Ericsson, the global telecommunications giant, has unveiled Cognitive Labs, a pioneering initiative designed to accelerate open-source Artificial Intelligence (AI) research. The announcement, made on Tuesday, highlights the company’s commitment to advancing AI technologies while fostering collaborations within the open-source community.
Jean-Christophe Laneri, the Head of Cognitive Network Solutions at Ericsson, explained that the focus of Cognitive Labs would be on innovative technologies like Graph Neural Networks (GNNs), Active Learning, and Large-Scale Language Models (LLMs) — key components in the development of Ericsson’s next-generation solutions.
While the initiative will not have a fixed physical location, Laneri confirmed that it would operate virtually, with teams and activities based initially in key cities, including Madrid, Malaga, and Cairo.
Laneri emphasized that Cognitive Labs will not only enhance Ericsson’s existing technologies but also extend their applications to fields such as healthcare, which could bring substantial benefits to society. The initiative, in collaboration with the open-source community, is expected to attract AI researchers, data scientists, and engineers, all working together to create improved solutions for Ericsson’s customers.
As part of this ongoing commitment, Cognitive Labs will house three distinct research labs:
- Geometric Artificial Intelligence Lab: Focuses on Geometric AI, graph generation, and temporal GNNs.
- Machine Learning and Reasoning Lab: Aims to optimize training models and reinforcement learning.
- Fundamental Artificial Intelligence Lab: Dedicated to creating foundational AI models like LLMs.
Ericsson’s goal is clear: to produce world-class AI research that will not only benefit the academic community but also improve their products to better serve clients worldwide.
Credit: NAN (Text Excluding Headline)
