Data as Asset to Data as Liability

Image credit: Ashley Jurius on Unsplash

Many problems occur while working with data, therefore there is a need to address it more while working with data.

“Data is the new oil.” Breathlessly repeated at conferences and in classrooms the world over, this cliché has come to represent the worst of big data’s hype. The phrase has also unintentionally served as a harbinger of data’s risks and harms, which are growing at an alarming and accelerating pace.

The warning signs are everywhere. Sensed data may not be reliable—authentication techniques employing blockchain may be one solution, but the risks of faked data leaking into mission-critical applications are growing fast. And this won’t be our first encounter with bad data—racial bias is now widely understood to be endemic in large data sets, even by those who’ve created and profit from them. Still, many issues go unaddressed because data work is undervalued by AI researchers and practitioners, who would rather work on algorithms and analytics. And important tools like synthetic populations, which could help address many data risks, are often improperly used.

Will the data practices prove sufficient to stand close scrutiny or the extreme pressures of unanticipated stress? How will the ethics of data be treated in the future? How to ensure that procedures, rules and mindsets are in place in the forthcoming future?