About
Email: shane@vetta.org
I’m Shane Legg, a post doctoral researcher studying reinforcement learning and theoretical neuroscience with Prof. Peter Dayan at the Gatsby Unit, University College London.
My research career began in New Zealand as an undergraduate working for the WEKA machine learning project at the University of Waikato. For my MSc I moved to the University of Auckland where I worked on universal prediction (Solomonoff Induction) and mathematical provability with the complexity theorist Prof. Cris Calude. In 2000 I moved to New York to work at Webmind, and after this company folded I went to work for Adaptive Intelligence in Los Angeles. In 2003 I decided to return to academia and took up a PhD student position working on theoretical models of super intelligent machines (AIXI) with Prof. Marcus Hutter at the Swiss IDSIA research lab. My PhD research was published in international conferences and journals, as well as being reported in popular science magazines such as Le Monde de l’intelligence and New Scientist.
Upon the completion of my PhD thesis Machine Super Intelligence in 2008, I was awarded the $10,000 Canadian Singularity Institute for Artificial Intelligence Prize. My thesis research on the definition of intelligence is now featured on the main Wikipedia page for Artificial Intelligence. After a year at the Swiss Finance Institute working with Prof. Enrico de Giorgi on mathematical models of cognitive bias in investor behaviour, I was awarded a post doctoral research grant to study abroad by the Swiss National Science Foundation. I’m now working on self tuning temporal difference learning with Prof. Peter Dayan at the Gatsby Computation Neuroscience Unit, which is part of University College London.