With the advent of digital, our lives have changed radically. Internet allows you to have access to almost all human knowledge in a universal way. Whenever we use our address personal email or we go on a trip and immortalize the most beautiful moments with tens or hundreds of photos taken by smartphones that are increasingly similar to professional cameras, thousands of cloud server consume electricity and redundant hard disk space in different parts of the planet. All, often, without the user spending a penny. But, as an American adage from the beginning of the last century says, referring to the practice of many bars of offering patrons free of charge small snacks, often salty, in order to stimulate the consumption of alcohol, "there is not such a thing as a free lunch".
In digital, not paying does not mean free
It is known that the big digital giants choose to offer these and other products without charging users because in return they get very precious data on their behaviors. As long as you are aware that don't pay doesn't mean free, the revolution that Google, Microsoft, and many other tech giants have started and continue to carry on has an unprecedented scope in terms of democratization of information and access to tools previously for the benefit of a few. Every day these services permeate our lives more and more unconsciously.
When we organize a trip we look for a hotel on Booking and choose where to dine on Trip Advisor. We do it because it's comfortable, practical and it works: the algorithm is powerful and the reviews of other users reinforce trust.
When we do a Google search we have the unconscious expectation of finding what we are looking for among the first results and, if this is not the case, it is much more frequent to quickly find ourselves redoing a search with different terms rather than going to the periphery of page 2. We have so much faith in the algorithm to doubt ourselves: when something goes wrong we probably searched badly.
The imperfection of algorithms
Algorithms, in the perception of most, are impartial and efficient. This is why our faith in them is blind and, rather than questioning them, we are willing to question ourselves. They work in an almost magical way, impossible to understand even for the most expert. However they work and that's enough for us.
However, an algorithm, that inscrutable black box on which we rely completely, is not perfect. Being built by human beings, it potentially contains all the mistakes that a human can make and all the choices that the same human has consciously decided to make.
Il machine learning and artificial intelligence they do not solve the problem: an algorithm is only as good as the dataset that is used to train it and, consequently, subject to the same previous limitations and biases.
Imagine a company that, having to recruit staff, chooses to use an algorithm trained on thousands of resumes of successful European and American managers to identify candidates most likely to make a career. This algorithm will tend to give higher scores to white men. This is not because they are the best candidates but because they are the ones who most frequently occupy the most important positions today. If not trained properly the algorithm discriminates and iterates the status quo. The opposite of what our example company was hoping to achieve. Even worse: it will hardly be noticed.
Algorithms are not infallible. They aren't perfect. They are not goals. By mistake or by choice: by construction.
The dominance of the giants of the web
Very often these big and small web giants have sizable positions in their sector with market shares close to 100%. Even if they do well, the impact they can have on the lives of consumers and the lives of other companies is huge.
If due to a technical error, or an arbitrary choice, one of these digital platforms were to black out all the restaurants whose names begin with the letter P or the hotels that have a blue facade, these blacked out companies would see their turnover drop rapidly, without well understand why it is happening and without great possibilities of appeal. This certainly places an economic issue.
Taking the concept to extremes, if one of these companies decides to try to influence an electoral campaign in a specific direction, perhaps because a particular party has a political position more similar to the company in question, it would be extremely difficult to detect. This poses a social issue.
These algorithms, these choices or errors, are practically impossible to control from the outside if not with empirical evidence. In a world increasingly controlled by algorithms, this poses an objective risk that is difficult to recognize and adjust.
The problem is the misperception of objectivity of these algorithms and the consequent feeling of autonomy and freedom of choice on the part of consumers.
This is amplified by an extremely relevant position that these companies have managed to build over time, thanks to their ability to innovate outside the box, to invest and to attract talent. They are private companies, but often de facto represent the entire market. It is difficult to regulate them because they lack the skills and understanding of what they actually do, but also because there are no easily applicable solutions.
A conscious use of the web
These digital businesses shouldn't be demonized, they are the growth engine of innovation, positively impact the lives of millions of companies and billions of people every day. They have paved the way for new business models, enabling universal access to information and knowledge. They are social levellers and democratizers. Together with the infrastructure, they allow us to work from home, connect with distant relatives or friends, and do business. However, it must be used consciously and we need to find the regulatory framework that enables these companies to continue to innovate while protecting the market from damage and incorrect behaviour. Not an easy task.
As a consumer it is always Better to search across multiple platforms to choose which food delivery to order tonight in order to have access to the maximum possible offer. Compare prices, quality and level of service offered on different sites before making an online purchase. In short, being a little less comfortable and a little more aware.
On the other hand, they also taught us this at school, the best way to find out is read multiple sources and then form an opinion own. Possibly critical.
°°°The author is Commercial Operations Director of Vodafone
