From Universal Induction to Intelligent Systems

The project will develop a universal theory of intelligent learning systems. Although there is continuing progress in artificial intelligence and machine learning, there is still no comprehensive mathematical theory what learning and intelligence actually are. Based on an unexpected unification of universal induction and sequential decision theory, the CI recently developed in several publications an elegant model that seems to fill this gap. Getting the foundations right usually marks significant progress and maturing of a field. The theory will provide a gold standard and guidance for researchers working on smart software. With the CI's expertise and an excellent team, the full potential of the theory of universal AI shall be explored.