Samsung AI Center in Warsaw is a place where a unique interdisciplinary AI community thrives, bringing together open and enthusiastic individuals who are conquering the world with their ideas and human-centered AI technology.
In our unit, we believe in human potential supported by AI, which is why we enhance it by developing artificial intelligence. AI that can be carried in your pocket, is safe and can be trusted. We aim to push the boundaries of knowledge, address user needs and creating further conveniences.
We focus on the dynamics of reasoning and expanding knowledge. We work on safety by developing "safety-by-design" and self-regulation concepts, ensuring easy understanding of AI models. We want to enable action planning and improve command comprehension through context-aware prediction.
Join us! Together, we can achieve more!
A graduate of the Warsaw University of Technology, twice a visiting professor at the MIT Department of Mathematics. He conducts research in the areas of numerical and randomized linear algebra and control theory. He develops algorithms for fast machine learning, data compression and reconstruction, and knowledge representation. Author of the theory of meta-factorization, which explains internal reactions between data structures and enables the construction of new matrix decomposition algorithms. He formulated the optimality conditions for energy-efficient control of processes on an infinite horizon and the optimality conditions for Nash equilibrium points in non-cooperative games induced by KKT conditions. A loving father and husband, dancer, and enthusiast of reading and writing.
Dr. Eng. Michał P. Karpowicz
Head of Samsung AI Center Warsaw
Areas we work in
Context & Coherence Lab
How can we understand the context of images, sound, text, and the sensory input we receive from multiple sources? Understanding context is essential for prediction, planning, reasoning, reducing imagination, and preventing hallucinations. Without context, it is difficult to explain anything. We aim to teach AI models to comprehend context to assist them in making predictions, creating coherent and meaningful content, and taking logical actions accordingly.
Knowledge & Reasoning
We aim to understand the process of continuous knowledge creation and supplementation, long-term memory retention, discovering relationships between observations, and building world models that drive imagination and reasoning. We want to comprehend and shape the dynamics of these processes, ensuring their stability and the speed of convergence. Ultimately, we seek to translate these discoveries into technologies that enable answering complex questions posed in natural language, supporting reflection, planning, and drawing conclusions.
Safety & Alignment
We aim to understand the process of continuous knowledge creation and supplementation, long-term memory retention, discovering relationships between observations, and building world models that drive imagination and reasoning. We want to comprehend and shape the dynamics of these processes, ensuring their stability and the speed of convergence. Ultimately, we seek to translate these discoveries into technologies that enable answering complex questions posed in natural language, supporting reflection, planning, and drawing conclusions.
