The Female Frontier: Advancing AI and Women's Journey in Science
In science and technology, few fields are as dynamic and impactful as Artificial Intelligence (AI). It's a field where innovation meets practicality, pushing the boundaries of what's possible and reshaping our future. As we celebrate the International Day of Women and Girls in Science on February 11, we are thrilled to present an exclusive interview with Mateja Jamnik, Professor of Artificial Intelligence in the Department of Computer Science and Technology at the University of Cambridge, UK. Exploring her journey in the world of science, her current research, and the challenges and triumphs she has faced as a woman in STEM, will help us shed light on the cutting-edge developments in AI but also echo a powerful message on the importance of diversity and inclusion in science.
What inspired you to pursue a career in science, and how did you get started in your field of study?
The beginning was when I was a kid at school, and I really loved doing math. I had really inspirational teachers who presented it in a way that it was fascinating. I saw the beauty and elegance in problems and the solutions, as well as finding solutions myself. As I continued with my education, I started to also be interested in people and how people come up with these kinds of beautiful and elegant solutions. I wanted to computationally model these systems, trying to figure out if we could make systems do this kind of intuitive thinking on machines, and that's how I ended up going from studying that into a science masters and doing a PhD in Artificial Intelligence.
With your professional focus on the exciting field of Artificial Intelligence, could you describe your current research interests and projects?
I'm really interested in how people think and solve problems, and I'm trying to build computational systems that can do it in the same kind of way. So I model this on AI systems in terms of enabling them to find intuitive solutions, giving a human explainable reasons for their solutions, and being able to navigate and be flexible and adapt to the user in a way that is amiable to the human user. I'm also looking at how there is this huge wealth of human knowledge that is digitally recorded and trying to see whether machines can capitalize and use this human informal knowledge to make machines more capable.
What aspects of your work do you find most exciting, promising or rewarding in today’s world, especially in advancing the explainability of AI systems?
Out of these four things that I mentioned, one of them is explainability, and that's driven by the fact that we have this prolific success of machine learning methods in AI, and especially deep learning methods, but these are inherently black boxes. So they get an input, they predict a solution or an answer, but they don't give us reasons for it. We want to know the reasons why these solutions have been constructed by an AI system to help us make better decisions. So, we need explainability methods that will break this black box paradigm and make deep learning methods inherently explainable. We're pioneering new techniques for that.
Have you faced any unique challenges or obstacles as a young female scientist and how have you navigated those challenges?
I guess I was brought up in a system where the disparity between men and women in terms of science, in terms of engagement was not as strong as it is, for example, in the UK. And so I always thought that I could do it, and I haven't really been obstructed by these stereotypes. But it certainly influences you when you have no female lecturers in your entire undergraduate degree. For example, it makes you feel like you are an exception and you have to work harder, but I never let that stop me. So I wanted to set an example. That's also why I started women@CL, the Women in Computer Science Research National Network in the UK. This programme of activities has been replicated across the UK in universities and other institutions. It helps to inspire young women and retain those who already are working in computer science.
Considering the dynamic and changing nature of AI, how do you stay up-to-date on the latest developments in your field?
It happens day to day that new findings are published. First of all, I have an amazing team of people that I work with: PhD students, postdocs and collaborators. So I see them regularly, and we all read these papers and discuss and inform each other of the latest findings that are relevant to our research. I have students who come to me and say, "I just came across this work that was published yesterday on the topic that we've been looking in this direction; perhaps we can use it." Basically, it requires time, and it requires more than just my own effort; it's a team effort.
How has your perspective on the potential and limitations of AI evolved throughout your career, and what do you hope to achieve or discover in your future research?
I think there has been a shift in the perception of AI. AI was very unpopular when I was doing my PhD, known as the AI winter, and there was no funding, among other issues. But then a few technological advances broke through because they demonstrated its potential. Suddenly, these machines were able to do amazing things, and people became scared, wondering if they were going to take over the world in Terminator-style affairs.
Now, I think that there's still some of that fear around, but the questions have shifted or changed. People are accepting that this is part of life and will always be because it's an enabling technology, not just in computer science. It's enabling all other technologies and fields to progress rapidly. So, we have to embrace it, and we are, even in our day-to-day lives. We're embracing it in the tools on our smartphones and devices, and in home appliances. So now the questions have shifted towards how we can ensure privacy, security, ethics, and capitalize on this new technology in ways that will benefit humanity in general.
What role do you see science, or AI, playing in addressing some of the world's most pressing challenges, such as climate change, public health, and social justice?
It's already playing a role. For example, AI systems are deployed in courtrooms, for instance, in recommending sentences for people who have been found guilty. Doctors are using AI systems to help them make better diagnosis and treatment decisions. We work, for example, with oncologists, especially looking at and working with breast cancer specialists, and helping them build new AI tools to aid in this process. In our department, we also have an extensive group researching the use of AI for addressing climate change challenges. So yes, absolutely, they are already helping. it's already been deployed, and as we do more research and find more solutions, I think it's just going to continue to be even more present and helpful.
How do you mentor the next generation of scientists to prepare them for future developments and opportunities in AI?
I mentor a lot of students, of course. I teach undergraduates, supervise master's students, and I have a large team of PhD students and postdocs. My main goal is basically to help them become independent researchers who will become the world experts in the field and topic that they're very excited and passionate about, and who will be able to communicate their science not just to specialists but to the general public as well, and that they will be conducting this work and research in scientific and transparent ways.
How do you successfully collaborate with colleagues and other scientists, and what strategies have you found most useful for building strong working relationships?
There are two aspects to this. One is that my work is inherently interdisciplinary. In my day-to-day work, I collaborate closely with cognitive scientists, educationalists, medics, and mathematicians, and in fact with lawyers as well. The main thing that helps, or the main first challenge that we have to overcome in order to establish a successful collaboration, is that we have to be able to communicate because we use different technical terminology. We have different expertise. So we really have to learn from each other, listen to each other, and try to understand where we're coming from. Another aspect is that I think we also have to communicate our science to the general public, and we are collectively responsible for the image or the perception that AI holds out there in the public.
How important do you believe it is for scientists to attend scientific conferences, on-site or virtually, for the purpose of communication and collaboration?
In computer science, which is a fast-moving field, we find that the most respected and top venues for publishing our results are conferences and conference proceedings. Journals and such take too long for such a fast moving field. There are many conferences that present new cutting-edge research. This is our normal mode of operation: we go to conferences to present our work, but most importantly, to network with others and to learn about new work and establish new collaborations.
As a professor, do you incorporate multimedia resources, such as scientific videos, into your teaching strategies?
Yes, we do use multimedia resources to supplement our teaching. They're there for me to explain how it works, so I definitely integrate my material with other resources, including videos. I tell Master and PhD students to take advantage of such resources, for example, there are courses for scientific writing, presenting and discussing their research. So, we supplement all the material we prepare with useful online resources, including videos. Our students are very ambitious and seek out these extra resources, which we provide for them.
Given your extensive career, is there any professional moment that has been particularly impactful or memorable so far?
There's not just one moment that has been particularly impactful or memorable in my career; they all have their impact. However, a couple of moments are dear to my heart. One is being in my office with PhD students, discussing the depth of their ideas. During these discussions, when I ask them to clarify their thoughts, you can see them coming to a realization, a moment of clarity that helps them answer their own questions. This motivation and creativity in the new generation of my students are incredibly rewarding. I tell them that by the time they finish their PhD, they will be the world expert, far better than me, in their chosen topic.
Another significant moment was when I was doing a TEDx talk, which influenced all my subsequent teaching. The training on how to make science accessible had a profound impact. In terms of technical work, the things we're doing on explainability and deploying with oncologists in practice to help specialists make better decisions, is really exciting.
What advice would you give to young women who are interested in pursuing a career in science, technology, engineering, or math (STEM) fields?
I would say go for it, follow your dreams. You need to be excited about the work you're doing; it should almost feel like a hobby, and there are no barriers that you can't overcome. One important piece of advice I would offer is to always surround yourself with supportive people and find a mentor, or perhaps several mentors if they specialize in different aspects of your journey, who will encourage you and be there for you to ask for advice and support you while you are going through this process. It really helps!
Share this article