Josep Maria Martorell: "We are in the hottest moment of scientific research in history."

There's a crossover between technology, science, and artificial intelligence that could revolutionize medicine, research, and other things we probably can't even imagine today. One of the people most familiar with this intersection is Josep Maria Martorell. He has been Director General of Research, Associate Director of the Barcelona Supercomputing Center, and for months has been a partner at the investment firm Invivo Partners, where he focuses on investing in artificial intelligence projects.

The Google CEO was asked what artificial intelligence means, and he compared it to the appearance of fire. Do you agree?

— I'd be a bit cautious and say it's comparable to the emergence of the internet, which has changed the way we live. And if we think about it, nothing we do every hour of the day would be possible without the internet; we'd do it differently. I believe that AI, which is still a technology, will also transform everything. How we live, how we work... everything.

Let's talk about power. Who owns AI?

— Basically, in the hands of two major world powers, the United States and China. We've moved from geopolitics to geotechnology. Until now, countries argued, and unfortunately continue to argue, over borders or raw materials. Now there's a third playing field: technology.

So, is AI the new oil?

— The components that make AI possible are the new oil. AI depends on three things: algorithms, data, and computing. Behind the algorithms is talent. And the global battle for talent is exceptional, and it's a battle that we in Europe can still play well. The second issue is data; it's even more brutal, so it's important to have good regulation. And the third area, which we're sure to lose here, as Europeans, is computing, that is, chips.

Why did we lose him?

— We don't have factories to make these chips, nor do we have large companies to design the components. We can remedy this in the coming years, because there's a resounding European commitment; the question is whether we'll be fast enough.

The most important manufacturing continues in Taiwan, right?

— Yes, and it's outrageous. 90% is still there. But if we're talking about high-precision chips, the United States has really stepped up its game and is making exceptional public investments, and we'll see the fruits of this in the coming years. In Europe, there's no way to attract a high-precision, highly complex chip factory project.

Because?

— We're talking about something extremely complicated. Just so people understand, we're talking about a significant investment, €20 billion, no joke. From the day you say "let's get started" until the day you release the first chip, it may take 10 years.

But not having it leaves you with a sovereignty problem.

— Especially in terms of design, not so much manufacturing, which ultimately goes through Taiwan, but also through Korea or the United States, and an alliance strategy could be developed. But design is concentrated in the hands of very few companies, either Chinese, with all the security and regulatory issues, or American. The problem is the lack of sovereign European companies in chip design. And the investment for this is much lower than building a factory. There are interesting initiatives here; there are some in Catalonia. It's too early to say anything, but perhaps in two, three, or four years, Europe could put some company in a relevant position globally.

This century's struggle for hegemony is between China and the United States. Whoever wins the technological battle will have won the battle?

— I'm not sure there's a winner. I mean, we have a lot of this idea that winner takes all, because we came from a very monopolar world, where the United States ruled and the rest followed suit. Now the feeling is that we're moving towards a multipolar world, where there are different territories, and in each of them, one of those major players with a lot of influence. Not one that ends up taking everything.

You claim that opportunity lies in talent.

— We have a great opportunity. First, Europe has nothing to envy anywhere else in the world in terms of scientific and technological talent. We are helped by the geopolitical situation and the way some countries around the world are closing in on themselves.

Is Donald Trump helping us?

— Well, every time the US government decides to make life more difficult for people who want a visa, people look for places where obtaining a visa is easier.

And are you noticing that many more people are coming?

— Absolutely. I think any academic institution has seen a very strong change in the last year. Especially with European expatriates there who are deciding to return, and Asians and Latin Americans, who had a very strong natural tendency to go to the United States, and now may be more interested in Europe. Both trends, combined, I believe offer us a spectacular and unique opportunity.

And is Barcelona a good place for them?

— Where would you have lived before, Helsinki or Barcelona? Barcelona, with its salaries, can't compete with many places in Europe, but there are spectacular living conditions. It's not just the climate; there are many other important things. The city's cultural life, a decent education system, a healthcare system that's unbelievably free and universally accessible... The combination of all this puts us in the top five attractive places in Europe, of course.

But are we talking about people who used to go to Harvard or Columbia and now come to Barcelona?

— Yes, because the political situation there is unattractive, and also for another important reason: we are predictable. We Europeans may be boring and slow, but we are predictable. That is, if a government says it's launching a scientific project, it does. The US government can decide on a Friday to change or cancel an entire program.

Do you think European governments understand what the battle is about?

— There's an awareness that science is a national asset, something that was unthinkable 20 or 25 years ago. And there's an awareness that this is the foundation for some of the future opportunities for the country and the economy. Later, we'll criticize whether they do it better or worse and whether they put in enough money or whatever. But 15 or 20 years ago, we had to debate why this was important, and today we can skip over it.

What does AI mean for research?

— A complete paradigm shift. I like to say that we are probably at the hottest moment in scientific research history. The massive use of artificial intelligence to interpret large volumes of research data allows you to do things that were previously impossible.

An example?

— The easiest example, which I'm not saying is simple to understand, is last year's Nobel Prize in Chemistry.

Let us remember this.

— It was given to three people who aren't chemists—most of them engineers—who don't work at a university, but work in a company, and they solved the problem of protein folding. The important thing is that it's a problem that humanity had been unable to solve. So AI doesn't just accelerate processes; it solves problems. That's why I believe that in the coming years we will see many scientific advances in all disciplines thanks to the intersection of the most advanced science and the widespread use of artificial intelligence.

And does this change the scientific method?

— This puts the scientific method in jeopardy. But continuing with the Nobel Prize example, there's an important question. The algorithm gets the answer right, but for us, solving a problem means finding the rules that explain why it happens that way. And AI doesn't do that. So I wouldn't say it changes the method, but it does put it in jeopardy. Until now, we've said: I have a theory, I go to the lab, and see if it works. Now I have algorithms that help me find new conclusions.

And could this mean ending animal testing in the future?

— I'd like to. We can already simulate human hearts. We can't completely save ourselves yet, but I think it will be possible in the future. And now let's talk about making things. in vitro or live, that is, in the laboratory or with animals, and we will move on to doing in silico. That is, computer simulations.

This means creating another "me", but with data?

— Imagine the power of this. When a company makes a medical device to put in your heart, it won't be standard anymore, it will be completely customized.

With so much data, there are risks, such as not being able to provide health insurance because they know you're at high risk for some disease.

— This is where regulation comes in. So, what does European regulation say, which is widely criticized? I think it's simply from people who haven't read it. European regulation says, look, you can regulate artificial intelligence technology based on size, how big or how powerful an AI is. You can develop whatever AI you want. Now, I'll regulate whatever use you make of it. Do you want to use AI to decide whether to grant credit or insurance? You can do so as a support tool. But the law tells you that the decision must always be explainable. So, can you use AI? Yes. Can you simply believe that AI is leaving many people in a vulnerable situation? No. I would say, generally speaking, that we're pretty well on track with this.

What is the limit of AI?

— There are three major limits. One has to do with investment capacity. Talent, data, and computing cost money. Another is natural resources, because managing this entire system consumes energy, and energy is limited. And the third, fundamental, is physical. Chips are printing at a width of 3 nanometers; a nanometer is one-millionth of a millimeter. We can't go much lower.