DeepLearning

Gitta Kutyniok

Gitta Kutyniok

Born in Bielefeld, Germany • Birth year 1972 · Studied Mathematics and Computer Science at University of Paderborn in Germany • Highest Degree Habilitation in Mathematics • Lives in Munich, Germany • Occupation Professor for Mathematical Foundations of Artificial Intelligence

I had never planned to become a professor of mathematics, and if someone had told me when I was young, I would have said: This is impossible. Due to my excitement for mathematics in school and the fact that my mother and my grandfather were both teachers, I first wanted to become a high school teacher myself. And this is how I then started my studies, choosing computer science as a minor. Although the change from high school mathematics to university mathematics was difficult and required a lot of hard work, I enjoyed my studies very much. I however could not get excited about didactics for high school teaching, hence I switched to diploma studies in mathematics. And since at the University of Paderborn, it was quite easy to pursue a diploma in computer science at the same time, I enrolled in this as well.

(…) In retrospect, this period trained me to follow my own path and be very independent.

In my last year, a professor working in abstract harmonic analysis approached me with an offer for a Ph.D. position. I was hesitant about whether this was the right career path for me. Eventually, I accepted the offer but quickly realized that not pure mathematics was my passion but applied mathematics. Hence, in agreement with my supervisor, I chose a more applied topic and got assigned a second supervisor in Munich. This arrangement was not optimal. However, in retrospect, this period trained me to follow my own path and be very independent.

One of the reviewers of my Ph.D. thesis then offered me a position as a Visiting Assistant Professor at the Georgia Institute of Technology. Since I was hesitant about what to do next, I embraced this opportunity, taking it as a chance to see whether I am good enough for continuing as a post-doc. My time as a Visiting Assistant Professor was again hard, since I had never taught a course before, and I now even needed to teach in English. But research-wise a whole new world opened to me; having now collaborators with similar interests as myself, namely the area of applied and computational harmonic analysis. I then spent another year in the US with a research fellowship at both Washington University in St. Louis and again at the Georgia Institute of Technology. It was a very productive time for me, leading to a Habilitation in Mathematics at the University of Giessen in Germany.

I overcame my shyness and approached [some professors in the US whose work I had always admired] for an invitation (…).

Due to the uncertainty of obtaining a professor position in Germany, I applied for a Heisenberg Fellowship from the German Research Foundation to visit some professors in the US, whose work I had always admired. I overcame my shyness and approached them for an invitation and eventually got the amazing chance to visit first Princeton University, then Stanford University, and finally, Yale University, learning about new research areas such as compressed sensing.

Returning to Germany, I started as a full professor at the University of Osnabrück. This was a very fulfilling experience, and I loved building up my own research group. However, it was a very small department, and finding good students was hard, and I soon started looking for other positions.

I was again lucky and was offered an Einstein Chair at the Technical University of Berlin. Soon after, the advent of deep learning started and affected my research area significantly. I decided to embrace this paradigm shift and delve research-wise into artificial intelligence. Looking back, this was one of the best decisions in my life.

For the first time, I am now not the only female professor in my department.

This might have also led to a personal offer from Ludwig-Maximilians-Universität München for a Bavarian AI Chair for Mathematical Foundations of Artificial Intelligence, which I was surprised and delighted to receive. Due to the excellent conditions for AI research in Munich and Bavaria, I accepted the offer and moved to Munich. For the first time, I am now not the only female professor in my department. In fact, I have several wonderful female colleagues, which is an entirely new experience for me.

In general, I learned in my career that one should be open to opportunities, as they often arise unexpectedly, and also not be shy to approach colleagues for advice and help. If you ask whether being a woman has impacted me in my career, I have to say that the first time I realized that one is treated differently was when I became a professor. As committee meetings increased, I learned the hard way that men do not behave better or worse, but just differently. Looking back, a course on gender-specific behaviors in professional environments, as it is, in fact, custom for higher positions in industry, would have helped significantly.  On the other hand, I also had and still have several amazing male colleagues who support me tremendously, also with advice, and I am deeply grateful to them.

Posted by HMS in Stories
Nicola Richmond

Nicola Richmond

Born in UK • Studied Mathematics and Computer Science in Edinburgh, UK • Highest Degree PhD in Algebra and Algebraic Geometry • Lives in London, UK • Occupation VP of AI

As a child, I enjoyed solving logic puzzles and spent a lot of time teaching myself BASIC on a Commodore VIC-20 that my dad had given to my brother for Christmas – my brother wasn’t remotely interested in the computer – I was obsessed by it!

My love for the problem-solving aspects of mathematics was solidified at school. I was lucky to have amazing mathematics teachers who made my learning journey both interesting and enriching. After regularly getting decent marks in school tests, I realised that I also had an aptitude for the subject and specialised early on by taking double mathematics A’ Levels.

(…) The inherent precision and rigour in mathematics helps keep my wandering mind constrained!

I went on to study mathematics as an undergraduate at Edinburgh. While there, I gravitated to pure mathematics – I love the logical nature of abstract mathematics and how concepts and rules can be linked together to develop new ideas and prove theorems – the inherent precision and rigour in mathematics helps keep my wandering mind constrained! I intended to pursue an academic career in mathematics, but with permanent academic positions in short supply, I settled on IT as a sensible Plan B and stayed on at Edinburgh to take an MSc in computer science. After that, I headed to Leeds to study for a PhD in representation theory of finite-dimensional algebras; and this was the end of my pure mathematics adventure – a career involving computing beckoned!

Looking back, there were several junctions along the road where I could have taken a different direction. The first was leaving my IT consultancy role to join Unilever on a two year contract. This introduced me to the world of chemoinformatics which I could link to mathematics by considering molecules as graphs of atoms connected by bonds. When my contract at Unilever came to an end, and with no sign of the recruitment freeze lifting, I decided to go to Sheffield as a post-doctoral researcher to work on developing a (commercialised) approach to facilitate computer-aided drug design.

Just over a decade was in the computational chemistry department, developing methods to find small molecules with medicinal properties.

Following the post-doc, I spent 18 years at GSK. Just over a decade was in the computational chemistry department, developing methods to find small molecules with medicinal properties. I then made an internal move to focus on bringing novel data analytics methods into GSK. This GSK chapter exposed me initially to the world of deep learning and its application to computer vision, and then later to new drug modalities, like antibodies, when I was responsible for a portfolio of digital, data and analytics projects.

The final four-year leg of my GSK journey I spent in the newly-formed AI/ML organisation. There, I learned the virtues of good engineering best practice and agile development, which was excellent preparation for my current role as VP of AI at BenevolentAI. I was also put in charge of building and leading the GSK.ai Fellowship Programme, which ignited a passion for developing, mentoring and nurturing junior staff members.

While I no longer have the opportunity to indulge in pure mathematics, mathematics is omnipresent in what I do.

Now at BenevolentAI, I focus on the company AI strategy and our centre of functional excellence in AI. While I no longer have the opportunity to indulge in pure mathematics, mathematics is omnipresent in what I do. I spend a lot of time reading the AI literature, which really combines probability theory, statistics, linear algebra, calculus and optimisation, and thinking about how we can leverage AI to accelerate drug discovery.

Young students often struggle to visualise how the study of mathematics may translate into practice. Many believe they’ll end up being a banker, an accountant or a mathematics teacher (which are of course worthwhile professions). I never really planned my career-journey, I did what felt right at the time, and I would never have imagined that I’d end up using my skill-set to find life-changing medicines for patients. So here’s my advice: we’re living in challenging economic times, so be flexible and responsive – seek out and embrace new opportunities that play to your strengths; and most importantly, follow your passion for mathematics – it can take you anywhere!

Posted by HMS in Stories