The most worrying examples are not those of the beginning of this article that we are all more than aware of. It is the algorithms and not the hardware. It is relatively simple to detect how hardware affects us, therefore it is “easy” to solve. Algorithms, however, the most representative product of these decades, worry me more.
We are still working with biased data because we have not solved the previous step and, despite this, algorithms are, little by little, marking our interactions and entering all the areas of our lives. There does not seem to be much understanding among the people who code these algorithms about the problems with the data on which they are based. Artificial intelligence ranges from voice recognition systems that do not recognize female voices (there were always authorities), to online dictionaries, or algorithms that decide whether a specific CV will ever be seen by human eyes.
Limitation of opportunities, that is what this article is about.
— Sara Antolín.