Text written in Basque and translated automatically by Elia without any subsequent editing. SEE ORIGINAL

It makes no sense to separate science and letters in this way

Physics and mathematics, language and artificial intelligence, singing food and scientific dissemination, all converge in the trajectory of Ixak Sarasua Antero. He recalls that he has always been very curious and that books and devices related to the sciences were common in the house. He was also well acquainted with these subjects at school and, when choosing university studies, did not hesitate much.

“I decided to study physics because I thought it would give me the key to understanding the basic systems of the universe. I also loved quantum, cosmology and so on, and then I saw that you also work on other skills. For example, studying physics gives you some tools for mathematics and logical reasoning that are also useful in other fields,” he recalls.

He also thought that maybe he had to study mathematics, but he was happy with his classmates and made great friends. However, at the end of the degree, he did not want to study physics. “It’s very sacrificed, and when I did my final degree work, I saw clearly that I preferred to expand it to something else. So, although many in the class went on to do the Master in Quantum, I chose the Master in Modelling and Mathematical Research, Statistics and Computing.”

“In the end, data is used everywhere: in physics, in hospitals, even in language.”

He explains that in the Master they had subjects from different areas (cryptography, statistics, programming); therefore, it can be applied to a wide variety of works. “Now I’m in language technologies, and if I had known this from the beginning, I would have done a master’s degree in language processing.” In any case, after completing the master’s degree, the desire to carry out the thesis linked to the Basque language arose.

Sarasua, therefore, renounces the distinction between sciences and letters that the system usually makes. “There’s no point in separating like that. It is true that when I was studying the Physics degree I did not imagine that I would end up in the Orai artificial intelligence center, promoted by Elhuyar, but the Master opened up a lot to me. In the end, the data is used everywhere: in physics, in hospitals, even in language,” he explains.

The thesis is currently investigating the adaptation of large language models (PLL) to Basque and other languages with few resources. He warns, of course, that these languages are not all at the same level; many have less written texts than the Basque language. But they have the same difficulties compared to hegemonic languages. “Training HEHs requires a lot of data; in our case, a lot of text. This is one of the main difficulties.”

He also points out that this area of research is evolving very quickly and that, compared to other large languages, there are few researchers in low-income languages. Despite the difficulties, he is enthusiastic about his doctoral thesis.

Not everything is a thesis

Although the research requires a lot of time, it also takes time for other activities. Among them is music. And there it also shows that it does not fit into rigid limits. He plays more than one instrument, belongs to groups of different styles and enjoys singing meals as much as he does experimenting. He also works in the colonies, in the commission of txosnas of youth movements and in the association of Euskera.

“But with the language and the data, yes, I think so, I will continue.”

Another work he does voluntarily comes from the hand of a group of friends of his student time. In fact, some of these people have gone abroad to study or investigate, and it occurred to them that it would be good to report their topics, in Basque, on a website. They created the Asteroidea website. They were invited to write Sarasua, who immediately nodded. “But we have started very informally, with the intention of writing in Basque and in free style what we have learned,” he says.

In any case, collaborations have begun in some popular media and who knows if any of them will not be encouraged to be scientific disseminators. At the moment, Sarasua confesses that he is not clear about what he will want to do in the short to medium term (five years, for example). “I don’t think I’m going back to physics. Once I leave, it takes a lot of work to get back on track, and I don't see myself doing that. Artificial intelligence is also going very fast and there is a lot of noise around this concept. I don't know. But with the language and the data, yes; in that I think yes, I will continue. This is a great place to work,” he concluded.

Ixak Sarasua Antero was born in Aretxabaleta in 1999. After completing the Degree in Physics at the UPV/EHU, he has completed the Master's Degree in Modelling and Mathematical Research, Statistics and Computing at the same university. He is currently working on his doctorate at the Orai Centre.