Interview with Dr. Petar Veličković, a member of the Advisory Board

In early December, we received news that the world's leading scientific journal Nature published on its cover the research paper on the use of AI in representation theory and knot theory.
For the first time, and thanks to machine learning methods, mathematicians have gained new insights in this field, and among the three main authors of the research is a member of the Advisory Board of the Institute for Artificial Intelligence, Dr. Petar Veličković.

"My research ideal is to develop machine learning algorithms that are theoretically well-grounded, elegant to implement, and most importantly, useful to the relevant field experts. With our recently published work on the use of AI in pure mathematics, I would say that we have successfully achieved all these goals. Our models are faithful to the mathematical objects on which they operate, we have made all of our source code publicly available, and the mathematicians with whom we have collaborated have expressed great delight with the tools we have developed. The contribution of our tools has been crucial enough that we are credited with discovering new mathematical results, even though most of our efforts went into developing machine learning models," says Dr. Veličković.

Dr. Petar Veličković is a Staff Research Scientist at DeepMind, Affiliated Lecturer at the University of Cambridge and Associate of Clare Hall, Cambridge. He holds a PhD in Computer Science from the University of Cambridge (Trinity College). He is an approved ELLIS Scholar in the Geometric Deep Learning Programme and the first author of Graph Attention Networks and Deep Graph Infomax.

For publication, the young Serbian scientist adds:
"For all the reasons I have just listed, I consider this paper to be one of the greatest achievements of my scientific career so far. The fact that our work has been recognised by the journal Nature, where it graces the cover, only makes the feeling even more fantastic! But as with all my previous successes, I see this primarily as a stepping stone for what's coming next, i.e. a motivator to keep the momentum in the future as well."

The love of science is the crucial factor for outstanding scientific results.

"At least in my experience, success requires, above all, perseverance and a true passion for your field of study. In science - perhaps most especially in our recent mathematical project - we often not only do not have the 'right answers', we do not have the right questions, and the results we collect often leave us with more questions than we had before collecting those results. In these situations, it is very easy to give up on the project or look for ways of higher "certainty". It is precisely at such moments this passion becomes invaluable, and without the true desire to expand human knowledge, any other external motivation of a researcher could become meaningless," says Dr. Veličković.

Dr. Petar Veličković's list of informal interests is also very interesting. When he is doing research, he listens mostly to electronic and downtempo music. He goes to concerts and he is interested in cooking. His favourite TV series is Mr. Robot because it combines a great story, characters, well-chosen songs and scenes that represents modern cybersecurity threats.

"The best way to understand something is to explain it to someone else," is Veličković's guiding principle.

"Guided by this quote, I have been actively sharing my knowledge with the broader community, both within Serbia (where, for example, I was one of the initiators of the Computer Science Week at the Mathematical Grammar School; a computer science lecture series for gifted high school students) and abroad. Most recently, I was part of a team of lecturers that elucidated the concepts of geometric deep learning to talented African students as part of the African Master's in Machine Intelligence."

His motto in life is a verse from the song 'Outro' by the band M83: "Facing tempests of dust, I'll fight until the end".

"This quote perfectly combines several aspects of my personality and philosophy. For me, it represents the desire to persevere in whatever I decide to tackle, regardless of the circumstances," adds the 28-year-old scientist.

Although he has never formally studied, he is very interested in aerodynamics. He also follows Formula 1 races and spends his free time with his family and close friends.

"I have been very fortunate to meet some truly outstanding people during my education. Their support, as well as the support of my family, is probably the main reason for my success today," concludes Dr. Veličković.