We want to understand how intelligent information processing is possible in dynamical systems with (some) brain-like properties. This aspiring quest pulls us in the middle of machine learning, AI, computational neuroscience, theory of computing, and the freshly re-emerging field of non-digital neuromorphic (wide sense) hardware. We are mathematically minded and develop new modeling formalisms and algorithms. Our main pursuits are

Echo State Networks, a recurrent neural network paradigm which blends naturally with unconventional substrates

Conceptors, a method for modulating the dynamics in a neural network (or some other nonlinear substrate) such that a single device can be switched between different functional behaviors

Non-digital computing theory, working toward a fundamental theory of computing in brainlike materials that could replace the Turing paradigm (with a humble mind, being aware of the majestic proportions of this task)

Observable operator models, a fundamental re-formulation of stochastic systems