Professor Marcus Hutter

Honorary Professor

2xBSc,MSc,PhD,Habil

CSIT Building (108), N322

marcus.hutter@anu.edu.au

+61 2 612 51605

Bio
Research interests
PhD Students
Courses

Marcus Hutter is an Honorary Professor in the School of Computing at the Australian National University in Canberra, Australia. He received his PhD and BSc in physics from the LMU in Munich and a Habilitation, MSc, and BSc in informatics from the TU Munich. Since 2000, his research is centered around the information-theoretic foundations of inductive reasoning and reinforcement learning, which has resulted in 200+ publications and several awards. His book "Universal Artificial Intelligence" (Springer, EATCS, 2005) develops the first sound and complete theory of AI. He also runs the Human Knowledge Compression Contest (500'000€ H-prize).

Artificial intelligence, Bayesian statistics, theoretical computer science, machine learning, sequential decision theory, universal forecasting, algorithmic information theory, adaptive control, MDL, image processing, particle physics, philosophy of science.

Research Description

During the last 20 years, Marcus Hutter has developed an exciting mathematical model for a reinforcement learning agent behaving optimally in an arbitrary unknown environment - a kind of Super String theory for AI.
The theory (necessarily) integrates and draws from and contributes to many fields: computer science (artificial intelligence, machine learning), engineering (information theory, adaptive control), mathematics (probability, statistics), and also economics (rational agents, game theory), psychology (behaviorism, motivation, incentives), philosophy (inductive inference, theory of knowledge).
While nowadays most AI researchers avoid discussing intelligence, his theory even constitutes a mathematical, objective, non-anthropocentric, sound and complete, direct measure of rational intelligence and a formal though incomputable definition of a super intelligent agent, amenable to rigorous mathematical analysis.
It also allows us to regard existing approaches to AI as effective approximations, thus giving students and researchers a much more coherent view of the field.

Research Outline

"I work on a mathematical theory of general Artificial Intelligence, which can be used to address many theoretical, philosophical, and practical questions regarding super intelligent agents. Generally I'm attracted by fundamental problems on the boundary between Science and Philosophy."