I take the bus and find that my seat has already been reserved. Good! I get off at the fourth stop and enter a large hypermarket. Right at the entrance, in a carefully arranged basket, are the avocados and tomatoes that I buy every week, the juices that I like the most and my usual toothpaste, which was running out.

As I head to the checkout, a few bags of snacks appear on the conveyor belt, unasked for. Why not? It’s Friday and friends will probably be coming over for dinner afterwards and it could be fun. So we watch together the next episode of my personalized series, generated by an artificial intelligence system that has created just the characters and stories I like… How will it end?

Although this scenario is currently a fantasy, it could become a reality in the near future thanks to three elements that already exist: a world connected by mobile phones – used by more than 75% of the world’s population – a complete record of our preferences, location and activities that we allow companies to use , and the intelligence of today’s algorithms, which work behind the scenes, behind our backs.

The accommodating servant

Algorithm is a term of Arabic origin that means “recipe”: a set of simple steps that can help us solve a problem, if followed rigorously. An example would be the “algorithm” for putting on makeup: it involves a series of more or less mechanical tasks, with a certain order between them. And the end result is that we look great when we go out or when we are interviewed for a job.

However, in its early days, this idea of ​​a mathematical or computer recipe was very rigid. The pioneering woman to whom we owe the first algorithm that could be run on a computer was Ada Lovelace , and her role is key to understanding the history of algorithms . She worked in the first half of the 19th century, when no computer really existed. And she did so assuming that a computer would be a “machine,” something capable of performing only mechanical tasks.

Today we know that algorithms go much further. In addition to mechanical tasks, they are used as decision-makers and predictive tools, such as the programs that control robots that assemble cars in factories or the programs that regulate traffic lights to ensure smooth traffic flow.

Algorithms in your life

In the little experiment we started with in the article, there is an algorithm that helps us find a reserved seat on a bus, and another that consults the data accumulated from previous purchases made at the supermarket and helps us not to forget anything. It does not just repeat our habits: since it is based on the idea of ​​“learning”, it takes into account data such as the weather (if it is hot, we are likely to want a juice) or what day of the week it is. That is why it is able to predict which items I want to take home today.

To do this, they need to incorporate something called machine learning : no human programmer tells them what the relationship is between the data and the result. Today the shopping list is one thing, and tomorrow it will be another.

Within this type are the recommender algorithms . They started precisely on shopping websites, especially for music and books, and suggested new purchases based on those already made. In addition to that, as they are sites with many registered users, they incorporate information from other customers: we all know the phrase “other people also bought…”.

From servants to masters

Meanwhile, without having thought about it, another algorithm has analyzed my last conversations and has seen that I showed interest in playing some games at home with my friends, around a table, sharing those drinks. In this case, the algorithm has done something that we perhaps did not expect: it has served me the “content” that it thinks I will buy, even if it has no direct relation to my habits.

Since he knows my conversations as well as my tastes, because he can examine my social media profiles, he is in a position to offer me this product. And maybe then another. And then another. We already know how TikTok or any news channel works : they serve us what we want before we know we want it.

Fortunately, I managed to leave the store in time, and I will have time to watch the series. Since 2010, another type of algorithm has been spreading: the so-called generative algorithms . These allow the creation of new realities based on the experience of many network users and other sources of information.

It is well known that books , movies and even hoaxes and lies can be created in this way.

In these four possibilities we have seen different types of algorithms. But now we should answer the question: what are they used for?

It’s clear that at the end of the day I’ve had some satisfaction: I’ve arrived at my destination faster and the bus knew where I needed to go. The supermarket served me immediately and I’ve watched my series. But what’s the price?

Give up deciding

Some studies have conducted experiments that show that one of the functions of algorithms is to decide for us. Deciding stresses us out , so we often prefer to delegate that task. This is dangerous because the tools that seem free are actually intended to make money for the companies behind them. So they are a form of advertising without ads, and the algorithm serves precisely that purpose, to make us buy more.

There is also another dark area: if we get satisfaction with little effort, we are sure to fall into a loop. The more they give us, the more we want. But this is no longer so utopian, because it causes addiction. A famous experiment with rats proved the power that a pleasurable stimulus can have.

So when we want to know what algorithms are for, we should also ask ourselves: which ones are for me and which ones for others? It might be worth considering who the rat in the experiment is before swiping.

Author Bio: Alejandro Cervantes Rovira is University professor in Artificial Intelligence and Machine Learning at UNIR – International University of La Rioja