"AI for a good cause!"

He is one of the international stars of German science: Wolfram Burgard, one of the world's leading specialists in artificial intelligence and robotics. In the WELT interview, the winner of the "German AI Prize 2022" talks about the global fight for the best AI talents, about his path from Freiburg via Silicon Valley to Nuremberg and a horrifying experience of his parents.

WORLD: Mr. Burgard, you set up one of the leading chairs for artificial intelligence (AI) at the University of Freiburg before moving to the Toyota Research Institute in Silicon Valley in 2019 to work on the further development of autonomous driving. Your impression at the time: "In Germany, people are clearly lagging behind in the field of AI." What was your criticism based on?

Wolfram Burgard: She was referring to the high-tech companies there. If you think about the technology behind cell phones, search engines or self-driving cars, the US has a significant lead. We have to do a lot in Germany to be able to keep up here in the future.

WORLD: Are we doing enough, could the gap to the leading power USA be reduced?

Wolfram Burgard: I don't think Germany has fallen further behind. A lot has happened for us. Many new AI professorships have been advertised nationwide. Bavaria also started the high-tech agenda in 2019 with an investment volume of around 3.5 billion euros, part of which is used to promote AI. The Alexander von Humboldt Foundation has also created numerous AI professorships. Of course, many measures do not take effect that quickly, but they give us the chance at least not to fall behind.

WORLD: German companies are desperately looking for AI experts, but have little chance of success with the best minds. One of the reasons: top talent who go to Google, Amazon or Apple, which are opening more and more branches in Europe, after completing their doctorate, can earn up to 300,000 euros a year as a starting salary; one hears of interns who collect 10,000 euros a month. What consequences does this have for the economy and technological sovereignty in Germany?

Wolfram Burgard: There is a risk that the intellectual property of technologies that – like AI – increase the productivity of an economy will increasingly end up with foreign companies. Since the added value is based to a considerable extent on rights of this intellectual property, this would entail a loss of economic power. But I think many scientists and AI talents in Germany have now also recognized that Europe has a locational advantage in terms of quality of life. Of course life is very good in Silicon Valley. But many consider life in Germany to be even more pleasant. However, as the companies you mentioned are setting up more and more branches in Europe, it is becoming even more of a challenge to attract talented people to German companies. But I think that the big players in Germany also offer an attractive environment.

WORLD: You taught as a visiting professor in Pittsburgh and most recently worked in Silicon Valley – but you have now returned to Germany, where you took over the founding chairmanship of the Department of Engineering at the new TU Nuremberg in February. Wouldn't a professorship at Stanford University or in Berkeley, a move to Google or Amazon have been more appealing and lucrative?

Wolfram Burgard: I really enjoyed my time in the US, I had a great apartment, the weather is good, and I also had offers from American universities or from companies like Amazon. But you don't pursue your life plan alone, but together with your family. For example, my wife, who is a hotel manager in the Black Forest, would find it very difficult to persuade her to go to the USA. That was one reason for my return. The other was the chance to build something completely new here in Nuremberg, “from scratch”, as the Americans say. This is a unique opportunity. We feel a bit like a start-up, with all the freedom you have and all the ideas you can bring to the table. Designing this big thing together with colleagues, making it successful and also doing something for the region is great.

WORLD: The TU Nuremberg is the first newly founded state university in Bavaria since 1978. International stars of science are of great importance for the 1.2 billion euro prestige object in Markus Söder's hometown. But have you really never had any doubts as to whether development work in the German provinces is right for you? When you think of Bavarian cutting-edge research, you first think of Munich, don't you?

Wolfram Burgard: Bavaria does a lot more than other federal states and with the high-tech agenda I mentioned and the top professorship program, excellent conditions have been created for top-class research. A different wind is simply blowing in Bavaria! Of course, Munich is very high up in Germany with the Technical University and the Ludwig Maximilian University. But that doesn't mean that you can't ignite a small fire next to this big flame, which also develops luminosity and radiance. Our goal is to build an excellent team here, to put the TU Nuremberg on the AI ​​and robotics map and thus to strengthen industry in the region in the long term.

WORLD: The Israeli historian Yuval Noah Harari describes AI as "the most important technology of our time" and believes it is possible that the country with the best AI can control the whole world. Do such central questions, where the position of man in the world is just as important as geostrategic dominance, play a role in your scientific work? Do they make you more cautious and thoughtful, or nervous and ambitious?

Wolfram Burgard: I'm a little inspired by the idea that also drives Silicon Valley: namely "AI for good". The potential benefits of AI are so great that if we want to make the world a better place, we have no choice but to work on it. AI offers so many possibilities, from road safety to robots that do tedious work for humans to the democratization of medicine. As mentioned at the beginning, this is accompanied by questions of technological sovereignty. That is why it is so important that we invest heavily in the development and research of AI in Europe. Of course, it is not enough to just create professorships, we also have to build ecosystems that facilitate and accelerate the transfer of knowledge from universities to companies. This is the only way we will be able to build technologically leading companies in Europe that are competitive worldwide.

WORLD: Nicolas Chaillan, former software chief of the US Air Force in the Pentagon, said in the “Financial Times” that when it comes to the use of AI, political decision-makers are slowed down by ethical issues, with the result that China the innovative leadership in this field. Do you also see it this way: Is the West being held back by ethical issues?

Wolfram Burgard: Sometimes we might be overly critical and overcautious. Just look at how much of our lives we spend managing or blocking cookies. Data transfer between countries is also complicated, and data cannot simply be made anonymous: if all faces have a black bar over their eyes, for example, the system cannot learn properly because it thinks that everything with a black bar over their eyes has, must be a human. Nothing useful comes out of this. As a result, certain AI systems don't work well in Europe because you didn't have European images to train them. Certainly, security and privacy must be taken seriously and guaranteed. But sometimes it's difficult to make the necessary progress because privacy makes it very difficult for us.

WORLD: Is the impression deceptive, or has the development of AI actually gained significant momentum in the past year or two?

Wolfram Burgard: I have to honestly say it's fascinating: I've never experienced a phase like today myself. Computer science and robotics have always been relatively fast in terms of development progress. But what we have experienced in recent years is significantly more than I expected. We're currently experimenting with voice systems to use them for robotic navigation, and it's really impressive what you can do with them today. Of course, technological advances are frightening because we all don't know how they affect us. I see it positively and believe that the world is getting better.

WORLD: Your specialty is the navigation and control of mobile robots. Dare to take a look into the future: When will the high goal of future mobility be achieved – autonomous driving?

Wolfram Burgard: Last week my in-laws called me from Kelheim and told me that they had been traveling there in a self-driving minibus. That would have been a bit "scary" in modern German, but actually everything worked well. There you can see the so-called Level 4, automated driving in a restricted area. I think we're going to have to wait a very long time for Level 5, real autonomous driving around the world.

Prof. Dr. Wolfram Burgard, founding chair of the Department of Engineering at the TU Nuremberg, is one of the most decorated computer scientists in Europe and the most internationally cited expert in the field of machine learning. In 2009, the 61-year-old was awarded the Leibniz Prize, which is one of the world's most renowned research prizes. He also received the Advanced Grant from the European Research Council and the Technical Field Award, and is also a member of the Heidelberg Academy of Sciences, the Leopoldina and a Fellow of the Association for the Advancement of Artificial Intelligence and President of the IEEE Robotics and Automation Society. Together with Sebastian Thrun and Dieter Fox he wrote the standard work "Probabilistic Robotics". Wolfram Burgard is married to a Black Forest hotel manager, has a son (he works at the nuclear research center CERN) and lives near Freiburg and in Nuremberg.