published on 17 May 2017 in air
Algorithm and society
Fig. 1 – Graphic representation of the algorithm
It underlies many of our everyday actions, dictates what happens on the web, affects our “web reputation” on social media and chooses the ads that we run into on the Internet. It’s the algorithm, a mostly unknown entity, that day after day advances by a few centimetres in controlling of our lives and that – by definition – consists of a mathematical procedure to solve a problem through a finite number of steps. But let us try to understand a little more.
Etymology and definition
When the meaning of a word or concept is not entirely clear to us, it is good practice to try to shed some light by starting from its etymology. According to the Treccani dictionary, the term algorithm derives from the medieval Latin algorithmus or algorismus, from the name of the place of origin, al–Khuwārizmī, of the Arab mathematician Muḥammad ibn Mūsa of the 9th century, so named because a native of Khwarizm, a region in Central Asia.
Fig. 2 – Internet of things
The algorithm is therefore something that has to do with mathematics and since society has gradually become more computerised and digitised, the algorithm has gradually acquired more importance. Starting from huge masses of data – big data – that, for example, we produce through our smartphones, the algorithm makes a series of ultra fast calculations, which hierarchically arrange the information collected. It is an algorithm that “decides” which posts to make appear on our Facebook wall and it i again an algorithm that, based on our previous clicks, establishes which advertisements to propose to us before a YouTube video. In general, we can say that the algorithm is a form of reorganisation of the numerous raw data that each of us produces, for example, when making purchases, when on the move, when interacting through social networks.
To get an idea of the quantity of data we are talking about, suffice it to say that every day 3 billion Internet users are exchange 144 billion emails and that every day on Facebook, its subscribers exchange 4.5 billion likes. All this information, if interpreted correctly, allows reality to be encoded and, in a sense, our future actions to be predicted, such as which objects we will be more likely to buy. Precisely for this reason, some consider big data to be like oil, as if it were, in a sense, the new fuel of the society of the future. Already today, in fact, algorithms control part of our lives, they affect and determine them, influencing our choices. So let’s try to understand how they work.
Four families of digital computing
Without going into too much detail, let’s try to add a bit to our knowledge of the magical world of algorithms by trying to differentiate them into four main categories. According to the sociologist Dominique Cardon, the first step is to understand which are the “families” into which we can divide the different types of algorithms that now dominate the web: those based on popularity, on authoritativeness, on reputation and onpredictive measurement systems.
Fig. 3 – Icons of social media, such as Fb, Twitter, Linkedin, etc.
The first family, that of popularity, measures website audiences by simply counting visitor clicks. Part of this group is, for example, Google Analytics, the tool which allows webmasters to have the figures of their websites, by recording the IP addresses that connect to the web pages. The second family, which Cardón calls that of authoritativeness, is typically that used by Google, which indexes websites based on the quantity of hypertext links: the more a website is linked from another website, the more it is considered authoritative and, therefore, worthy of being at the top of the hierarchy established by the search engine. The third group of algorithms, on the other hand, goes hand in hand with social media. It’s that of the Facebook likes, based on which, each of us, depending on our presence on social media, on the extent of our network and on the pervasiveness of our message, raises or lowers the bar of our “glory meter”, also called “e-reputation”. Last but not least, the fourth family is that of “prediction” and is based on machine learning, a particular statistical technique that, for example, allows Netflix or Amazon to suggest new products based on our past choices.
The syntax of our time
Each of these families has advantages and disadvantages, but what is important to understand, regardless of the classification, is that algorithms are acquiring a huge importance in our lives, starting with the fact that they regulate our access to information, be it of a political, commercial or cultural nature. If in fact Galileo Galilei spoke of a book of nature written in mathematical language, we can say that – today – the syntax of the world we live in is certainly based on the algorithm. The conclusion? To have full awareness of reality, to have effective control of our choices and to take full advantage of the opportunities brought about by the dissemination of the Internet, it is essential to have adequate intellectual tools to understand the algorithms. Only in this way can we control the technology instead of being controlled by it.
- Dominique Cardon, What do algorithms dream of?