Université de Sherbrooke

Published purpose

This project involves foundational research for developing quantum algorithms for classically intractable tasks in artificial intelligence (AI). Quantum physics principles and phenomena discovered in the last decade will be explored for their use in computationally challenging tasks in AI. Specifically, the project will embed primitive problems for these tasks into quantum circuits with thermalizing quantum dynamics in constrained Hilbert spaces and develop techniques to characterize the dynamical properties of these quantum systems. Using these physical principles, quantum algorithms will be formulated for complex problems in AI, namely probabilistic inference and automated planning, and the quantum resources required to tackle these will be estimated.