Big data could eventually result in would-be criminals being caught before they commit a crime, according to Kenneth Cukier, data editor of The Economist.
Cukier, who was speaking yesterday as part of the London arm of Big Data Week, said: “We’ll have algorithms that can predict behaviour and mean we could be punished before committing a crime.”
The suggestion is highly reminiscent of the Tom Cruise film Minority Report, which is based on the Philip K Dick story of the same name. In the film, a specialised PreCrime department uses foreknowledge provided by psychics to catch criminals before they break the law.
Big data is already in use in some police forces. Los Angeles police made the headlines in 2012 for using a crime prediction algorithm to indentify likely crime hotspots and arrive before the criminals, which resulted in a 25% drop in thefts.
Cukier believes that this is just one aspect big data’s future role in our lives.
“We are just at the beginning,” he said. “It is going to invade all aspects of human endeavour and that’s a good thing.”
Big data is already unlocking knowledge about everything from voting patterns to cancer diagnosis, and has the potential to provide remarkable levels of detail about human behaviour. It could eventually provide a level of knowledge about us that would have previously only been thought possible with psychic abilities.
However, for some big data represents a threat. In particular it raises serious privacy concerns both from a data collection point of view and in terms of behaviour prediction.
Even the crime-catching technology would raise significant moral concerns should it come to fruition, as it would raise the issue of how pre-criminals would be punished given that they have not actually committed a crime.
Cukier likened big data to nuclear technology, in that it has both beneficial and damaging applications.
“I think there are thousands of ways big data could inflict incredible harm on society,” he said, adding that the big data industry needed to “keep going” with the technology despite the likely problems.
Recent news that Google Flu Trends overestimated the illness’ prevalence by half has led some to question the usefulness of big data.
However Cukier believes that the press’ “decimation” of big data over the news is largely unfounded.
He explained how the data it was compared to is just for visits to health clinics, which means flu suffers who stayed at home would not have been counted.
Given that many people may have avoided a clinic trip due to a lack of health insurance, a lack of transport or simply the desire to stay in bed, this could equate to a huge number of people.
“It’s possible – not totally likely, I’ll admit – that the Google searches were a better indicator of flu than the official data,” said Cukier.
First body image courtesy of Predpol. All others screenshots from Minority Report.