Data science and artificial intelligence, are becoming increasingly important in democracies

the Data science and the Artificial intelligence They played an increasingly visible role in recent democratic elections. Using technology to detect supporters of a political party, manipulate them and influence their vote is an increasingly common practice. What was at one point a method used in secret, is now being applied in a more visible and exposed manner.

This is how the researcher from the University of Zaragoza, Jorge Gracia del Rio, puts it in his article. “Algorithms that will choose the next prime minister”was originally published in Conversation And its replication in various media at the international level.

Garcia suggests that the use of data science and artificial intelligence should lead to a discussion that requires going beyond the technical level, to move it into the ethical and moral realm, because it questions the practice of democracy As we know it so far.

The specialist refers to the role of data science and in particular tools such as natural language processing (NLP) or psychometric studies of Kosinsky that allow user profile identification through sentiment analysis (sentiment analysis) to influence their political position through biased and even false information, which in turn has a direct impact on the election results.

The author notes that “…one of the first successful use cases of big data technologies and social network analysis to adjust the electoral campaign was that which Barack Obama did for the United States presidential election in 2012. In his campaign (and in many others after that), traditional polls of voting intention, based on phone calls or in-person interviews, combined with network analysis Social “.

He points out that these are cheap methods that can be applied in near real time to know and gauge voter opinion. For this purpose, natural language processing techniques are applied, especially those related to sentiment analysis. Gracia del Río explains that these technologies analyze messages sent by thousands or millions in blogs, tweets, and so on. It seeks to measure the positive or negative features of opinions expressed in relation to a particular politician or electoral political message. The author identifies limitations in the process, because these are measurements with a biased sample, since the most active users on social networks tend to be young and tech-savvy, i.e. most of the population is not represented, and therefore Its scope is limited when it comes to predicting election results, but it is very useful for “studying voting trends and the state of people’s opinion”.

In the text, the author refers to what happened in the election in which Donald Trump won the presidency of the United States and the role that the company played in that process. Cambridge Analytica

“Big data and psychological profiles have a lot to do with the victory that polls fail to predict. It wasn’t mass manipulation, but different voters received different messages based on predictions about their vulnerability to different arguments, receiving information that was biased, fragmented, and sometimes contradictory to other messages from the candidate. assignment to Cambridge Analyticawhich was involved in controversy over the unauthorized collection of information about millions of Facebook users,” says the author.

method Cambride AnalyticaDemonstrates, on the basis of psychometric studies of Kosinskywith which it was possible to check “… how can you accurately get a user profile with a limited number of likes as if your family or friends had done it … The problem with this approach is not in the use of technology, but in the ‘secret’ nature campaigning, psychological manipulation of vulnerable voters through direct appeals to their sentiments, or the deliberate dissemination of fake news via bots.”

Gracia del Río encourages us to think broadly about this topic and highlights how important it is to discuss the scope of data science and, in particular, the use of algorithms that can influence a country’s democratic future.

According to him Edelman Trust Barometer 2023Social networks can act as a force for growthpolarization In the world, because far from creating a space for common conflict resolution, it becomes a catalyst for conflict itself.

Social networks have become a place of mistrust, and this may be because they can be manipulated much more than traditional media and search engines, where there is more trust according to the study.

The same study shows that Republicans in the United States are more likely to believe the differences between Republicans and Democrats are irreconcilable. This illustrates the way the algorithm allowed for the creation of a deeper division between supporters of different political parties and a larger gap between each other’s beliefs.

Article He reveals that other than the fact that technology can determine the “next head of government”, the time may come when artificial intelligence will rule us. Gives an example of the case of Denmark with the establishment of artificial party which is driven by artificial intelligence. This was intended to represent the “20% of the Danish population who do not go to the polls” in parliament.

In the face of a possibility approaching more and more, moral and ethical questions enter again, because as Gracía del Río states, machines are manufactured and controlled by humans (at least for now) who are still subject to corruption even though the machine is not there. itself and does not allow us to understand the logic that was implemented to generate a particular response, which is why, in the words of García del Rio, it does not allow “interpretation”.

Therefore, the role of data science and artificial intelligence in democracies is increasing.