Cristian Canton: "We want sovereignty over chip manufacturing."

BarcelonaCristian Canton (Terrassa, 1980), who was previously director of artificial intelligence (AI) at Meta's US headquarters, will take over as the new associate director of the Barcelona Supercomputing Center (BSC) in the coming days. He will succeed Josep Martorell, who is leaving the center after nine years. Canton is leaving after more than 15 years at large corporations such as Meta and Microsoft to join the BSC management team, where he will seek to transfer his experience to the private sector. In a conversation with ARA, he discusses this professional change and reviews the institution's future challenges.

You're an AI expert and have worked for many years at major tech giants. How can you apply this experience to the BSC?

— On the one hand, there's the entire technological aspect. I've learned new methods of management, technological development, and knowledge transfer that I can bring here. On the other hand, I've incorporated a very American way of doing things, typical of the private sector, of setting ambitious goals and finding ways to achieve them. I'm not going to incorporate this approach 100%, because it probably wouldn't work completely, but it's a vision that can be adapted to the BSC. We are one of the most important centers in Europe, and we will work to advance science.

Given your track record, should we expect more public-private partnerships in the coming years?

— I hope so. A good part of the impact we want to generate from the BSC involves transferring technology, that's why we have the AI Factory project, which should accelerate the access to AI in the business community. The BSC has already created many spin-offs and collaborates with large corporations; we'll see if we can have more partners; we're studying it. There are some very interesting companies; we can learn from them, and they from us. There will certainly be opportunities we should explore, and there are many avenues for collaboration.

Under your leadership, what do you hope to achieve at BSC?

— The transformation of the center into an AI center. Until now, AI at BSC has been highly fragmented into different research areas. We have people using AI to design chips and people using AI to mitigate climate change. This can lead to inefficiencies and duplication. A transformation is necessary. We'll have time; we'll have to think it through carefully, because these will require conceptual and cultural changes, but I hope to implement them as soon as possible.

What can we do with AI and what challenges do we face?

— AI allows us to be much more efficient in some tasks that until now were very complicated. For example, it can be a tool for doctors to diagnose diseases or to study pollution in Barcelona much more precisely. AI can be used to train a group of people with the data they're trained on. Here we have the AINA project and other initiatives to create a sovereign, Catalan, European AI, with the data we want and that represents our values and our culture.

There are voices that say Europe has fallen behind in this race.

— There's this tendency to always be first in everything. It's true that the models of Google, Facebook, and other companies are monstrously large, have a lot of resources, and are used massively. However, I think Europe has something very important: a very interesting regulatory framework. These regulations will allow us to implement responsible and inclusive AI. Elsewhere in the world, they're more lax about moving fast and winning this race, and sometimes it's not about winning, but about doing things the right way. If you take a little longer to get there and do it right, I think that's the way to do it.

The director of BSC, Mateo Valero, has repeatedly said that his dream is to turn Barcelona into the city of chip design and that already exists. the DARE projectAre you tasked with further exploring this area?

— We want sovereignty over chip manufacturing; it's important. Right now, the entire world is heavily dependent on large chip manufacturers, and if we had our own chips, we would end that dependence. So, we must continue down this path, yes, because having high-powered chips will allow us to train larger AI models and conduct deeper experiments. I would like Barcelona to be the Silicon Valley of Europe, and we will be if we develop chips, pioneer AI, and are also able to transfer it to society and ensure it has a return.

Will the US take advantage of the situation to attract research talent dissatisfied with the policies of the Donald Trump administration?

— We collaborate with many countries, including the United States. The BSC is a place where doors are always open. Perhaps some people will choose us as a center for their doctoral studies instead of the United States, but I believe we have enough prestige and are attractive enough to avoid having to actively seek out professionals. If this scenario allows us to incorporate talent that other countries reject for the wrong reasons, they will be welcome.

What legacy do you want to leave at BSC?

— I had to leave for the United States in 2009, but I would have stayed. I left because the professional opportunity was there. I'd like people to see in a few years that Catalonia has become a place of opportunity, that it's not worth leaving because it's the best possible place; to create an environment to be leaders in many areas of knowledge. We now have MareNostrum 5, which is one of the largest in Europe in supercomputing, and we're already thinking about MareNostrum 6, which we should start planning soon. We have options, resources, and talent. It will be a very long road, but I think it's possible. Otherwise, I wouldn't have come.