The angels' share
Big data has received a lot of bad press recently but its potential for more ethical applications remains, writes David Burrows
Back in 2013, in an essay for Foreign Affairs magazine, Kenneth Cukier and Viktor Mayer-Schoenberger predicted that big data was poised to reshape the way we live, work, and think. “Big data will become integral to addressing many of the world’s pressing problems,” they wrote. But five years on, and the Facebook-Cambridge Analytica scandal has thrown the spotlight on the demons that such data can create. “This is the story we have been waiting for, so people will pay attention not just to Facebook but to the entire surveillance economy,” says Siva Vaidhyanathan, a professor of media studies at the University of Virginia.
It’s alleged that the personal data of about 50 million Americans was harvested from the social networking site, “improperly shared” with the political consultancy, and then possibly used to influence the 2016 US presidential election. In the UK, Dominic Cummings, the director of Vote Leave, has also been forced to deny allegations of links between his campaign and Cambridge Analytica. The news brought to mind a retail white paper I wrote in 2015 that began:
“Researchers at the University of Cambridge have just created an algorithm that uses Facebook ‘likes’ to determine an individual’s psychological traits. Apparently, the computer can judge personality more accurately than workmates, friends or family members. In fact, given enough ‘likes’ to crunch (300), it could even rival a spouse on a range of psychological traits including conscientiousness and neuroticism. This might take us one step closer to Minority Report, but the finding has important social and economic implications – the ability to judge personality is an essential component of social living: who to marry, hire or elect as prime minister.”
Oh.
The more brands know about us, the easier it becomes to persuade us to buy their wares. Facebook has an almost hypnotic power to pull in punters – its adverts increase audience reach, ad memorability, brand linkage and likeability. Junkfood brands, for instance, have fattened their profits, thanks to a platform that costs a fraction of traditional media such as television. The fact that online advertising restrictions are voluntary, and full of loopholes, makes the whole thing even sweeter.
The recent scandal leaves a sour taste, but I hope it doesn’t persist. Sure, we need to weed out the devils in this sorry tale, but data can still be a powerful tool for social good – and that includes Facebook.
For example, researchers from Boston Children’s Hospital, Massachusetts, found that in areas where there were lots of Facebook ‘likes’ for healthy activities, the obesity rates were lower; where the social media love was for TV shows, people were more likely to be overweight. Online social networks such as Facebook represent a high-value, low-cost datastream for looking at health at a population level, says John Brownstein, one of the team involved in the study.
Indeed, the tight correlation between Facebook users’ interests and obesity data could help to generate real-time estimates of obesity levels in an area, and to launch more effective, targeted public health campaigns.
It’s just one of several examples where data is being massed, mined and manipulated for a purpose. Telefónica Brazil, for instance, is using mobile network data to monitor air pollution in São Paulo, Brazil’s largest city. Pollution problems can now be predicted up to two days before they occur, so traffic can be diverted via alternative routes and areas of high pollution can be flagged to residents.
Purpose and profit
Sometimes there is purpose and profit. A project in Spain combined pollution and consumer data to show that people spend €25m-€41 million (£22m-£36m) less on days when ozone pollution is 10% worse than usual. When particulate matter pollution is 10% worse than usual, spending falls by €20m-€30m (£17.6m-£26.4 m). The results provide a timely kick up the backside for governments, including the UK’s, to improve air quality.
And these examples are just the tip of the iceberg. In time, data from Uber and taxis could be combined with that from smartphones to better understand how people get around, and in turn design more attractive and effective public transport systems, reducing pollution and carbon emissions. Drones are already being used to relay information on when crops will be ready for harvest. But there is potential for this to be married with weather forecasts and consumer demand data to balance supplies and cut wastage.
Whether it’s climate change or food scarcity, deforestation, poverty or obesity, data will help to show us how dire the situation is, predict how much worse it will get, and – critically – identify where to focus our efforts to turn things around. The data will be both public and private, so dealing with it requires delicacy.
“Corporations have a lot of data that, if unlocked in one way or another, could be used for a variety of public good objectives,” says Stefaan Verhulst, co-founder and chief research and development officer of The Governance Lab. But because not many yet understand how their data can be used for public good, he adds, there is “quite often no imagination of what the value of private data might be”.
The devil’s work or heaven-sent opportunity?
David Burrows is a freelance journalist