Interview Transcript
Q1: Oh, professor, thank you so much for coming here today. Could you tell us your full name, date of birth, and place of birth?
Agnieszka Wykowska: Thank you so much for inviting me here. It’s a great pleasure. My name is Agnieszka Wykowska. I was born in Krakow, Poland, on the night of July [19], 1977. Thank you.
Q2: And tell me a little bit about your parents, family, and upbringing.
Agnieszka Wykowska: So, my parents are actually also academics, also engineers. My father was a professor at the University of Science and Technology in Krakow, in Poland, and my mother also worked there at the same university.
I grew up in the very first part of my life in a communist country—a different regime there. But then things changed when I was around 12 years old, so I don’t remember very much those “great times” of Poland.
When I was a little kid—around 8 to 10—we moved to the United States. My father had a contract there for two years in Colorado. So I went to an American school there as a kid, and that was something very, very different from what kids back then in Poland usually did. That’s also where I learned English.
When we came back, the system had already changed completely. So from learning Russian in school as obligatory, I could actually learn French and continue English. That was a big change. So yeah, that’s my childhood in a nutshell.
From there onwards, I attended high school in Krakow. At university, I first studied philosophy in Krakow, and then I moved to Munich, where I studied neuro-cognitive psychology.
Q3: Did you go to Munich—at the Ludwig Maximilian University in Munich?
Agnieszka Wykowska: Yeah, LMU.
Q4: What was that like?
Agnieszka Wykowska: I really, really enjoyed my time, both in Poland and later on in Munich. In Munich, the course was very, very intense because it was just a two-year master’s. We studied really all the time and we didn’t have any breaks because we had internships during the summer and the winter. It was an intense but really great course in cognitive neuroscience.
Back in Krakow, studying philosophy gave me a very different background in terms of asking broader questions in science, abstracting away from the small, detailed questions that you typically ask in more empirical research. Having a broader outlook on things was the very basis of my education that I think still resonates in my research today.
Q5: Did you work outside academia after that, or did you go straight through?
Agnieszka Wykowska: No, I continued in academia. I did my PhD and then continued as a postdoc, still in Munich. I stayed in Munich for a very long time and continued with postdoc positions until I moved to Sweden for a very short time, where I got a professorship position. From there, I moved to Italy to the Italian Institute of Technology (IIT).
Q6: How did you get interested in robotics in the first place? Was it popular culture while growing up?
Agnieszka Wykowska: Absolutely not! I mean, maybe there was popular culture, but it was not my interest back then.
It happened while I was in Munich. As I mentioned, I studied cognitive neuroscience and philosophy first. I was interested very much in AI when I was studying philosophy—the topic of natural and artificial intelligence, the Turing test, and all these philosophical questions around it. But when I studied cognitive neuroscience, I didn’t have as much interest in robotics.
What happened after my PhD is that my PhD supervisor was involved in a very large excellence cluster called “Cognition for Technical Systems” (CoTeSys). These are big initiatives in Germany linking different universities together, with many principal investigators working together on large-scale projects. CoTeSys was an initiative between LMU and the Technical University of Munich (TUM).
She offered me a postdoc position in that cluster, which involved engineers, psychologists, and neuroscientists working together. I didn’t really know at first what this would mean—of course I knew robotics from visiting labs—but I felt like, “I’ll give it a try. Maybe it’s going to be interesting to bring our psychology and cognitive neuroscience perspective into robotics.” So I said yes to the offer of a postdoc position. That’s how it started. Then I got totally in love with the theme and the topic, and here I am, many years later, still working on it.
Q7: Do you remember some of the robots from those early projects?
Agnieszka Wykowska: Yeah, I was familiarized with the robots in the CoTeSys project. One specific social robot made at TUM that I worked with and was interested in was named EDDIE. It was very similar to the Kismet robot from MIT. Kismet was sort of the first inspiration to go into those social aspects of robotics.
Q8: What were some of the main research questions you had at that moment in robotics?
Agnieszka Wykowska: When I started in robotics, I was interested in social signals: How do humans attune to robots? What are the natural ways of expressing social signals by robots? How do people receive those robots?
Given my cognitive neuroscience background, I’m interested in human-robot interaction (HRI) more on the human side. For me, the interest is how our brain processes social information from robotic faces, robotic bodies, and various behaviors on the robot.
Q9: How did your research questions evolve over time?
Agnieszka Wykowska: I started with social aspects like gaze. I was very interested in directional gaze, especially when I started supervising a PhD student who was working on that topic.
From there onwards, I went in the direction of attribution of intentionality to robots: What are the behaviors in robots that make us perceive them as intentional agents? In fact, I wrote an entire European Research Council (ERC) project on that topic of the “intentional stance”—adopting an intentional stance toward robots and attributing intentionality to them. That proposal was funded, and it basically changed my entire career. I became an independent researcher with that grant and moved to Italy with it.
I started that topic in 2017, and it’s still present in my research today. More recently, I’ve been moving toward topics like sense of agency, sense of joint agency with robots, cobots, cognitive load, and cognitive offloading. The topics change a bit, but the approach is always the same: trying to understand the human side of a human-robot interaction scenario.
Q10: Have you faced any major challenges throughout your career?
Agnieszka Wykowska: I would say the biggest challenge for this kind of research—and in general for HRI research—is interdisciplinarity. I come from a different background, and while I now have enough experience working with roboticists, it is still a challenge to manage a group where many researchers are cognitive neuroscientists or psychologists. I have to make sure their careers develop in a way that allows them to eventually become independent researchers while staying in this interdisciplinary field.
The typical career path requires focusing strictly on psychology or strictly on robotics, whereas here we are constantly bouncing between the two disciplines. That’s also why I travel a lot for various conferences—today I’m at a robotics conference, tomorrow I’ll be in Rome for a neuroscience conference. We’re in psychology, neuroscience, robotics, and HRI communities all at once. It’s demanding to stay present in all these communities while ensuring team members acquire the right methods and skills needed for this research.
So interdisciplinarity is a big challenge, but it’s very exciting and I like it a lot. I enjoy trying to talk to roboticists and telling them, “Please take into account the human brain, human psychology, and HRI.” On the other hand, when I speak to psychologists, I try to convey the message that robots are great tools to understand our human nature—which is also not trivial for a classical psychology researcher.
Q11: So you have to speak different languages.
Agnieszka Wykowska: Exactly, exactly.
Q12: People are talking about Large Language Models (LLMs) and tools like Gemini everywhere now. Do you see that as a major breakthrough personally?
Agnieszka Wykowska: Definitely for communication, this really changes the game a lot. Things are much easier now to have a natural communication and interaction with a robot compared to even five years ago.
For example, in a project we’re doing looking into cognitive rehabilitation for the elderly, having a natural conversation between a robot and an elderly person five years ago would have been a nightmare. Everything had to be Wizard-of-Oz—we had people typing the responses, the robot speaking them, and huge delays for the user. Now implementing LLMs for communication purposes has become much easier. That’s a total game changer.
Q13: Speaking of funding, who are the major funding sources for your work?
Agnieszka Wykowska: European projects primarily. As I mentioned, I had an ERC Starting Grant, followed by an ERC Proof of Concept Grant. I also have another project funded by the European Innovation Council (EIC), as well as national grants from Italy through the post-COVID Recovery Fund.
My work is primarily fundamental basic research, so I don’t have that many industrial projects. We are currently involved in a semi-industrial project with an institution in Italy focusing on workers and worker well-being (working on cognitive offloading). But most of my research is so fundamental that it’s far from direct industrial applications. On the application side, we focus more on healthcare—cognitive rehabilitation for the elderly and robot-assisted training for children with autism.
Q14: Is there anything you would have liked to pursue but didn’t get the chance to in your career so far?
Agnieszka Wykowska: Well, my career is not over yet! So I still hope that whatever I have in mind I will have a chance to pursue.
One thing that is a huge challenge that I would love to do is more longitudinal studies with robots—for example, with children with autism at home, observing how interactions continue over a long period as children grow up into teenagers and young adults over many years. It would be beautiful to see whether what we observe in short-term interactions translates to longer-term interactions with robots and other humans.
I don’t see that as possible right now given the current state of technology. We can’t really have robots at home yet that interact naturally without technicians constantly being around to intervene. That’s something I would really like to do one day, though I’m not sure if I’ll be able to do it within my career timeframe.
Q15: Looking at where robotics is heading, how do you see this field helping in the next 10 to 20 years?
Agnieszka Wykowska: There is a huge interest right now in robotics and humanoids, and a lot of enthusiasm, so there is a massive opportunity at the moment.
I hope things will go in the direction of having robots as assistants for humans, hopefully at home—which is probably the hardest environment—but at least seeing robots in healthcare institutions as a standard rather than something unique and experimental.
Our work has shown that robot-assisted training for kids with autism works—they improve their social skills. This is a very promising avenue, but it’s still very experimental with only a handful of institutions worldwide using this approach. I hope this becomes a more common, daily experience for children.
Q16: What do you consider the most important contribution of your research in robotics?
Agnieszka Wykowska: If I were to think about the most important contribution, it is indeed bringing closer those two worlds of cognitive neuroscience and robotics/HRI together.
While there were neuro-inspired robotics approaches before, what I mean is bringing cognitive neuroscience methodscloser to HRI research—using cognitive neuroscience methods to understand how the human brain responds to and interacts with a robot, what cognitive mechanisms are involved, and whether they are the same mechanisms involved in human-robot interaction compared to human-human interaction.
I’ve been advocating for bringing these two worlds together my entire career. Years ago at robotics conferences, people weren’t interested in these topics at all. Now I see more and more of these approaches being heard. By the same token, at the beginning I would go to psychology conferences with slides featuring robots and only 3 or 4 people would attend my talk. Five to eight years later, I’d go to psychology conferences and the rooms would be full because people realized that humanoid robots specifically are sophisticated tools to probe how the human brain works in interaction.
Additionally, on the healthcare application side, developing training protocols tailored to specific cognitive mechanisms that can be trained with a robot is something I’m very proud of. Working with vulnerable populations, rehabilitation centers, and hospitals requires a lot of preparatory work to ensure families and parents are comfortable. Seeing kids improve is something I’m extremely proud of what we achieved as a team.
Q17: Have there been real-world applications in medicine stemming from this?
Agnieszka Wykowska: Yes. As I mentioned, we’ve worked with the elderly and with children. In our ERC Proof of Concept project, we focus on adolescents, helping them enter adulthood in a more independent, autonomous manner by training skills needed for autonomous adult life.
This is a population that isn’t well addressed in research—people usually focus either on very small kids or the elderly. Adolescents with special needs are often left alone with issues that empower or hinder them when entering adult life, such as going through a job interview or exam, which is extremely stressful for vulnerable individuals. We train them to get used to an exam situation so they are less stressed, enabling them to pass job interviews. That is a promising real-world application.
Q18: Has anything surprised you when bringing these robots into the real world?
Agnieszka Wykowska: Kids are always surprising! You get all sorts of anecdotal stories about how kids respond to robots.
Members of my team told me an anecdote about a kid saying they would like to invite the robot to their birthday party, but since the robot didn’t have legs, they would wait until it got legs before inviting it (since we were using only the upper torso of the iCub robot). Stories like that are beautiful and funny, but they also show how much kids socially attune with the robot.
In terms of logistical surprises: we designed an entire setup to guarantee safety, placing a transparent barrier between the robot and the kid so no fingers get caught in motors, along with barriers underneath the table and on the sides. The robot was essentially in a cage, but they interacted through a little drawer where they could pass objects back and forth without direct physical contact. You would think with all those barriers there was no way a kid could get to the robot, but kids always find a way! We had cases where a kid jumped right over the barrier to touch the robot before the therapist next to them could even react.
When you work with kids, there are always surprises, which is part of the fun. No matter how much you plan as an adult, there will always be something you didn’t predict.
Q19: What would you recommend or say to students interested in entering your field?
Agnieszka Wykowska: I would say the most important thing is to master the core methods of the field you come from. I am not entirely in favor of overly generic interdisciplinary education at the start. You need to have very strong foundational skills either in robotics or in neuroscience.
Acquire those strong methodological skills during your studies first, and then maintain an openness to learn about new methods and research questions from other disciplines. That combination of strong core methods plus open-mindedness is the key to success in our field.
Q20: Is there anything else you would like to add?
Agnieszka Wykowska: Continuing on that advice for younger people: research is not always easy. There are obstacles we have to overcome creatively. Things take forever—paperwork, bureaucracy, ethics approvals, convincing families, writing grant proposals that get rejected—so you have to get used to rejections and difficulties along the way.
However, doing research is an extremely rewarding human activity. When you have the passion for it, you will find a way regardless of the obstacles. It is so exciting to push the frontiers of knowledge forward, and I am extremely happy that I can do research and be where I am today.
Q21: Thank you so much!
Agnieszka Wykowska: Thank you. Thank you very much.