A late morning in April 2017. The postman rings and says: “Hi, here’s the Amazon parcel you were going to order this afternoon…“ Science fiction?
Not much longer, if it were up to the online seller holding a patent that describes the model of “anticipatory shipping”. If a customer shows strong interest in a product, it would be proactively dispatched to a nearby distribution center and is ready to be shipped as soon as the customer places the order.
This is a quite innocent example of the era of the algorithm - an era we are just entering and in which needs and patterns of behavior are analyzed not only subsequently but also anticipatorily. Needless to say, algorithms are nothing new. The eponym of the calculation rule, the Iranian polymath Muhammed al-Chwarizmi, lived more than 1000 years ago.
Two relatively new developments have boosted the topicality of algorithms: the extensive collection of digital data (big data) and the availability of virtually unlimited computing power. In combination, these factors are gaining an alarming level of power. Google’s algorithms make the difference “only” between the success or failure of online shops - the NSA algorithms used in the drone war probably between life and death.
But how can the era of the algorithm be designed along a reasonable concept? Let’s look at the factors: The genie of exponentially increasing computing power - Moore’s law - is out of the bottle and won’t return there. Enforcing the disclosure of algorithms would be useful and create transparency. Too bad it’s utopian - after all, an algorithm is a business idea. Google disclosing their algorithm is about as likely as Coca Cola printing handouts of their soda recipe.
Consequently, the only possible parameter can be personal data, and it should be the goal of a modern civil society to give control of this data back to its creators. A hot topic, not only for politicians but for software developers as well.