Keyan Ghazi-Zahedi

Complexity Postdoctoral Fellow

Postdoctoral Fellow, Max Planck Institute for Mathematics in the Sciences

In recent years, it has become evident that a general theory of intelligence requires a better understanding of the interplay of brain, body, and environment. Biology provides us with numerous examples of functions that are normally attributed to the brain but result, at least partially, from the interactions of the body with the environment. Two examples are insect’s eyes and the human hand. The receptive fields of many insect’s eyes are laid out in such a way that they significantly reduce the burden for the brain to detect motion. The softness of the human hand and the friction of its skin allow us to grab fragile objects without the need to precisely control the position of each finger. In both examples, computation is, in one way or another, outsourced from the brain to the morphology. This has led to the notion of Morphological Computation.

Morphological computation seems to be a fundamental principle in biology. Unfortunately, there is no theory that accounts for the different kinds of morphological contributions to intelligence. Finding such a theory is what drives Keyan’s research. In particular, he works on quantifications of morphological computation. A important question in this context is the following: is it possible to record a behavior and calculate from observations alone, how much of the behavior is actively controlled by the brain and how much of it results from morphological computation? He is currently applying preliminary results of his theory to quantify the naturalness of human motion. This will be used to support the visual diagnostics of patients with neuro-degenerative diseases. His research is currently also applied in the context of biomechanics to measure the contribution of muscles to various human motions. During his time at the SFI, Keyan extended his theory to other complex systems such as social groups. In this context, one can ask how much of an individual’s behavior is controlled by its environment. This is in many ways analogous to quantifying morphological computation.

Keyan currently is a senior researcher at the Information Theory of Cognitive Systems Group at the Max Planck Institute for Mathematics in the Sciences and a lecturer at the University of Leipzig. He conducted his Ph.D. project at the Fraunhofer Institute for Autonomous Intelligent Systems and earned his degree in Neuro-Informatics from the University of Osnabrück. He received his MSc in Computer Science from the University of Tübingen.