Over the summer, I had the privilege of completing my Laidlaw Internship in Action (LIA) at the German Research Center for Artificial Intelligence (DFKI) in Kaiserslautern, Germany. DFKI is one of Europe’s leading research institutions in the field of artificial intelligence, where cutting-edge ideas in machine learning, robotics, and data science are constantly being developed and applied to real-world challenges. During my time there, I worked with a team of researchers and educators to design a curriculum and set of learning materials on machine learning for pre-university students, as part of an upcoming project called the AI Advent Calendar.

The idea behind the AI Advent Calendar was both simple and ambitious: throughout December, students would open one “door” per day, each containing a small task or question related to AI. By solving one problem per day, they would gradually be introduced to the fundamental concepts of machine learning-step by step, and in an engaging, gamified format. The ultimate goal was to make AI not only accessible, but also exciting, to a younger audience who might one day become the next generation of data scientists, engineers, or responsible AI developers.

Designing for Diversity: The Challenges of Building a Curriculum

One of the first challenges we faced was designing content for a highly diverse audience. The target age range for the AI Advent Calendar was broad, encompassing both younger students just beginning to explore programming and older students preparing for university-level studies. This meant we had to consider, very carefully, what concepts could reasonably be expected of the students, and what concepts needed to be introduced from scratch.

Even basic mathematical ideas such as gradients, derivatives, and vectors, which are foundational in machine learning, could not be taken for granted. Some students might already have encountered these topics in advanced mathematics classes, while others might never have heard the terms before. Therefore, we had to constantly ask ourselves: to what extent should we explain these ideas? How much depth is appropriate before we risk overwhelming our audience?

A related challenge arose from the project’s ambition to extend beyond Germany. The AI Advent Calendar was intended not just for local students, but potentially for participants around the world. This brought in the issue of curriculum standardization-or rather, the lack of it. The level of mathematical and computational literacy among pre-university students can vary drastically across countries and educational systems. What a 16-year-old in one country learns in calculus might not be covered until university level elsewhere.

This made our design work a constant exercise in balancing accessibility and intellectual rigor. We wanted to inspire curiosity without making the material seem too difficult or too simplistic. In some ways, this task mirrored the broader challenge faced by AI education itself: how do we make a highly technical and abstract field comprehensible to a general audience, without diluting its essence?

Communicating Complex Ideas: From Precision to Playfulness

One of the lessons I learned early on in the project was that communicating AI concepts is an art as much as a science. As undergraduates trained in mathematics and computer science, we were used to the concise, formal language of equations and algorithms. But this precision, while essential in research, can sometimes act as a barrier when trying to reach younger learners.

Explaining a concept like a neural network, for instance, to a teenager unfamiliar with calculus, requires creativity. We had to move away from the purely technical language and instead think visually and intuitively. What metaphors or analogies could we use? How could we represent abstract ideas in a tangible way? We explored using media such as static coding demonstrations, interactive graphs, and simple animations. Each choice came with trade-offs between interactivity, clarity, and development time.

In many ways, this process reminded me of the famous phrase, “Explain it to me like I’m five.” It forced us to strip concepts down to their core intuitions-without losing meaning. Interestingly, this exercise also deepened my own understanding of machine learning. Having to rephrase, simplify, and visualize ideas made me see them in a new light.

Collaboration was key in overcoming these communication challenges. Our team frequently bounced ideas off each other, critiquing drafts and experimenting with different teaching styles. It was fascinating to see how our approaches differed: some of us preferred minimalist explanations, emphasizing clarity and simplicity; others leaned toward narrative or story-based formats, embedding questions in short tales or real-world scenarios to make them more engaging. Both styles had merit, and finding the right balance between them was an ongoing process of trial, error, and discussion.

What we initially thought would be a relatively straightforward task-writing short daily questions-turned out to be surprisingly difficult. Each question required not only technical accuracy but also pedagogical thoughtfulness. How do you design a task that is solvable yet thought-provoking, that teaches something new yet doesn’t intimidate the learner? By the end of the internship, we had come to appreciate that good educational design demands as much creativity as any form of art.

What to Teach: Balancing Depth, Breadth, and Relevance

Another major aspect of the project was deciding what content to include. In the modern age, AI is a buzzword that evokes both fascination and fear. For many people-especially younger audiences-AI seems mysterious, even magical. We wanted to help demystify it, while also conveying its real potential and limitations.

As a whole, there are three things to explore

1. What AI can do- introducing students to the exciting capabilities of modern AI systems, from image recognition and natural language processing to robotics and generative models.
   
   
2. How AI works- giving them a very basic sense of the mechanisms behind AI, such as training data, pattern recognition, and learning from examples.
   
   
3. The societal and ethical dimensions- encouraging students to think critically about AI’s impact on privacy, fairness, bias, and the future of work.
   
   

We also wanted to highlight that AI is not an arcane “black box” beyond human understanding. While many algorithms are indeed complex, they are built upon understandable mathematical principles-linear algebra, probability, optimization, and calculus. Exposing students to the mathematical side of AI, even lightly, was important to convey that this field is not magic; it is logic, structured reasoning, and creativity combined.

However, as undergraduates immersed in mathematics ourselves, we had to be careful not to overestimate what was appropriate for our audience. It was tempting to dive into technicalities, but the goal was not to turn students into experts overnight-it was to ignite curiosity. The balance we had to strike was between accessibility and authenticity: simplifying without oversimplifying.

Ultimately, we structured the content to progress from conceptual to practical, then to reflective. The early questions focused on building intuitive understanding (“How does a computer learn from examples?”), while later questions delved into small-scale experiments (“What happens when data is biased?”) and ethical discussions (“Should AI make decisions about people?”).

This experience taught me a lot about curriculum design and educational psychology, particularly in STEM outreach. As someone passionate about machine learning, I realized that being able to teach it clearly is just as important as understanding it deeply.

Learning from Researchers and Research Culture

Beyond the curriculum design, one of the most enriching parts of my internship was learning from the researchers at DFKI. Although I had already been exposed to research through my studies at Oxford, that environment is, by nature, highly academic and structured within a university setting. DFKI, by contrast, is a research institute independent of a university, focused on applied AI and collaboration with industry. Experiencing this different model of research firsthand was incredibly illuminating.

When we arrived, the lab was approaching a major paper submission deadline, and the atmosphere was electric. The researchers were working long hours, iterating over experiments, refining figures, and perfecting arguments. It was inspiring to witness the dedication and intellectual drive that goes into producing cutting-edge research. Each researcher brought their own unique strengths: some excelled in technical depth, others in creative ideation or synthesis. It reminded me that research is fundamentally a human endeavor-a collective pursuit shaped by diverse minds and personalities.

The team was also remarkably international, with members from across Europe, Asia, and beyond. Conversations about research often flowed seamlessly into discussions about personal journeys-how they had entered academia, what motivated them, and what they hoped to achieve. Some were PhD students testing whether academia suited them; others were senior researchers who had already made their mark but were still exploring new directions.

This diversity of perspectives was deeply inspiring. It broadened my view of what a career in research could look like, and how one’s relationship with academia evolves over time. I also gained a greater appreciation for interdisciplinary collaboration-how computer scientists, linguists, psychologists, and ethicists all have roles to play in shaping the future of AI.

Outside of work, the researchers were equally supportive and welcoming. They invited us to attend their lectures at both DFKI and RPTU (Rheinland-Pfälzische Technische Universität), where we saw firsthand the variety of teaching styles and academic traditions in Germany. They also made time for fun-one evening, we even went out for laser tag together. These experiences made me realize that professional and personal connections often intertwine, and that strong research teams thrive on camaraderie as much as intellect.

Experiencing Life Abroad: Beyond the Research

Although I am already an international student at Oxford, living and working in Germany offered a different kind of overseas experience. Studying abroad and working abroad are not the same. A workplace immerses you more deeply into local rhythms, habits, and interactions that you might miss as a tourist or even as a student.

Kaiserslautern itself is a charming city, surrounded by forests and filled with a mix of students, researchers, and families. Over time, I began to notice subtle cultural differences-in communication, work habits, and even the rhythm of daily life. Germans, for instance, are precise and structured, yet there’s a strong appreciation for leisure and balance. I admired how even busy researchers would take time for coffee breaks or short walks, reflecting a healthy approach to work-life integration.

While English was widely spoken in the workplace, living there gave me opportunities to practice basic German. I remember the small victories of being able to order food confidently at Fred’s Chicken Van or ask for directions at the train station. These moments, though simple, were deeply rewarding-they reminded me of the value of stepping outside one’s comfort zone.

Beyond Kaiserslautern, I had the chance to visit several cultural and historical sites, including the Völklingen Ironworks, a UNESCO World Heritage Site. The factory’s exhibitions, especially one on African art and industry, were unexpectedly profound. They offered a lens into how technology, culture, and history intertwine-something that resonated with my growing awareness of AI’s global and ethical dimensions.

But some of the most memorable experiences came not from planned trips but from chance encounters. On trains, in cafés, or at research events, I met people from all walks of life-students, travelers, professionals-each with their own story. Some were fleeting connections, lasting only a few train stations, yet they left lasting impressions.

These interactions made me reflect on the universality of human aspiration. Everyone, in their own way, is striving-whether to learn, to contribute, or to find meaning. Coming from Oxford, which can sometimes feel like a self-contained bubble, it was refreshing and humbling to see the wider world at work. It reminded me how much there is to learn simply by listening to others.

Reflections and Takeaways

Looking back, my internship at DFKI was far more than a technical or professional experience-it was a journey in communication, empathy, and personal growth. Designing educational materials on machine learning taught me how to see my field through the eyes of others. It forced me to ask not just what I know, but how I can share what I know effectively.

It also deepened my appreciation for clarity in communication, something I have been consciously working to improve. The ability to distill complexity into simplicity is valuable not only in teaching, but also in research, where conveying your ideas clearly can make or break a project.

Equally, the experience of working alongside passionate researchers showed me what it truly means to pursue knowledge for its own sake. Their commitment, curiosity, and openness left a strong impression on me, and reaffirmed my desire to contribute to research that bridges technology and society.

Finally, living abroad taught me humility and adaptability. It reminded me that no matter how much one learns academically, personal growth often comes from the unplanned, everyday interactions-ordering food in another language, getting lost in a new city, or sharing stories with strangers on a train.

Conclusion

In summary, my LIA experience at DFKI in Kaiserslautern was one of the most formative chapters of my academic and personal journey. It combined the intellectual challenge of teaching complex concepts with the human experience of collaboration, cross-cultural exchange, and self-discovery.

I am deeply grateful to DFKI, Ruby-Anne, the Laidlaw Scholars Foundation, and the Oxford SDG Impact Lab for their support throughout this project. Their guidance made this opportunity possible, and their encouragement allowed me to grow-not just as a student of computer science, but as a communicator, collaborator, and global citizen.