Editorial of July 2018


 by Alessandra Silveira, Editor 
 and Sophie Perez Fernandes, Junior Editor

Artificial intelligence and fundamental rights: the problem of regulation aimed at avoiding algorithmic discrimination

The scandal involving Facebook and Cambridge Analytica (a private company for data analysis and strategic communication) raises, among others, the problem of regulating learning algorithms. And the problem lies above all in the fact that there is no necessary connection between intelligence and free will. Unlike human beings, algorithms do not have a will of their own, they serve the goals that are set for them. Though spectacular, artificial intelligence bears little resemblance to the mental processes of humans – as the Portuguese neuroscientist António Damásio, Professor at the University of Southern California, brilliantly explains[i]. To this extent, not all impacts of artificial intelligence are easily regulated or translated into legislation – and so traditional regulation might not work[ii].

In a study dedicated to explaining why data (including personal data) are at the basis of the Machine-Learning Revolution – and to what extent artificial intelligence is reconfiguring science, business, and politics – another Portuguese scientist, Pedro Domingos, Professor in the Department of Computer Science and Engineering at the University of Washington, explains that the problem that defines the digital age is the following: how do we find each other? This applies to both producers and consumers – who need to establish a connection before any transaction happens –, but also to anyone looking for a job or a romantic partner. Computers allowed the existence of the Internet – and the Internet created a flood of data and the problem of limitless choice. Now, machine learning uses this infinity of data to help solve the limitless choice problem. Netflix may have 100,000 DVD titles in stock, but if customers cannot find the ones they like, they will end up choosing the hits; so, Netflix uses a learning algorithm that identifies customer tastes and recommends DVDs. Simple as that, explains the Author[iii].
Continue reading “Editorial of July 2018”