Public Engagement

In May 2026, our students Iva Janković, Anthos Makris, and Mohammad Kouli took part in Falkirk Science Festival.

They delivered two activities for the children, which allowed them to introduce machine learning at different levels of complexity...

The first was building a decision tree to learn how to classify cats and dogs, which gave an intuitive first idea of how a model can learn to classify objects by asking simple questions and identifying useful features. 

For the second activity, participants learned how to create masks for segmentation of malaria cells that can capture all their variants. This showed how similar ideas become highly relevant in biomedical AI, where models need to handle biological variability and generalise well across different cell appearances. 

"A key takeaway for us was how effective hands-on activities can be in making both the basic logic and the real-world importance of machine learning accessible to a young audience." - Iva

In April 2026, our 2024 cohort organised a sold-out discussion event for Edinburgh Science Festival called 'Health or Hype: AI in Biomedicine'.

The session started with a thought-provoking question: "how comfortable would you be if your doctor was using AI to diagnose and treat you?"

Participants were then split into smaller groups to discuss AI innovations in healthcare at three stations:

• Discover: using unsupervised models (e.g. clustering), can we better understand diseases and their subtypes?
• Diagnose: using vision models, can we better and more efficiently detect diseases in medical images?
• Treat: using Large Language Models (LLM), can doctors choose a more optimal treatment plan for their patients?

The event concluded with a panel discussion, led by professors and clinicians Sara Brown and Charlie Lees, and our programme director Ian Simpson. Participants could ask questions about topics that came up during the event or more general questions about the use of AI in biomedicine.

Participants left with a better understanding of how AI works, as well as specific examples of AI tools used in healthcare to discover, diagnose, and treat diseases. They discussed the benefits and risks of these applications and how they are currently being implemented and evaluated.

When asked again at end of the event “how comfortable would you be if your doctor was using AI to diagnose and treat you?”, participants' opinions had changed for the positive.

In March 2026, as part of the Federation of European Neuroscience Societies (FENS)' Brain Awareness Week, our students Lachin Soufizadeh and Nikoo Moradi delivered a workshop titled “Brains in the Making: Shaping How We See Epilepsy”.

They delivered it to secondary school pupils at Castlebrae Community Campus. The session combined neuroscience with hands-on activities, including building brain structures using Play-Doh.

Lachin and Nikoo expressed that it was very rewarding to see how quickly students engaged with the material and how their understanding of the brain and epilepsy developed over the session.

In January 2026, the University of Edinburgh's Science Magazine published "In The Hallucinating Machine: Why AI Needs to Dream", written by our student Elisa Castagnari.

The article explores why LLMs may hallucinate for the same reason humans dream: to break free from overfitting and spark creativity. 

Read the full article at the link below.

• The Hallucinating Machine: Why AI Needs to Dream 

In November 2025, our student Elisa Castagnari took part in I’m a Computer Scientist, Get Me Out of Here! 

This student-led STEM enrichment activity connects school students with computer scientists through real-time, text-based Chats, helping them discover computing based careers.

Elisa shared with us a bit about her experience: "During the sessions, students were engaged and curious about my work and career path. We discussed how AI turns health data into useful knowledge and whether it could replace humans in healthcare. I emphasized that ethical accountability requires humans to remain in the loop for validation, especially in clinical settings. We also talked about how rapidly evolving technology affects our work. I explained that while tools change quickly, the foundations last longer, and designing modular, adaptable software ensures it remains useful. Finally, I highlighted that adaptability is essential across all engineering disciplines, enabling engineers to innovate and solve real-world problems effectively."

Find out more about I'm a Computer Scientist at the link below.

• I'm a Computer Scientist

In July 2025, our students Núria Fàbrega Ribas, Artur Miralles and Zhijie Yao ran an interactive workshop for 15 students in their penultimate year of secondary school who are considering a Computer Science degree.

As part of the Computer Science & Engineering stream of The Sutton Trust Summer School at The University of Edinburgh, the session introduced students to the world of Large Language Models (LLMs) and Natural Language Processing (NLP).

The workshop began with a presentation covering the basic of LLMs: how they work, how they're trained, and how they're used in real-world applications. Students then applied what they’d learned in a hands-on activity where they used various LLMs and NLP techniques to analyse thousands of real visitor reviews from Edinburgh Zoo.

Working through a beginner-friendly Jupyter Notebook, students explored how to automatically detect animals mentions in text, analyse positive and negative opinions about animals, and create short summaries of the reviews. Along the way, they also got their first experience with programming in Python, seeing how code can be used to work with language and extract meaning from data. To wrap up, students debated how they might improve the zoo experience based on the patterns they observed in the reviews.

The materials for the workshop are freely available at: https://github.com/nuriafari/NLP_workshop_zoo_reviews 

In March 2025, Jamie Davies and Rodrigo Lara Molina gave a talk to students from three Edinburgh secondary schools, about their research projects at AI4BI. This was part of the Widening Participation program at The University of Edinburgh, which focuses on inspiring young people to be life-long learners.

The talk focused on how AI can be used in two biomedical applications: accelerated cardiac MRI, and experimental design to tackle antibiotic resistance.

To explain these projects, they both used analogies accessible to the students:

• Jamie explored mobile phone face recognition, self-driving cars, and text-to-image generation, to introduce computer vision and explain how he is building software for image reconstruction in accelerated cardiac MRI.
• Rodrigo explained how he could frame his project as training a machine to play a specific video game, where the aim is to experiment with bacteria to understand how resistance to antibiotics develops. Thus, the machine would learn how to be a scientific-AI-assistant which can propose the most informative experiments to study antibiotic resistance. 

The students asked a lot of great questions around AI: how models are built, what are they useful for, and when and how they can go wrong. We hope some of these students will end up studying with us at the School of Informatics, although the main objective was to inspire them to learn more and pursue their interests, whatever they might be!

In April 2025, our students Melina Müller and Bianca Branco, alongside other University of Edinburgh PhD candidates Eileen Xu, Christina Steyn and Clara Sanchez-Izquierdo Lozano, ran their own workshop as part of Edinburgh Science Festival, called Spacebound Minds.

The group performed a story in which Mission Control and mental health researchers interviewed an astronaut struggling with her mental health after spending a long time at the space station. The children attending were tasked with designing their own rooms at the space station, thinking about what they need (other than food, water and oxygen) to be happy and healthy on another planet.

The workshop encouraged young people to learn and reflect on mental health and well-being topics. It focused on the themes of sleep, physical exercise, and social contact, delivering (kid-friendly) scientific explanations for how and why these activities impact our mental health.