Finn S. Fassbender

I’ve spent the last few years following my own double degree in Medicine and Machine Learning—mostly because I couldn’t pick one, and nobody told me I wasn’t supposed to do both. It turns out that’s exactly where I needed to be.

Somewhere between the hospital and the lecture hall, I realized that the questions I cared about most live in the gap between what we can predict and what we can actually change. In medicine, you’re constantly making decisions under uncertainty, and the stakes couldn’t be higher.

That’s what pulled me into causal inference. Not because I wanted to build better predictors, but because I want to understand whether an intervention actually helps this patient, not just the average one.
The future is inherently unknowable. Every patient is, in a very real sense, an out-of-sample data point. If a model tells me it’s “uncertain” without telling me why (is it because the patient is unusual, the data is noisy, or the situation is just fundamentally hard to predict?) then I can’t actually take a calculated risk. I’m just guessing about my guess.

With personalised medicine, the goal isn’t to ignore all the hard-won evidence from thousands of others and just back my best hunch about the patient. It’s to build a model representation that can sit alongside that evidence—to help me understand how this person in front of me might be similar to, or different from, the populations in the trials I’ve read and past patients I’ve treated.
Electronic health records are essentially a massive log of people trying things and seeing what happens. We just haven’t been good at reading them causally. Yet.
True evidence-based medicine in the age of AI means learning to combine the general and the specific, rigorously.

Outside of research, I spend a lot of time on stuff that probably won’t improve my h-index but might make medicine a little less broken. I co-lead Ankizin, Germany’s largest free, community-owned medical flashcard deck, which started as a frustration with paywalled study materials and grew into something more than 20,000+ students now use regularly.
I’m also the National Officer for Research for the bvmd (the German Medical Students’ Association), where I’m trying to fix doctoral training: Right now, the system expects med students to do PhD-level research without structure, without funding, and often without real supervision. It doesn’t have to be this way.

Medical education feels stuck sometimes. We still train students like the internet doesn’t exist, like the goal is to memorize facts rather than to think. I’m interested in what happens when we shift that, when we treat education as cultivating the competence to ask better questions instead of just recalling facts.

At the end of the day, I’m just someone who wants to help build medicine that’s a little more precise, a little more equitable, and a little more curious. Preferably without the corporate pressures. And with uncertainty estimates I can actually use.