I am a post doctoral researcher at the Swiss Finance Institute working on portfolio selection, in particular, I am developing simulations to study the effect of deviations from strictly rational behaviour.
I recently defended my PhD thesis in artificial intelligence that I wrote under the supervision of Marcus Hutter at the Dalle Molle Institute for Artificial Intelligence (IDSIA). My thesis concerns the AIXI model of artificial super intelligence which is based on Kolmogorov complexity theory, Solomonoff induction and sequential decision theory. In particular, my thesis considers the performance of AIXI in different classes of environments, its connection to definitions of intelligence, a formal definition of machine intelligence, and limitations on computable systems due to Gödel incompleteness. I also study and/or publish in reinforcement learning, genetic algorithms, theoretical neuroscience, complexity theory and algorithmic probability theory.
In January 2009 I will start a two year post doctoral research fellowship at the Gatsby Unit for machine learning and theoretical and computational neuroscience, which is a part of University College London. I will be working on temporal difference learning and hierarchical temporal generative models. I believe this to be the key technology required to build a practical artificial general intelligence.