More than 100 people attended the first version of Data, Beer & Pizza at IMFD
Data (lots of data), pizza and beer were the essential ingredients of the first Data, Beer & Pizza, a meeting that seeks - in a relaxed atmosphere of camaraderie - to disseminate the scope of data science in its most diverse fields of application. The meeting, which was attended by more than 100 people, was organized by the Millennium Institute Foundational Research on Data (IMFD) with the support of the External Environment Outreach program of the Millennium Science Initiative and took place in Galería Mackenna, in downtown Santiago.
How do computers understand human beings? How does trolling manifest itself through social networks and with what type of topics does it arise the most? Are computers all-powerful? These were the three topics addressed in the talks given by IMFD researchers Jorge Perez, Magdalena Saldaña and Marcelo Arenas, respectively, in the first version of this event held last April 16, 2019.
How do computers understand us?
Jorge Pérez, associate researcher of the IMFD and academic of the Department of Computer Science (DCC) of the University of Chile, was in charge of opening the day with a playful talk where he explained how a computational system is trained to learn to join concepts and put together sentences like the ones we use every day in society, using what we believe is unique and distinctive of human beings: natural language.
Through examples, the researcher explained how computers, when learning to construct sentences using information generated by human beings, reproduce - without understanding its meaning - the same biases that society has around various concepts. As reported by the electronic media Emol, Perez showed the training of an artificial intelligence "through a work with vectors and analysis of words that were taught to a system, extracted from articles published in Wikipedia", according to the publication.
During the exercise, in which Jorge Pérez programmed "live" in front of the audience, it was shown how the computer associated words related to gender or politics, for example, with certain adjectives, reflecting a possible bias of the machine, acquired in its learning process. "A mission for scientists is to avoid teaching machines biases", said Jorge Perez in the talk, "after showing, with examples, how some words were contextualized according to the texts in which they were found to result in relations that clearly had discrimination problems", reports Emol.
Chileans and trolling in social networks
Magdalena Saldaña, an academic at the UC School of Communications and researcher at IMFD, addressed incivility in social networks -the academic concept for trolling- in her talk "Haters gonna hate".
The researcher commented that, internationally, it is estimated that one out of every five comments (that is, 20%) falls into the category of incivility. Now, the situation changes when we look at the display of opinions on social media networks in Chile: on average, in our country one out of every four comments falls into the category of incivility and if it is election time, 30% of the posts correspond to what is clearly distinguished as "trolling".
"The importance lies in the fact that uncivil comments have an effect on audiences: people become more intolerant, the credibility of the media declines and even affects the perception of the news," adds the IMFD researcher. She pointed out that the more uncivil comments there are in a news item, the more people tend to evaluate the information as negative, without having read it.
They are the all-powerful computers
Marcelo Arenas, director of the IMFD, was in charge of closing the day. With his talk entitled "Are computers all-powerful?", the full professor of the Department of Computer Science (DCC) of the Catholic University, refuted the belief that computers have reached a processing capacity for which there are no limits.
The researcher explained that, although in some aspects computers may be surpassing human beings, this does not make them all-powerful or infallible. In his presentation he stated that there are physical limits to the capabilities of computers, imposed by the hardware.
To demonstrate this, he used some examples of computational logic problems whose solution, if processed by the most powerful computers imaginable, would take longer to process than the universe has existed. "And that is something we should be grateful for, because if a computer could solve some of these problems, everything that science has developed in cryptography and that is key to the security of online financial transactions would be useless," concluded Arenas.
Data, Beer & Pizza is part of a series of science outreach talks aimed at the general public, which will be held during 2019 and will be communicated through the IMFD website and its social networks.