I am Florian Dorner, a doctoral fellow with the Max Planck ETH Center for Learning Systems. I hold an MSc. in Mathematics from FU Berlin and an MSc. in Science, Technology and Policy from ETH Zurich.

My work aims at a better understanding of the societal impacts of Artificial Intelligence and how policy and technical research can interface to shape these impacts.


Algorithms and Policy

I aim at better understanding of how advanced algorithmic systems and Artificial Intelligence interact with society. In particular, I am interested in how this interaction shapes society and how policy can positively influence algorithmic systems' societal impacts.

My experience in machine learning combined with my studies on technology policy and the social sciences puts me in a unique position to integrate relevant perspectives from different fields, as exampled by my work on algorithmic collusion.

Machine Learning

My most recent work centered on fairness in Machine Learning with a focus on connections to adversarial robustness and learning from human feedback.

Before this, I worked on Reinforcement Learning (RL) and have written a Master's thesis on exploiting modularity in RL as well as a review paper on trends in RL data efficiency. I also helped building smart traffic lights deployed at real crossings at CertaintyLab.


I care about improving both personal and institutional forecasting abilities to enable better decision making.

I am especially interested in how to evaluate the quality of probabilistic forecasts, especially before they fully resolve, or if they are partially self-fulfilling.

I sometimes participate in forecasting tournaments and am a certified Superforecaster with Good Judgment Inc. I have also published on Forecasting progress in Artificial Intelligence.