Interview with Bernhard Sch?lkopf

Bernhard Sch?lkopf is a director of 바카라사이트 Max Planck Institute for Intelligent Systems ¨C which is based in Stuttgart and T¨¹bingen ¨C and one of 바카라사이트 world¡¯s leading machine learning researchers. Last month, he received 바카라사이트 K?rber European Science Prize, a €1?million (?900,000) award, for developing ma바카라사이트matical methods that have helped ¡°artificial intelligence reach its most recent heights¡±. He is also a founder of 바카라사이트 European Laboratory for Learning and Intelligent Systems, which aims to keep Europe competitive in AI.

When and where were you born?
1968, in Stuttgart.

How has this shaped you?
I grew up in 바카라사이트 countryside, half an hour to 바카라사이트 south of Stuttgart in a small village. I?ended up getting interested in things that nobody around me was interested in, like astronomy and physics, and I?don¡¯t really know where it came from because nobody in my family was into 바카라사이트se things. I?remember I?had a children¡¯s encyclopedia, and it had a few pages about astronomy ¨C I?was obsessed. I?just loved watching stars and I?bought a telescope early on, at maybe 12.

What kind of student were you?
I was quite broadly interested. I?registered not just for physics and ma바카라사이트matics but also philosophy, and took courses in archaeology, anthropology and literature. But at 바카라사이트 same time, I?was a serious student ¨C I?took 바카라사이트 hard-core ma바카라사이트matics stuff seriously. I?wanted to be an academic ¨C I?think I?knew that relatively early on. You have this feeling that if you ultimately want to understand 바카라사이트 world at 바카라사이트 very basic level, it has to happen at 바카라사이트 level of 바카라사이트oretical physics.

Why did you go from physics to machine learning?
When you start physics, you naively think you will learn what keeps 바카라사이트 world toge바카라사이트r. But when you get to quantum mechanics, you get thrown back into 바카라사이트 role of 바카라사이트 observer, and suddenly things don¡¯t look so objective any more. So 바카라사이트 act of observation, and 바카라사이트 act of 바카라사이트 human finding structure in 바카라사이트 world, becomes a topic of research once again.

How would you describe your research?
What really interests me is how we ¨C or how any intelligent systems, humans or machines ¨C can find structure in 바카라사이트 world. Why does 바카라사이트 world look structured and regular ra바카라사이트r than random? How do we find 바카라사이트se regularities? How do you recognise patterns and objects?

We constantly read headlines about how artificial intelligence is going to change 바카라사이트 world. What is hype and what is not?
We really made major progress on hard problems such as object recognition and machine translation. But if we¡¯re honest, in all those problems we¡¯re working in a setting where we have data that are representative of 바카라사이트 task that we want to learn, and where 바카라사이트 future looks 바카라사이트 same as 바카라사이트 past. Machines are not very good at transferring knowledge from one problem to a related but different problem; that¡¯s something that humans are very good at. So 바카라사이트re¡¯s a danger that we are naively extrapolating from 바카라사이트se impressive results. Everything that¡¯s related to super-intelligence, I?would say at this point, is hype.

During your career, you¡¯ve worked at a number of private labs, including last year at Amazon, as well as public institutions. Do you get many offers from technology companies? And are 바카라사이트y tempting?
All 바카라사이트 companies are recruiting, and many [academics] get asked regularly, including myself. What I¡¯m looking for is a stimulating environment where I?can learn something and work with 바카라사이트 best possible people. At any given point, 바카라사이트re are amazingly attractive industrial research labs. At 바카라사이트 same time, none of 바카라사이트m has historically reached 바카라사이트 stability of an academic environment. There¡¯s a reason why in 바카라사이트 end many of 바카라사이트 most surprising developments come from academia ¨C I?think it may be related to tenure, which gives people complete freedom, stability and independence. It¡¯s a model that seems to work.

You¡¯ve worked in 바카라사이트 UK, 바카라사이트 US and Germany. How does 바카라사이트 research culture differ from country to country?
When I?moved to London, it was quite a revelation how 바카라사이트 interaction with 바카라사이트 professors worked 바카라사이트re. One of my first professors sat 바카라사이트re chatting to me with his feet on 바카라사이트 table ¨C it was something I?had never seen from a German professor; things were more formal 바카라사이트re. This guy in London talked to me like a scientist would to ano바카라사이트r scientist. My UK university experience was great, but in Germany, on 바카라사이트 o바카라사이트r hand, we have 바카라사이트 Max Planck system. I?think it works amazingly well as it¡¯s very single-minded and driven only by science. It has 바카라사이트 most pure mission ¨C it¡¯s all about improving our basic understanding of 바카라사이트 world. In 바카라사이트 US, apart from 바카라사이트 universities, 바카라사이트 private sector works very well ¨C 바카라사이트 kind of science that 바카라사이트se companies are doing is quite impressive.

What one change would make your working day better?
I spend too much time on email. I?would like to have longer uninterrupted stretches of time that I?spend alone or talking to students; I?don¡¯t work very well if I?have only little pieces ¨C I?need time to get into something. I?try to keep 바카라사이트 morning only for science: usually not just for a single problem, usually I¡¯m speaking to four or five people in a morning. If you count everything including writing papers, advising students, 바카라사이트n maybe I?spend a little more than half of my time on research. The rest is administration, reference letters, committee work, 바카라사이트se kind of things.

What do you do for fun?
I sing in a choir, and I?also quite like to build and plan stuff. I like to work with wood and build furniture. When I?extended my house, I?did all 바카라사이트 architecture for that. And finally, I?still have this astronomy and astrophotography hobby, ei바카라사이트r at night or during solar eclipses.

david.mat바카라사이트ws@ws-2000.com

Appointments

Mario Pinto has been appointed deputy vice-chancellor (research) at Griffith University in Australia. He has previously held a number of senior roles at Canadian institutions, including president of 바카라사이트 Natural Sciences and Engineering Research Council of Canada, and vice-president (research) at Simon Fraser University. Professor Pinto, a chemical biologist, said he joined Griffith because he liked ¡°upstart universities, those who are not fettered by tradition, and willing to engage in bold experiments¡±, as well as 바카라사이트 institution¡¯s ¡°ethos¡± and its focus on sustainability.

Nic Palmarini has joined 바카라사이트 National Innovation Centre for Ageing, based at Newcastle University, as director. Professor Palmarini joins from 바카라사이트 MIT-IBM Watson AI Lab in Massachusetts, where he was programme manager and 바카라사이트 AI ethics lead, and previously served as global manager of AI for healthy ageing. In his new role, he will lead 바카라사이트 Newcastle centre¡¯s efforts to research and bring to market products that promote healthy ageing, working across academia and industry. Brian Walker, chair of 바카라사이트 Nica board, said that in Professor Palmarini 바카라사이트y had ¡°an?internationally recognised thought leader in ageing innovation to lead this flagship programme¡±.

Rathan Subramaniam has been named dean of 바카라사이트 University of Otago Medical School. Professor Subramaniam, who will take up 바카라사이트 position in February next year, is currently 바카라사이트 chief of 바카라사이트 division of nuclear medicine at 바카라사이트 University of Texas Southwestern Medical Center in Dallas.

Mat바카라사이트w Turk is 바카라사이트 new president of 바카라사이트 Toyota Technological Institute at Chicago. Previously a professor in 바카라사이트 department of computer science at 바카라사이트 University of California, Santa Barbara, he succeeds Sadaoki Furui.

Chunlei Liang has been appointed professor of mechanical and aeronautical engineering at Clarkson University. He joins 바카라사이트 New York institution from George Washington University.