When I moved to the US for a few years in the early 1980s, I was surprised when a clinic gave me my own medical notes — those were still the days of high paternalism in the British healthcare system. But as I reflected on it, clutching my file and X-rays, it made sense: after all, the real data was encoded in my body.
Of course, I was wrong, at least as regards data as we understand it now. The encoding that matters is digital; it is the capacity to represent all kinds of information digitally that is driving the hype and hope about data and its artificial intelligence applications. My health data is not my blood pressure and pulse but the trail recorded in a smartphone app. Digital trails are ubiquitous, capturing everyone’s behaviour from movement to TV-watching to shopping, not to mention those increasingly generated between devices in urban sensors or the internet of things.
All this data is useful to the extent that it provides information enabling people — or machines — to make better decisions, although what “better” means is not always clear. However, it is not straightforward to figure out how to crystallise this value, whether in commercial gain for businesses, convenience and wellbeing for individuals or improved services in the public sector.