Based at AMLab (University of Amsterdam)
We advance the foundations of Ideal Machine Intelligence. We view AI not as an artificial mind, but as a representational instrument — an interface between the abstract and the concrete. Our research focuses on building systems grounded in the fundamental laws of nature and geometry, enabling AI to better comprehend the world while helping us make sense of it.
"Matter and mind are not disconnected, but inextricably related. The laws of nature, math, and the abstract manifest in our observations of the physical universe."
Whether viewed through a metaphysical lens or a practical one, our core premise remains the same:
Artificial Intelligence needs to reason about data that is fundamentally grounded in our
shared reality.
Whether mental (mathematics, language) or physical (images, scientific data), all data is a
representation of phenomena
taking place in this reality. For an AI to truly reason, it must recognize the fundamental laws
governing the
processes that generate this data. By grounding our models in this intrinsic geometry, we build the
necessary
bridge between raw observation and conceptual understanding.
Since the structure of reality is geometric, the representations AI builds of it must respect
that geometry: equivariance is not optional, but essential.
The Philosophical Stance. Mind and matter are aspects of one reality, governed by the same geometric structure. AI, on this view, is not a mind but a representational instrument — and like any instrument, it must be structurally compatible with the reality it represents and the minds that use it.
The Computational Paradigm. We embed the symmetries of physics (equivariance) and manifold structures into our architectures, ensuring that learned representations remain mathematically consistent with the physical world.
Scientific & Practical Impact. Our research is driven by critical use-cases where grounding is essential: distinguishing signal from noise in scientific discovery (e.g., computational chemistry) and ensuring reliability in medical imaging and robotics.
Mathematical foundations for symmetry-preserving networks.
Continuous function representations grounded in geometry.
Robust analysis of biological structures and diagnostics.
Stochastic generation on non-Euclidean manifolds.
Atomic point clouds and molecular simulations.