Big Data: cutting-edge or perilous? In their talk on "Big Data and the Workforce: Between Self-Tracking and Corporate Panopticon" on Day 3, part of our focus topic "Work", Andrea Kocsis from the German union ver.di and quantum physicist Andreas Dewes discussed data retention in the workplace, moderated by consultant and researcher Johannes Kleske. Algorithms are like your new, unpopular colleague in the office. They will evaluate and optimize the workflow, all the while diligently collecting their co-workers' data. How to deal with this situation, was the core question of the debate between the two panelists.
Algorithms are seen as capable of correcting the mistakes and bad decisions people have made. But that isn't really the case, says Andreas Dewes. "Algorithms do not automatically make fair decisions," he says. "They are nothing but recipes for computers." And it's us people choosing the ingredients for that recipe: Algorithms base their decisions on the data that is initially fed to them in training. So if you train your custom algorithm with data from your white male employees, for instance, it will in future tend towards these kind of data inputs. And probably not hire black applicants. Dewes fancies the idea of a kind of technical safety inspection for algorithms, to have them examined by an independent authority.
Robots at Work
But robotics and big data also bring benefits to the workplace. We can let technology do mindless or bodily tiring work for us. A logistics company, for example, introduced data goggles for their workers. Instead of them having to carry around a list with shelf numbers, the goggles would show them the way. "The workers loved it, because they had another free hand now," says Andrea Kocsis. But there was always the danger that some jobs would be lost in this manner too.
Parcel deliveries can now be tracked using GPS, so that customers know where their parcel is and when it will arrive. There have been Amazon employees who were given notice after having been inactive for 30 minutes of their working day, even for being on the toilet. Dewes mentions these cases as examples of surveillance in the workplace, and cautioned against demanding 100-percent productivity from employees. If someone is not always equally productive, says Dewes, that in itself may help them to stay balanced and to do better work in the long run.