis more accurate with time.
As Police forces shrink, this could help them concentrate on the population that holds most risks.
Although it is still in development and limited in scope, NDAS is destined to be widely used in all of UK once it is fully operational.

Shortcomings and ethics

The whole process gives a shivering feeling, ethically speaking. But there are also matters of concern about the more technical aspects.
One of them is the data collection. The data comes mainly from the Police database. This means that less populated areas will be left behind and the most populated ones will be under heavy scrutiny.
In the same way, belonging to specific ethnic groups could target people for intervention.
The whole system, pretty much like in the novel, poses the same old nagging question. Will the crime occur indeed if nothing is done?
Time machine still being in R&D (as far as we know ;-), there is no way of answering that crucial question.
Identifying risky individuals sure raises ethics questions. The benevolent spirit of the entire project may have overlooked some issues. The Alan Turing Institute in the UK has brought these concerns to the team working on NDAS. It seems the two will now collaborate to produce a more ethics-oriented system.

Ethics is an adjustable variable

Uk is not the only place where such technology is in use. Elsewhere, it can pinpoint hotspots where crimes are more likely to occur. This could be a valuable strategy as remodeling urbanized areas could help lower criminality.
Detecting “crime indicators” could help widen knowledge on what causes criminality and offer options to act on a wider scale. Pinpointing individuals who have already committed offenses may come a bit late. Acting on the levers that brought about those offenses could prove more efficient (in the long run).
Unfortunately, some nations are less bothered by ethics. Profiling tools such as NDAS (coupled with surveillance) are already used to detain people in regions of China.
As always, tech breakthroughs come with their load of “we can but should we?” questions. Profiling is but one of them. Ethics should be integrated early on in R&D so we don’t end up in a Philip K. Dick novel.